Keywords: Networked control systems
Abstract: Networked and robotic systems in emerging applications are required to operate safely, adaptively, and degrade gracefully while coordinating a large number of nodes. Distributed algorithms have consolidated as a means for robust coordination, overcoming the challenges imposed by the limited capabilities of each agent. However, plenty of problems still exist to break down the barriers of fast computation, make effective use of measured data, and understand large-scale limit effects. In this talk, I will present ongoing work in the control of infrastructure networks and large-swarm coordination, along with a discussion on modeling approaches, analysis tools, and architectural trade-offs going from small to large-sized robotic networks.

Keywords: Networked control systems, Linear systems, Stability of linear systems
Abstract: Fixed eigenvalues {i.e., fixed modes} present obstacles to the decentralized stabilization and decentralized spectrum assignment of a multi-channel linear system, because they will appear in the closed-loop spectrum of the system with any given finite-dimensional linear time-invariant decentralized control. This poster introduces a simple approach to fully avoid the fixed eigenvalues of a jointly controllable and jointly observable multi-channel linear system using two techniques: state space extension and distributed control. State space extension means to add auxiliary integrators to channels and distributed control is to allow the passing of appropriate data among neighboring channels. The type of data to be transmitted is explicitly characterized. If the neighbor graph associated with the multi-channel system is strongly connected, this approach enables us to construct an augmented multi-channel linear system which has no fixed eigenvalues, regardless of bounded transmission delays.

Keywords: Communication networks, Networked control systems, Multivehicle systems
Abstract: This poster treats the problem of distributed planning for communication-aware information gathering by networked mobile robots. One of the major challenges with distributed planning is how to make a convergence of optimized planning under unknown or coupled decisions by a single robot. This poster uses a distributed sequential approach to tackle the issue. The key idea of the planning scheme is each robot makes its decision based on the decision results of other robots higher in a given hierarchy. The utility considered in a single robot is formulated as expected information gain that considers sensing and communication. To derive information measures for communication-aware information gathering, mutual information was applied when measurements were sent over a packet erasure channel network. Computing the mutual information under nonlinear and non-Gaussian properties of targets, sensing, and communication at the planner phase is another challenging issue. This poster presents sampling procedures using a specific measurement set and particle methods to calculate mutual information. A series of simulations from a scenario of RF emitters tracking using a team of fixed-wing small unmanned aircraft systems (sUAS) describe the presented approach can improve estimate accuracy and information gain compared with another planning approach that assumes the perfect communication. Simulation results also show the presented approach yields better results than other approaches that simplify the communication model or communication-aware information gain. Finally, this poster describes the ROS/Gazebo-based simulation implementation and flight experiments using a team of real fixed-wing sUAS.

Keywords: Energy systems, Embedded systems, Machine learning
Abstract: The capabilities of supervised learning for capacity fade estimation in battery packs have been progressing in applicability and relevance over the last decade. Concurrently, hardware developments for on-board vehicular battery management systems are becoming more and more capable with enhanced memory and power management. This has paved the way for implementation of complicated algorithms incorporating data-driven technologies, especially for data intensive tasks such as capacity fade estimation. Leveraging this expanding capability, this poster presents an implementation of a capacity fade estimation approach refined for application on a single-board computer appropriate for on-board application in a vehicle. The techniques incorporate clustering and a feedforward network based learning system, with implementation achieved on a versatile Raspberry Pi platform. The data used in training and validation of the implementation was generated in-house for an off-the-shelf battery pack.

The algorithms employed in this work consist of training and prediction on a multiple layered neural network, employing clustering techniques on reduced amounts of data for further data reduction. Although any suitable microcontroller could be utilized for this work, the simple Raspberry Pi platform was chosen for proof of concept. Moreover, the capabilities of this platform provide an ideal foundation for the next phase of the work which will employ cloud connectivity for an on-board application in an existing electric vehicle project.

Keywords: Neural networks, Formal verification/synthesis, Model Validation
Abstract: Neural networks have been showing impressive ability to solve complicated tasks including modeling complex nonlinear dynamical systems considered in this project. However, due to the vulnerability of neural networks against adversarial disturbances or attacks and the black-box nature of neural networks, such neural network models are only restricted to the applications with the lowest levels of requirements of safety. This project addresses the reachability analysis and safety verification problems for a class of feedforward neural network models of nonlinear dynamical systems. An interval-wise reachability analysis results of feedforward neural networks are proposed in the framework of interval arithmetic. An efficient simulation-guided method is developed to compute over-approximations of the output set of the feedforward neural network of interest. Then, by recursively using the simulation-guide method, the reachable set over-approximation of the neural network model of a dynamical system can be obtained over a finite-time interval. The safety verification of neural network models can be performed by checking the emptiness of the intersection between reachable set over-approximation and unsafe regions. An example of Maglev system is provided to illustrate the effectiveness of the developed approach.

Keywords: Control software, Linear systems, Optimal control
Abstract: We introduce an open-source MATLAB toolbox for System Level Synthesis (SLS), a novel framework for robust and optimal linear controller synthesis in the distributed setting. As the systems we seek to control become larger, considerations such as communication delay, actuation delay, actuation sparsity, and disturbance containment become key constraints in the controller synthesis problem. The SLS-MATLAB toolbox includes solutions for these constraints, as well as additional features for distributed MPC, robust controller synthesis, actuator co-design, and controller refinement. Further, it contains built-in simulation and visualization tools for easy representation of system responses to arbitrary disturbances. The SLS framework is used in a number of ongoing works in the control community; these include interfacing robust control with learning, nonlinear distributed control, and distributed adaptive control. With this toolbox, we hope to make the SLS framework more accessible to members of the control community as both a ready-to-use tool and a base upon which new controller synthesis algorithms can be built.

Keywords: Optimal control, Filtering, Kalman filtering
Abstract: There is a wide variety of applications that require sorting and separation of micro-particles from a large cluster of similar objects. Existing methods can distinguish micro-particles by their bulk properties, such as their size, density, and electric polarizability. However, those methods are not selective with respect to the individual geometry of the particles. In this work, we describe the use of resonant amplification of light propelling forces to achieve high selection sensitivity with respect to the radius of the particle, through the use of Whispering Gallery Mode (WGM) resonances of fluid-suspended dielectric microparticles. We demonstrate that the evanescent field around a tapered optical fiber fed with sim20 mW power from a 1064 nm laser can selectively move polystyrene microspheres of up to 20 mum in diameter through distances of more than 50 mum. Since WGM force allows the entire procedure to be controllable and tunable by adjusting the direction and the power level of laser, we explore a simulation scheme of a WGM-based device for micro-particle separation. In this design, particles flow in through an inlet and leave through an outlet. Particles will pass over two actuation regions provided by waveguides carrying laser of a specified wavelength to generate the evanescent field. A camera is used to observe particles such that a feedback control can be applied to the power of laser. Challenges in this proposed structure include unknown disturbances to the fluid flow, limited laser power, and uni-directional control over each actuation region. We then combine Expectation Maximization with Kalman filtering to both estimate the unknown disturbance and filter the measurements into a position estimate. Hybrid controllers are developed and compared with regular setting of a Linear–Quadratic–Gaussian (LQG) controller in the simulation.

Keywords: Autonomous systems, Nonholonomic systems, Simulation
Abstract: In autonomous vehicles (AVs), several studies have been conducted, such as driving, path planning, artificial intelligence (AI), and decision making. The AVs tracking control is required efficient and robust controllers to conduct large computations loads such as AI and decision making. First, in the case of path planning, a rapidly-exploring random tree has a large amount of calculation, so it is not easy to apply it to a real AVs. You need to create a vehicle's moveable path and create a path that requires less computation, so you can spend more time to have environmental recognition and making decisions. Besides, some researchers have studied linear controllers, sliding mode control, and model predictive control (MPC) as AV controllers. Among those controllers, MPC is the most common controller. However, since AVs have the nonlinear dynamics, a controller using nonlinear MPC (NMPC) also appears instead of using a linear MPC. To calculate the NMPC, we generate much computational load every sampling time. This is not suitable for path planning and controllers to proceed with AI algorithms. Therefore, we introduce a more straightforward, robust, and optimal method for optimizing the movement of AVs. In this paper, we introduce path planning using polynomial and tracking control based on PID. The SPP curve generator not only takes into account the movement of the vehicle but also uses the polynomial equation. The polynomial equation makes it faster than other curve generators. Also, we design a vehicle position control and vehicle velocity control based on a PID controller. The solution combined with SPP curve generator and PID generates not only a robust controller but also moving paths in AVs, allowing sufficient time for processing not only sensor data but also AI and decision making. Therefore, it is thought that it will be able to guarantee computation load and robustness in the future research of AVs field.

Keywords:
Abstract: We envision a future of zero crashes, zero emissions and zero congestion, and we have committed ourselves to leading the way toward this future. General Motors has been pushing the limits of transportation and technology for over 100 years. Today, we are in the midst of a transportation revolution. And we have the ambition, the talent and the technology to realize the safer, better and more sustainable world we want. As an open, inclusive company, we’re also creating an environment where everyone feels welcomed and valued for who they are. One team, where all ideas are considered and heard, where everyone can contribute to their fullest potential, with a culture based in respect, integrity, accountability and equality. Our team brings wide-ranging perspectives and experiences to solving the complex transportation challenges of today and tomorrow. At General Motors, innovation is our north star. As the first automotive company to mass-produce an affordable electric car, and the first to develop an electric starter and air bags, GM has always pushed the limits of engineering. We are General Motors. We transformed how the world moved through the last century. And we’re determined to do it again as we redefine mobility to serve our customers and shareholders and solve societal challenges.

For additional information see https://www.gm.com/

Keywords:
Abstract: The MATLAB and Simulink product families are fundamental applied math and computational tools at the world's educational institutions. Adopted by more than 5000 universities and colleges, MathWorks products accelerate the pace of learning, teaching, and research in engineering and science. MathWorks products also help prepare students for careers in industry worldwide, where the tools are widely used for data analysis, mathematical modeling, and algorithm development in collaborative research and new product development. Application areas include data analytics, mechatronics, communication systems, image processing, computational finance, and computational biology.

For additional information see https://www.mathworks.com/

Keywords:
Abstract: Mitsubishi Electric Research Laboratory (MERL), located in Cambridge, MA, is the North American R&D organization for Mitsubishi Electric Corporation, a 40B global manufacturer of electrical products including elevator and escalators, HVAC systems, electrical power systems, satellites, factory automation equipment, automotive electronics and visual information systems. Controls researchers at MERL collaborate with corporate R&D laboratories, business units in Japan and academic partners around the world to develop new control algorithms and control technologies that extend the performance envelope of these systems.

For students who are interested in pursuing an exciting summer of research, please check out our internship program and learn more at facebook, google, or @MERL_news.

MERL interns work closely with top researchers, and gain valuable industry experience – an impressive 1:1 intern to researcher ratio.

Internships are expected to lead to publications in major conferences and journals.

We offer competitive compensation and relocation assistance.

Boston is a fantastic student-oriented city, home to some of the best universities in the world. The summer season is especially lively as MERL and Boston are teeming with interns and visitors from all over the world.

Keywords:
Abstract: Quanser is the world leader in mechatronics, robotics, and control platforms optimized for the academic setting. Our leadership in producing innovative lab solutions makes us a trusted partner with academic institutions to help strengthen their reputation with transformative research and teaching labs. The Quanser approach of innovation, collaboration and education has produced a number of notable technology firsts that pioneered many critical contemporary trends, including efficient validation platform for control research, and high-performance real-time control on common microcomputers.

For additional information see https://www.quanser.com/

Keywords:
Abstract: SIAM publishes textbooks and monographs in print and electronic format. Visit our booth to browse new and bestselling titles, all available at discounted conference pricing. If you’re interested in writing a book, an editor is available to explain how SIAM partners with authors to publish books of outstanding quality and accessible pricing.

More info: https://www.siam.org/Publications/Books

Keywords:
Abstract: Silver Sponsor: Cancelled

Keywords:
Abstract: dSPACE offers universities and research institutions flexible systems that provide all the options necessary for the model-based development of mechatronic controllers in an academic environment. From architecture-based system design and block-diagram-based function prototyping to automatic production code generation and hardware-in-the-loop (HIL) tests, dSPACE products are successfully being used in the classroom and in research projects at internationally renowned universities. To actively support high-end research at universities and the high-quality education of young talents, dSPACE offers its hardware and software products in special kits for universities at a very attractive price. Learn more at dspaceinc.com / offers for universities.

For additional information see https://www.dspace.com/en/inc/home.cfm

Keywords:
Abstract: At Springer Nature, our aim is to advance discovery. For over 175 years, we’ve dedicated ourselves to the academic community, creating value across the publishing process. We deliver an unmatched breadth and depth of quality information which spans top research publications (Nature), outstanding scientific journalism (Scientific American), highly specialized subject-specific journals across all the sciences and humanities, professional publications, databases, and the most comprehensive portfolio of academic books. We use our position and our influence to champion the issues that matter most to the research community – standing up for science, taking a leading role in open research, and being powerful advocates for the highest quality and ethical standards in research.

For additional information see https://www.springer.com/gp/authors-editors

Keywords:
Abstract: Processes (ISSN 2227-9717) provides an advanced forum for process/systems related research in chemistry, biology, materials and allied engineering fields. The journal publishes regular research papers, communications, letters, short notes and reviews. Our aim is to encourage researchers to publish their experimental, theoretical and computational results in as much detail as necessary. There is no restriction on paper length or number of figures and tables.

Experimental, theoretical and computational research on process development and engineering

Chemical and biochemical reaction processes

Mass transfer, separation and purification processes

Mixing, fluid processing and heat transfer systems

Integrated process design and scaleup

Process modeling, simulation, optimization and control

For additional information see https://www.mdpi.com/journal/processes

Keywords:
Abstract: Founded in 1919, Halliburton is one of the world's largest providers of products and services to the energy industry. With 60,000 employees, representing 140 nationalities in more than 80 countries, the company helps its customers maximize value throughout the lifecycle of the reservoir – from locating hydrocarbons and managing geological data, to drilling and formation evaluation, well construction and completion, and optimizing production throughout the life of the asset. Halliburton’s technology organization provides cutting edge research and innovative solutions to maximize asset value for our customers.

For additional information see https://www.halliburton.com/en-US/default.html  

Keywords: Optimization, Human-in-the-loop control, Networked control systems
Abstract: We analyze a human and multi-robot collaboration system and proposed a method to optimally schedule the human attention when a human operator receives collaboration requests from multiple robots at the same time. We formulate the human attention scheduling problem as an integer optimization problem which aims to maximize the overall performance among all the robots, under the constraint that a human has limited attention capacity. We first present the optimal schedule for the human to determine when to collaborate with a robot if there is no contention occurring among robots' collaboration requests. For the moments when contentions occur, we present a contention-resolving Model Predictive Control (MPC) method to dynamically schedule the human attention and determine which robot the human should collaborate with first. The optimal schedule can be then determined using a sampling based approach. The effectiveness of the proposed method is validated through simulation that shows improvements on robot performance.

Keywords: Optimal control
Abstract: The Pure Pursuit strategy is ubiquitous both in the control literature but also in real-world implementation. In this paper, we pose and solve a variant of Isaacs' Two Cutters and Fugitive Ship problem wherein the Pursuers' strategy is fixed to Pure Pursuit, thus making it an optimal control problem. The Pursuers are faster than the Evader and are endowed with a finite capture radius. All agents move with constant velocity and can change heading instantaneously. Although capture is inevitable, the Evader wishes to delay capture as long as possible. The optimal trajectories cover the entire state space. Regions corresponding to either solo capture or isochronous (dual) capture are computed and both types of maximal time-to-capture optimal trajectories are characterized.

Keywords: Optimal control, Uncertain systems, Process Control
Abstract: We consider the problem of optimal feedback control of unknown dynamic systems. Optimal feedback control is important for systems with time-varying initial conditions, which cannot rely on feedforward controllers. For example, components in particle accelerators are turned on and off hundreds of times per second, with pulse widths of ~1 ms and repetition rates >100 Hz, whose dynamics and initial conditions vary slowly over time due to external disturbances, such as temperature fluctuations. Our approach allows us to quickly learn an optimal feedback control, which can then be applied for all initial conditions even as the system begins to drift and change. For linear systems we reproduce the cost minimizing linear quadratic regulator (LQR) optimal controller that could have been designed had the system been known.

Keywords: Optimal control, Decentralized control, Optimization algorithms
Abstract: We study the design of an optimal static decentralized controller with a quadratic cost and propose a variant of the homotopy continuation method using a damping technique. This method generates a series of optimal distributed control (ODC) problems via a continuous variation of the system parameters. Diverging from the classical theme of tracking a specific trajectory of locally optimal controllers for these ODC problems, we focus on the possibility of leveraging local-search algorithms to locate the globally optimal trajectory among several locally optimal controller trajectories. We analyze the continuity and asymptotic properties of the locally and globally optimal controller trajectories as the damping parameter varies. In particular, we prove that under certain conditions, there is no spurious locally optimal controller for an ODC problem with favorable control structure and a large damping parameter. As a result, the proposed method is able to locate the globally optimal trajectory with a suitable discretization in the space of the damping parameter. To demonstrate the effectiveness of this technique, it is shown that even for instances with an exponential number of connected components, damping could merge the trajectories of all local solutions to the trajectory of the global solutions. We further illustrate the convoluted behavior of the locally optimal trajectories with numerical examples on random systems.

Keywords: Optimal control, Robotics, Autonomous systems
Abstract: In this paper, optimal control theory is used to generate energy-optimal tree-to-tree routes by taking into account the map of the forest environment and the quadrotor's dynamics. After computing the energy consumption between all pairs of trees, a Travelling Salesman Problem is formulated and solved using Integer Linear Programming along with Sub-Tour Elimination Constraints to determine the optimal global route. The first scan is not only time-optimal but is also complete. Thus, an infection map is established after the first scan. For subsequent scans, an online solution is proposed to update the path-planner and only include a subset of the trees by considering the probability of infection of a tree based on neighboring trees.

Keywords: Optimal control, Decentralized control, Robust control
Abstract: We design resilient sparse state-feedback controllers for a linear time-invariant (LTI) control system while attaining a pre-specified guarantee on H-infinity performance measure. We leverage a technique from non-fragile control theory to identify a region of resilient state-feedback controllers. Afterward, we explore the region to identify a sparse controller. To this end, we use two different techniques: the greedy method of sparsification, as well as the re-weighted l-1 norm minimization. Our approach highlights a tradeoff between the sparsity of the feedback gain, performance measure, and fragility of the design. To best of our knowledge, this work is the first framework providing H-infinity performance guarantees for sparse feedback gain design.

Keywords: Optimal control, Mechatronics
Abstract: This paper investigates the design of a cascading linear quadratic tracking controller to improve the tracking performance of a dual-stage nanopositioning system. This approach separates the controller design between actuators, allowing for more flexibility in assigning actuator inputs. Specifically, the short-range actuator is given a reference dependent on the long range actuator’s residual tracking error compared to the full desired trajectory, effectively enabling more aggressive effort from the secondary actuator. This new control paradigm is validated in simulation on one axis of an experimental multi-axis dual-stage positioner, where individual actuator measurements are assumed to be available. Simulations are conducted with slow and fast triangular reference trajectories using the cascading design and a standard multi-input multi-output tracking controller. The plant is varied to include high-order dynamics and nonlinearities (specifically hysteresis) to further demonstrate performance. Results show a reduction of maximum and root-mean-square error by approximately 40% for plants with umodeled dynamics and hysteresis, while the root-mean-square error is reduced by nearly 90% for a model matching the plant’s dynamics.

Keywords: Optimal control, Numerical algorithms, Output regulation
Abstract: In this paper, we investigate an iterative method for computing optimal controls for general affine nonlinear quadratic tracking problems. The control law is computed iteratively by solving a sequence of linear quadratic tracking problems and, in particular, it consists of solving a set of coupled differential equations derived from the Hamilton-Jacobi-Bellman equation. The convergence of the iterative scheme is shown by constructing a contraction mapping and using the fixed-point theorem. The versatility and effectiveness of the proposed method is demonstrated in numerical simulations of three structurally different nonlinear systems.

Keywords: Optimal control, Optimization algorithms
Abstract: We present an efficient optimization framework that solves trajectory optimization problems by decoupling state variables from timing variables, thereby decomposing a challenging nonlinear programming (NLP) problem into two easier subproblems. With timing fixed, the state variables can be optimized efficiently using convex optimization, and the timing variables can be optimized in a separate NLP, which forms a bilevel optimization problem. The challenge is to obtain the gradient of the objective function which itself needs an optimization to compute. Whereas finite differences must solve many optimization problems to compute the gradient, our method is based on sensitivity analysis of parametric programming: the dual solution (Lagrange multipliers) of the lower-level optimization is used to compute analytical gradients. Since the dual solution is a by-product of the optimization, the exact gradients can be obtained ``for free''. The framework is demonstrated on generating trajectories in safe corridors for an unmanned aerial vehicle. Experiments demonstrate that bilevel optimization converges significantly more reliably than a standard NLP solver, and analytical gradients outperform finite differences in terms of computation speed and accuracy. With a 25,ms cutoff time, our approach achieves over 8 times better suboptimality than the current state-of-the-art.

Keywords: Optimization, Optimization algorithms, Lyapunov methods
Abstract: In this paper, we propose a new family of continuous-time optimization algorithms for time-varying, locally strongly convex cost functions, based on discontinuous second-order gradient optimization flows with provable finite-time convergence to local optima. To analyze our flows, we first extend a well-know Lyapunov inequality condition for finite-time stability, to the case of arbitrary time-varying differential inclusions, particularly of the Filippov type. We then prove the convergence of our proposed flows in finite time. We illustrate the performance of our proposed flows on a quadratic cost function to track a decaying sinusoid.

Keywords: Optimization, Distributed control, Networked control systems
Abstract: We consider a general distributed optimization problem, aiming to optimize the average of a set of local objectives that are Lipschitz continuous univariate functions, with the existence of same local constraint sets. To solve the problem, we propose a Chebyshev Proxy and Consensus-based Algorithm (CPCA). Compared with existing distributed optimization algorithms, CPCA is able to address the problem with non-convex Lipschitz objectives, and has low computational costs since it is free from gradient or projection calculations. These benefits result from 1) the idea of optimizing a Chebyshev polynomial approximation (i.e. a proxy) for the global objective to get (epsilon)-suboptimal solutions for any given precision (epsilon), and 2) the use of average consensus where the local proxies' coefficient vectors are gossiped to enable every agent to get such a global proxy. We provide comprehensive analysis of the accuracy and complexities of the proposed algorithm. Simulations are conducted to illustrate its effectiveness.

Keywords: Optimal control, Multivehicle systems, Nonholonomic systems
Abstract: This paper considers a two agent scenario containing an observer and a non-maneuvering target. The observer is maneuverable but is slower than the course-holding target. In this scenario, the observer is endowed with a nonzero radius of observation within which he strives at keeping the target for as long as possible. Using the calculus of variations, we pose and solve the optimal control problem, solving for the heading of the observer which maximizes the amount of time the target remains inside the radius of observation. Utilizing the optimal observer heading we compute the exposure time based upon the angle by which the target is initially captured. Presented, along with an example, are the zero-time of exposure heading, maximum time of observation heading, and proof that observation is persistent under optimal control.

Keywords: Optimization, Optimization algorithms, Machine learning
Abstract: The problem of finding the minimizer of a sum of convex functions is central to the field of optimization. Thus, it is of interest to understand how that minimizer is related to the properties of the individual functions in the sum. In this paper, we consider the scenario where one of the individual functions in the sum is not known completely. Instead, only a region containing the minimizer of the unknown function is known, along with some general characteristics (such as strong convexity parameters). Given this limited information about a portion of the overall function, we provide a necessary condition which can be used to construct an upper bound on the region containing the minimizer of the sum of known and unknown functions. We provide this necessary condition in both the general case where the uncertainty region of the minimizer of the unknown function is arbitrary, and in the specific case where the uncertainty region is a ball.

Keywords: Optimization algorithms
Abstract: We present a generic solver for unconstrained control problems (UCPs) whose objectives take the form of an integral functional of the controllers. The solver generalizes and improves upon our previously proposed algorithm for the Witsenhausen's counterexample, which provides the best-known results. In essence, we show that minimizing the objective implies minimizing the marginal cost functions almost everywhere, and we perform the latter task pointwisely by the adaptive minimization technique, which speeds up the computation. We implement single-threaded and parallelized versions of the proposed algorithm. Our implementation runs 30 times faster than our previous algorithm on the Witsenhausen's counterexample, and we demonstrate the applicability of the solver and discuss the possible generalization to constrained problems and multi-dimensional controllers through three more examples.

Keywords: Optimization algorithms, Robust control
Abstract: Iterative first-order methods such as gradient descent and its variants are widely used for solving optimization and machine learning problems. There has been recent interest in analytic or numerically efficient methods for computing worst-case performance bounds for such algorithms, for example over the class of strongly convex loss functions. A popular approach is to assume the algorithm has a fixed size (fixed dimension, or memory) and that its structure is parameterized by one or two hyperparameters, for example a learning rate and a momentum parameter. Then, a Lyapunov function is sought to certify robust stability and subsequent optimization can be performed to find optimal hyperparameter tunings. In the present work, we instead fix the constraints that characterize the loss function and apply techniques from robust control synthesis to directly search over algorithms. This approach yields stronger results than those previously available, since the bounds produced hold over algorithms with an arbitrary, but finite, amount of memory rather than just holding for algorithms with a prescribed structure.

Keywords: Optimization algorithms, Optimization, Machine learning
Abstract: Convex composition optimization is an emerging topic that covers a wide range of applications arising from stochastic optimal control, reinforcement learning and multi-stage stochastic programming. Existing algorithms suffer from unsatisfactory sample complexity and practical issues since they ignore the convexity structure in the algorithmic design. In this paper, we develop a new stochastic compositional variance-reduced gradient algorithm with the sample complexity of O((m+n)log(1/epsilon)+1/epsilon^3) where m+n is the total number of samples. Our algorithm is near-optimal as the dependence on m+n is optimal up to a logarithmic factor. Experimental results on real-world datasets demonstrate the effectiveness and efficiency of the new algorithm.

Keywords: Optimization algorithms, Robust control
Abstract: A classical method for risk-sensitive nonlinear control is the iterative linear exponential quadratic Gaussian algorithm. We present its convergence analysis from a first-order optimization viewpoint. We identify the objective that the algorithm actually minimizes and we show how the addition of a proximal term guarantees convergence to a stationary point.

Keywords: Optimization algorithms, Numerical algorithms, Optimization
Abstract: We consider a routing problem that arises in Persistent Intelligence Surveillance Reconnaissance missions, and pose it as a multiple depot traveling salesman problem, with an objective to minimize the maximum tour length, and an additional set of constraints on the revisit period of every node. We aim to satisfy the revisit period constraints by enforcing equivalent tour length constraints. We present a Mixed Integer Liner Programming (MILP) formulation for this proposed problem which is solved in a branch and cut framework. We also present a market-based solution, an iterative procedure, where each salesman is assigned prices in each iteration and the node assignments are updated accordingly. The quality of the solutions and computation time of the market-based solution are compared to a MILP solution using several numerical simulations. The MILP formulation does not scale well for larger instances with more than hundred nodes, however, the market based approach is able to produce quality solutions to these instances with only a few seconds of computation time.

Keywords: Optimization algorithms, Cooperative control, Emerging control applications
Abstract: This paper considers the problem of distributed beamforming using multi-agent coordination. Each agent is equipped with an antenna and the agents represent the individual elements in an antenna array. The agents are tasked to coordinate their relative location, phase offsets, and amplitude to construct a desired beam-pattern. As a prospective solution, we propose a two time-scale optimization algorithm that consists of a fast time-scale algorithm to solve for the amplitude and phase while a slow time-scale algorithm re-position the agents. In addition, the framework is further extended to the scenarios where the exact parameters of the model are unknown and a model-free optimization based on real-time feedback is proposed. The numerical results given here indicate the proposed approaches reconstruct the desired beam pattern.

Keywords: Optimization algorithms, Autonomous robots, Robotics
Abstract: Path planning for mobile robotics is a topic that has been studied for many decades, with many different formulations and goals. Considering obstacle avoidance with the very simple goal of minimizing the path distance from a start to end location, even this focused problem has attracted many solutions. The aspect of the problem studied in detail here is motivated by the question: what extent of the map needs to be considered by an algorithm to guarantee that the shortest path solution is within the considered extent? The algorithm presented in this paper examines this question in detail, revealing that the area of consideration can be calculated in stages of progress through a known map. Using this bound, the paper then proposes a method for guaranteeing the shortest path, while attempting to minimize the calculation time and memory requirements caused by consideration of map areas that would not admit the optimal path.

Keywords: Optimization algorithms, Optimization
Abstract: In this paper, we used a multi-objective optimization to increase the efficiency and decrease the torque ripple frequency of a BLDC motor by changing the dimensions of the stator sluts. Torque ripple usually happens when there is a variation in torque production and it causes unwanted vibrations and speed fluctuations. Simulations are used to demonstrate the effectiveness of each parameter in both efficiency and torque ripple as our objectives so we can have several design points and based on these points it is possible to reach to a mathematical formula in which describes the optimization problem. In order to get an explicit formula from true unknown responses, different meta-modeling methods were used and to solve the proposed optimization problem, Genetic Algorithm(GA) method was utilized in order to get Pareto frontiers and compare the results. Also in this paper, the accuracy of each model was checked by measuring its statistical parameters.

Keywords: Optimization algorithms, Delay systems, Adaptive control
Abstract: This paper develops a multivariable extremum-seeking control technique for static maps with constant and known input-output delays. The technique is based on the estimation of the gradient of the map as an unknown time-varying parameter using an infinite-dimensional observer. Unlike conventional extremum-seeking controllers, the proposed delay-compensated controller does not require the perturbation signal to be periodic, nor does it require any averaging analysis to guarantee closed-loop stability. Using Lyapunov stability analysis, we show that the observer-coupled closed-loop system exponentially converges to a small neighborhood of the origin. The effectiveness of the controller is demonstrated using a numerical simulation.

Keywords: Energy systems, Smart grid, Stochastic optimal control
Abstract: Undoubtedly, evolving wholesale electricity markets continue to provide new revenue opportunities for diverse generation, energy storage, and flexible demand technologies. In this paper, we quantitatively explore how price uncertainty impacts optimal market participation strategies and resulting revenues. Specifically, we benchmark 2-stage stochastic programming formulations for self-schedule and bidding market participation modes in a receding horizon model predictive control framework. To generate probabilistic price forecasts, we propose an autoregressive Gaussian process regression model and compare three sampling strategies. As an illustrative example, we study a price-taker generation company with six unique generation units using historical price data from CAISO (California market). We show that self-schedule is sensitive to the error in the forecast mean, whereas bidding requires price forecasts that cover extreme events (e.g., tails of the distribution). We benchmark realized market revenue against optimal bidding with perfect information and find static bid curve, time-varying bid curve, and self-schedule modes recovery 95.29%, 94.85%, and 84.87% of perfect information revenue, respectively.

Keywords: Adaptive control, Energy systems, Smart grid
Abstract: This work studies the grid-connected microgrid of a medical facility that is equipped with photovoltaic (PV) energy generation and a local battery energy storage system (BESS). Following a systematic approach in which power flow and state of charge (SoC) dynamics of the BESS is modeled and estimated from experiments, an optimal SoC reference profile is computed given electricity price, predicted solar generation, predicted load, and power and energy constraints. With SoC measurements subjected to errors, a Kalman filter is designed and used in a feedback controller that uses information on the Kalman filtered SoC, PV power and measured power flow at the PCC to compute the desired power demand signal for the DER to ensure the SoC of the BESS remains within operational conditions and tracks the optimal SoC reference profile. Experimental results on SoC scheduling and power control are provided to demonstrate the implementation of the approach.

Keywords: Nonlinear output feedback, Energy systems, Smart grid
Abstract: In this paper, a filter-based control scheme is developed for a single-phase grid-connected inverter system with local load. Improving the quality of the local load voltage in the grid-connected mode and injecting clean current to the grid at the same time is the main objective of the proposed scheme. Moreover, unity power factor at the grid side for different types of the local load is ensured by the proposed controller. Furthermore, this control approach ensures the seamless transfer between grid-connected and stand-alone operation modes without adjusting the controller structure and without any resynchronization scheme. The proposed scheme is validated by a Lyapunov stability analysis as well as via instantaneous dynamic circuit simulation in PLECS software.

Keywords: Smart grid, Stochastic systems, Power systems
Abstract: Over the last few decades, electricity markets around the world have adopted multi-settlement structures, allowing for balancing of supply and demand as more accurate forecast information becomes available. Given increasing uncertainty due to adoption of renewables, more recent market design work has focused on optimization of expectation of some quantity, e.g. social welfare. However, social planners and policy makers are often risk averse, so that such risk neutral formulations do not adequately reflect prevailing attitudes towards risk, nor explain the decisions that follow. Hence we incorporate the commonly used risk measure conditional value at risk (CVaR) into the central planning objective, and study how a two-stage market operates when the individual generators are risk neutral. Our primary result is to show existence (by construction) of a sequential competitive equilibrium (SCEq) in this risk-aware two-stage market. Given equilibrium prices, we design a market mechanism which achieves social cost minimization assuming that agents are non strategic.

Keywords: Smart grid, Energy systems, Lyapunov methods
Abstract: Three-phase inverters for photovoltaic grid-connected applications typically require some form of grid voltage phase detection in order to properly synchronize to the grid and control real and reactive power generation. A phase locked loop is used to determine this type of phase detection. However, in the present work, a method is proposed whereby the phase angle of the grid can be accurately identified solely via the grid current feedback. This phase observer is incorporated into a current controller which can manage the real and reactive power. Moreover, the maximum power point of the photovoltaic arrays is achieved without using a DC-DC converter. The design of this combined observer/controller system is motivated and validated via a Lyapunov stability analysis. Simulation results under various operating conditions are provided for further validation.

Keywords: Smart grid, Optimization algorithms, Computational methods
Abstract: In recent years, it has become crucial to improve the resilience of electricity distribution networks (DNs) against storm-induced failures. Microgrids enabled by Distributed Energy Resources (DERs) can significantly help speed up re-energization of loads, particularly in the complete absence of bulk power supply. We describe an integrated approach which considers a pre-storm DER allocation problem under the uncertainty of failure scenarios as well as a post-storm dispatch problem in microgrids during the multi-period repair of the failed components. This problem is computationally challenging because the number of scenarios (resp. binary variables) increases exponentially (resp. quadratically) in the network size. Our overall solution approach for solving the resulting two-stage mixed-integer linear program (MILP) involves implementing the sample average approximation (SAA) method and Benders Decomposition. Additionally, we implement a greedy approach to reduce the computational time requirements of the post-storm repair scheduling and dispatch problem. The optimality of the resulting solution is evaluated on a modified IEEE 36-node network.

Keywords: Stochastic optimal control, Energy systems, Control applications
Abstract: We study the deadline scheduling problem for multiple deferrable jobs that arrive in a random manner and are to be processed before individual deadlines. The processing of the jobs is subject to a time-varying limit on the total processing rate at each stage. We formulate the scheduling problem as a restless multi-armed bandit (RMAB) problem. Relaxing the scheduling problem into multiple independent single-arm scheduling problems, we define the Lagrange index as the greatest tax under which it is optimal to activate the arm. For the formulated RMAB, we establish the indexability and the asymptotic optimality of the proposed randomized Lagrange index policy for large systems. Numerical results show that the proposed Lagrange index policy achieves 10%-83% higher average reward than the classical Whittle index policy (that does not take into account the processing rate limits).

Keywords: Energy systems, Adaptive control, Electrical machine control
Abstract: Blue economy industries, such as aquaculture or deep sea mining, are moving further offshore to take advantage of the vast scale of the ocean, but moving further offshore requires access to consistent, reliable power untethered to land-based power grids. With a high potential for low cost power generation in locations otherwise isolated from the grid, marine hydrokinetic turbines could serve to help meet this growing power demand. This paper presents a novel adaptive super-twisting sliding mode control strategy for permanent magnet synchronous generators (PMSG) driven by ocean current turbines (OCT). To ensure robustness and mitigate chattering during maximum power point tracking (MPPT), an adaptive gain adjustment technique is proposed for super-twisting sliding mode control. This technique does not require knowledge of the upper bounds of uncertainties, such as external marine environment variability or unmodeled dynamics. More specifically, the adaptive gain rate can vary with a sliding variable when system states are approaching or on the sliding mode, which constitutes the novelty of this paper. The adaptive dynamic gain enables the rapid establishment of the real 2-sliding mode, and this is accomplished without overestimating or underestimating the disturbance boundary. The Lyapunov function technique is used to analyze the finite time convergence of the closed-loop system. A numerical model of a 720-kW PMSG-based OCT is utilized for validating the effectiveness of the proposed control strategy, with simulated operating environmental conditions based on ocean current data collected from the Gulf Stream off Southeast Florida.

Keywords: Automotive systems, Estimation, Identification for control
Abstract: This paper focuses on the estimation of the maximum tire-road friction in the tire’s longitudinal direction. The proposed estimation scheme relies on readily available on-board sensor measurements, hence it does not require additional hardware. The estimation problem is divided over different subsystems. First, at the chassis motion level, the estimation of the vehicle’s longitudinal speed at its center of gravity takes place. Second, based on the wheels rotational dynamics, the longitudinal tire forces and slip ratio are estimated via a state observer design. Information on the longitudinal maximum friction coefficient is then extracted using a model-based identification technique, which relies on Pacejka’s Magic Formula tire model and the similarity method. Exponential convergence of the estimation error is guaranteed based on Lyapunov’s stability theorem. The implementation and validation of the proposed estimation scheme are carried out in a MATLAB/Simulink framework via co-simulation with CarSim. Accelerate-then-brake scenarios are investigated at constant and variable friction levels, ranging from low to high. The obtained results demonstrate the estimator’s ability to detect the maximum friction value, even at low tire slip values.

Keywords: Energy systems, Numerical algorithms, Predictive control for nonlinear systems
Abstract: This paper presents a millisecond-level look-ahead control algorithm for energy storage. The algorithm connects the optimal control with the Lagrangian multiplier associated with the state-of-charge constraint. It is compared to solving look-ahead control using a state-of-the-art convex optimization solver. The paper include discussions on sufficient conditions for including the non-convex simultaneous charging and discharging constraint, and provide upper and lower bounds for the primal and dual results under such conditions. Simulation results show that both methods obtain the same control result, while the proposed algorithm runs up to 100,000 times faster and solves most problems within one millisecond. The theoretical results from developing this algorithm also provide key insights into designing optimal energy storage control schemes at the centralized system level as well as under distributed settings.

Keywords: Energy systems, Machine learning, Identification
Abstract: A safe and stable lithium-ion battery is created by pairing a lithium-iron-phosphate (LFP) cathode with a lithium-titanate (LTO) anode. The open-circuit voltage of a LFP--LTO battery is flat over significant ranges of state of charge. The weak voltage state-of-charge relationship complicates state-of-charge estimation algorithms that require open-circuit voltage measurements to calibrate or initialize coulomb counting. An alternative to open-circuit voltage state-of-charge calibration is analyzing the small-signal frequency response, which is state-of-charge dependent. The present paper estimates the electrochemical impedance spectra (EIS) using system-identification methods. The battery EIS is extracted from on/off current perturbations via balancing resistors. A LASSO regularization method is applied to the extracted EIS data for the purpose of obtaining frequencies of interest that are state-of-charge dependent. Extracted frequencies of interest are then utilized to create a state-of-charge predictor. The resulting method is validated using experimental results.

Keywords: Energy systems, Machine learning, Randomized algorithms
Abstract: The available resource of offshore wind could technically provide all the needed renewable electricity to the U.S. grid. However, the cost of energy from current floating offshore wind turbines (FOWTs) designs are not economical due to inefficiencies and maintenance costs associated with transitioning terrestrial wind technology offshore. By co-designing lighter less expensive FOWTs with individual pitch control (IPC) of each blade, efficiencies could increase, and costs could decrease to make offshore wind economically viable. However, the nonlinear dynamics and breadth of nonstationary wind and wave loading present challenges to designing effective and robust IPC.

This manuscript presents the development, design, and simulation of machine learning control (MLC) for IPC of FOWTs. MLC has been shown effective for many complex nonlinear fluid-structure interaction problems. This project investigates scaling up these component-level control problems to the system level control of the NREL 5MW OC3 FOWT. A massively parallel genetic program (GP) is developed using MATLAB Simulink and OpenFAST that efficiently evaluates new individuals and selectively tests fitness of each generation in the most challenging design load case. The proposed controller was compared to a baseline PID controller using a cost function that captured the value of annual energy production with maintenance costs correlated to ultimate loads and harmonic fatigue. The proposed controller achieved 67% of the cost of the baseline PID controller, resulting in 4th place in the ARPA-E ATLAS Offshore competition for IPC of the OC3 FOWT for the given design load cases.

Keywords: Energy systems, Fault detection, Identification
Abstract: Li-ion batteries (LIB) are used in many applications because of their high-power/energy density, long life cycle, and low self-discharge rate. Li-ion batteries are susceptible to mechanical damages which may lead to an electrical short, thermal runaway, and possibly explosions or fires. However, in some situations such as in a mild car accident, battery packs can get moderate deformations without an electrical short or immediate thermal runaway. Currently, there is no reliable battery characterization method to determine the safety of the batteries for future use after sustaining mechanical damage. In this study, we investigated the connection between the mechanical indentation of Li-ion cells to their impedance spectra. After the initial characterization of four Li-ion pouch cells, the first part of the study included indenting a cell and monitoring their Electrochemical Impedance Spectroscopy (EIS) and charge/discharge cycling data and comparing them to the ones from the control (intact) samples. The second part of the study included incremental indentation of the cell and collecting EIS measurements at specific times after each increment. The control group went through the same EIS measurements and the time series data were compared to the indented cell. Our results showed that while some properties of batteries such as ohmic resistance remain relatively constant when the battery is subjected to incremental mechanical damage, changes in low-frequency impedance were observed with the mechanical loading, suggesting a criterion to measure the safety of pouch cells.

Keywords: Autonomous systems, Transportation networks, Smart grid
Abstract: This paper presents an algorithmic framework to optimize the operation of an Autonomous Mobility-on-Demand system whereby a centrally controlled fleet of electric self-driving vehicles provides on-demand mobility. In particular, we first present a mixed-integer linear program that captures the joint vehicle coordination and charge scheduling problem, accounting for the battery level of the single vehicles and the energy availability in the power grid. Second, we devise a heuristic algorithm to compute near-optimal solutions in polynomial time. Finally, we apply our algorithm to realistic case studies for Newport Beach, CA. Our results validate the near optimality of our method with respect to the global optimum, whilst suggesting that through vehicle-to-grid operation we can enable a 100% penetration of renewable energy sources and still provide a high-quality mobility service.

Keywords: Automotive control, Automotive systems
Abstract: Wheel slip regulation algorithms constitute a significant part of active vehicle safety systems in heavy commercial road vehicles. Any wheel slip regulation algorithm designed for set-point tracking requires a reference operating wheel slip irrespective of its control strategy. This reference slip is unique to each tire-road interface and changes with tire normal load. The present work proposes the optimal reference slip as one that minimizes braking distance, and introduces a recursive algorithm for its estimation. It further proposes a framework for reference slip estimation and wheel slip control. The credibility of the framework was established by testing in a Hardware-in-Loop setup consisting of a pneumatic braking system and IPG TruckMaker® a high fidelity vehicle dynamics simulation environment. The algorithm estimated the reference slip corresponding to different tire-road interfaces and resulted in stable braking maneuvers with 7-17 % reduction in braking distance.

Keywords: Automotive control, Control software, Sensor fusion
Abstract: This paper covers some of the design details of adapting an 850 horsepower NASCAR cup car that it can be driven by a person with quadriplegia.

Keywords: Automotive control, Optimal control, Optimization algorithms
Abstract: Under real-world driving conditions, connected and automated vehicles (CAVs) must plan and follow an energy-optimal and collision-free speed trajectory with a high updating rate, based on available information limited by its communication range. This paper presents a speed planner using analytical closed-form optimal solutions. Using the simplest vehicle model, we derive closed-form solutions as functions of boundary conditions (BCs) and summarize them without and with pure state variable inequality constraints imposed by speed limits and the preceding vehicle. Then we introduce multiple driving modes (e.g., eco-approach to a traffic signal) for responding to dynamically changing situations and show how to set BCs for each mode while retaining the nature of globally optimal solutions. Finally, we perform a large-scale simulation study to identify and quantify the energy impacts of CAVs for real-world driving routes. A simple but effective planner based on closed-form solutions shows a significant energy saving potential compared with human-driven vehicles and adaptive cruise controlled vehicles with connectivity.

Keywords: Automotive control, Automotive systems
Abstract: Gasoline engines remain the most dominant power source for light-duty to medium-duty ground vehicles. Improving gasoline engine efficiency is critical for reducing greenhouse gas emissions and fuel consumption. Cylinder deactivation has proved to be an effective measure in improving fuel efficiency. Nevertheless, the market share of cylinder deactivation remains limited, mostly for the engines that lack of valve deactivation hardware. This paper presents an alternative control framework which is capable of enabling cylinder deactivation without using the current valve deactivation mechanisms. The new cylinder deactivation control framework mainly consists of closed-loop lambda controllers, a dual-stage adaptive engine speed controller and a two-stage emissions control. Experimental results demonstrated that, in idling condition, the proposed control framework was capable of reducing fuel consumption by up to 25.33%, while maintaining smooth engine speed profile with minimal engine speed variations. The feasibility of emissions control using two-stage TWC control is also discussed and validated using experimental data.

Keywords: Automotive control, Automotive systems, Variable-structure/sliding-mode control
Abstract: With increasingly strict regulations on Diesel engine NOx emission, urea-based selective catalytic reduction (SCR) systems with high NOx conversion efficiency and low tailpipe ammonia (NH3) slip, are much needed. One of the major challenges in SCR operation is the intrinsic tradeoff between tailpipe NOx emissions and NH3 slip. A SCR system with high NH3 loading in the front and low NH3 loading in the rear, has demonstrated the potential to achieve low tailpipe NOx and NH3 emissions. Due to various uncertainties and highly dynamic exhaust conditions, the NH3 storage distribution along the axial direction can be much different from the desired NH3 storage distribution, which may result in high tailpipe NOx or NH3 emissions. The main issue with the existing SCR system is that it may take a significant amount of time for the NH3 storage distribution to recover from the undesired one to the desired one. To address this issue, this paper proposed a novel two-cell SCR system with a bypass valve over the upstream cell, and sliding mode control (SMC)-based control algorithms for each cell in the system to quickly recover the NH3 storage distribution from an undesired one to the target one. Simulation results over US06 cycle demonstrated that, the presented SCR system is able to reduce the time cost of profile transition of NH3 storage level, and achieve 95% or higher NOx conversion efficiency and less than 10 ppm NH3 slip after the desired NH3 storage profile is accomplished. Such a novel SCR architecture and advanced control algorithms can collectively improve SCR performance in highly dynamic operating conditions.

Keywords: Automotive systems, Modeling, Optimal control
Abstract: In this paper, we present a modelling approach to vehicle energy management based on Port-Hamiltonian systems representations. We consider a network of interconnected port-Hamiltonian systems that describes the powertrain components and auxiliaries in the vehicle. This description is suitable to obtain a systematic approach to formulate a decomposable optimal control problem for Complete Vehicle Energy Management. A cost function that describe total energy consumption of the vehicle is proposed in terms of internal energy and losses of each system connected system in the network, which has provided an insightful physical interpretation. Taking advantage of the modularity of the formulation proposed, we present a distributed optimization algorithm to find solutions to the energy management problem. To illustrate this modelling methodology, a case study to optimize the energy consumption of a battery electric vehicle is proposed.

Keywords: Automotive systems, Computational methods, Stochastic systems
Abstract: Online simulations conducted in vehicles can enable predictive control of automotive systems. This capability can be especially valuable for complex propulsion systems to manage performance, safety, and efficiency under changing drive conditions. Reliable online simulations require accurate models. However, modeling errors are unavoidable, and the inputs from the driver and environment are subject to uncertainty and generally unknown a priori, rendering the system stochastic. Furthermore, limited computing resources in a vehicle can prohibit solving stochastic systems, posing a major challenge. This paper seeks to alleviate these computational bottlenecks by utilizing generalized Polynomial Chaos to efficiently propagate and quantify uncertainty without loss of accuracy for online propulsion system simulations. To demonstrate the effectiveness of this method, uncertainty quantification is performed for simulations of vehicle launch where both model and input uncertainties are considered. A standard Monte Carlo method is used as a baseline for comparison. It is shown that, for the same accuracy, the proposed method is more than two orders of magnitude faster than a Monte Carlo method. A variance-based sensitivity analysis is also used to quantify the statistical contribution from each uncertainty source to the output. The outcome suggests that the proposed method is well-suited to automotive applications where fast and accurate on-board simulation capabilities are required.

Keywords: Automotive systems, Automotive control, Estimation
Abstract: Driver models are invaluable for planning in autonomous vehicles as well as validating their safety in simulation. Highly parameterized black-box driver models are very expressive, and can capture nuanced behavior. However, they usually lack interpretability and sometimes exhibit unrealistic-even dangerous-behavior. Rule-based models are interpretable, and can be designed to guarantee safe behavior, but are less expressive due to their low number of parameters. In this article, we show that online parameter estimation applied to the Intelligent Driver Model captures nuanced individual driving behavior while providing collision free trajectories. We solve the online parameter estimation problem using particle filtering, and benchmark performance against rule-based and black-box driver models on two real world driving data sets. We evaluate the closeness of our driver model to ground truth data demonstration and also assess the safety of the resulting emergent driving behavior.

Keywords: Optimization, Human-in-the-loop control, Networked control systems
Abstract: We analyze a human and multi-robot collaboration system and proposed a method to optimally schedule the human attention when a human operator receives collaboration requests from multiple robots at the same time. We formulate the human attention scheduling problem as an integer optimization problem which aims to maximize the overall performance among all the robots, under the constraint that a human has limited attention capacity. We first present the optimal schedule for the human to determine when to collaborate with a robot if there is no contention occurring among robots' collaboration requests. For the moments when contentions occur, we present a contention-resolving Model Predictive Control (MPC) method to dynamically schedule the human attention and determine which robot the human should collaborate with first. The optimal schedule can be then determined using a sampling based approach. The effectiveness of the proposed method is validated through simulation that shows improvements on robot performance.

Keywords: Optimal control
Abstract: The Pure Pursuit strategy is ubiquitous both in the control literature but also in real-world implementation. In this paper, we pose and solve a variant of Isaacs' Two Cutters and Fugitive Ship problem wherein the Pursuers' strategy is fixed to Pure Pursuit, thus making it an optimal control problem. The Pursuers are faster than the Evader and are endowed with a finite capture radius. All agents move with constant velocity and can change heading instantaneously. Although capture is inevitable, the Evader wishes to delay capture as long as possible. The optimal trajectories cover the entire state space. Regions corresponding to either solo capture or isochronous (dual) capture are computed and both types of maximal time-to-capture optimal trajectories are characterized.

Keywords: Optimal control, Uncertain systems, Process Control
Abstract: We consider the problem of optimal feedback control of unknown dynamic systems. Optimal feedback control is important for systems with time-varying initial conditions, which cannot rely on feedforward controllers. For example, components in particle accelerators are turned on and off hundreds of times per second, with pulse widths of ~1 ms and repetition rates >100 Hz, whose dynamics and initial conditions vary slowly over time due to external disturbances, such as temperature fluctuations. Our approach allows us to quickly learn an optimal feedback control, which can then be applied for all initial conditions even as the system begins to drift and change. For linear systems we reproduce the cost minimizing linear quadratic regulator (LQR) optimal controller that could have been designed had the system been known.

Keywords: Optimal control, Decentralized control, Optimization algorithms
Abstract: We study the design of an optimal static decentralized controller with a quadratic cost and propose a variant of the homotopy continuation method using a damping technique. This method generates a series of optimal distributed control (ODC) problems via a continuous variation of the system parameters. Diverging from the classical theme of tracking a specific trajectory of locally optimal controllers for these ODC problems, we focus on the possibility of leveraging local-search algorithms to locate the globally optimal trajectory among several locally optimal controller trajectories. We analyze the continuity and asymptotic properties of the locally and globally optimal controller trajectories as the damping parameter varies. In particular, we prove that under certain conditions, there is no spurious locally optimal controller for an ODC problem with favorable control structure and a large damping parameter. As a result, the proposed method is able to locate the globally optimal trajectory with a suitable discretization in the space of the damping parameter. To demonstrate the effectiveness of this technique, it is shown that even for instances with an exponential number of connected components, damping could merge the trajectories of all local solutions to the trajectory of the global solutions. We further illustrate the convoluted behavior of the locally optimal trajectories with numerical examples on random systems.

Keywords: Optimal control, Robotics, Autonomous systems
Abstract: In this paper, optimal control theory is used to generate energy-optimal tree-to-tree routes by taking into account the map of the forest environment and the quadrotor's dynamics. After computing the energy consumption between all pairs of trees, a Travelling Salesman Problem is formulated and solved using Integer Linear Programming along with Sub-Tour Elimination Constraints to determine the optimal global route. The first scan is not only time-optimal but is also complete. Thus, an infection map is established after the first scan. For subsequent scans, an online solution is proposed to update the path-planner and only include a subset of the trees by considering the probability of infection of a tree based on neighboring trees.

Keywords: Optimal control, Decentralized control, Robust control
Abstract: We design resilient sparse state-feedback controllers for a linear time-invariant (LTI) control system while attaining a pre-specified guarantee on H-infinity performance measure. We leverage a technique from non-fragile control theory to identify a region of resilient state-feedback controllers. Afterward, we explore the region to identify a sparse controller. To this end, we use two different techniques: the greedy method of sparsification, as well as the re-weighted l-1 norm minimization. Our approach highlights a tradeoff between the sparsity of the feedback gain, performance measure, and fragility of the design. To best of our knowledge, this work is the first framework providing H-infinity performance guarantees for sparse feedback gain design.

Keywords: Optimal control, Mechatronics
Abstract: This paper investigates the design of a cascading linear quadratic tracking controller to improve the tracking performance of a dual-stage nanopositioning system. This approach separates the controller design between actuators, allowing for more flexibility in assigning actuator inputs. Specifically, the short-range actuator is given a reference dependent on the long range actuator’s residual tracking error compared to the full desired trajectory, effectively enabling more aggressive effort from the secondary actuator. This new control paradigm is validated in simulation on one axis of an experimental multi-axis dual-stage positioner, where individual actuator measurements are assumed to be available. Simulations are conducted with slow and fast triangular reference trajectories using the cascading design and a standard multi-input multi-output tracking controller. The plant is varied to include high-order dynamics and nonlinearities (specifically hysteresis) to further demonstrate performance. Results show a reduction of maximum and root-mean-square error by approximately 40% for plants with umodeled dynamics and hysteresis, while the root-mean-square error is reduced by nearly 90% for a model matching the plant’s dynamics.

Keywords: Optimal control, Numerical algorithms, Output regulation
Abstract: In this paper, we investigate an iterative method for computing optimal controls for general affine nonlinear quadratic tracking problems. The control law is computed iteratively by solving a sequence of linear quadratic tracking problems and, in particular, it consists of solving a set of coupled differential equations derived from the Hamilton-Jacobi-Bellman equation. The convergence of the iterative scheme is shown by constructing a contraction mapping and using the fixed-point theorem. The versatility and effectiveness of the proposed method is demonstrated in numerical simulations of three structurally different nonlinear systems.

Keywords: Optimal control, Optimization algorithms
Abstract: We present an efficient optimization framework that solves trajectory optimization problems by decoupling state variables from timing variables, thereby decomposing a challenging nonlinear programming (NLP) problem into two easier subproblems. With timing fixed, the state variables can be optimized efficiently using convex optimization, and the timing variables can be optimized in a separate NLP, which forms a bilevel optimization problem. The challenge is to obtain the gradient of the objective function which itself needs an optimization to compute. Whereas finite differences must solve many optimization problems to compute the gradient, our method is based on sensitivity analysis of parametric programming: the dual solution (Lagrange multipliers) of the lower-level optimization is used to compute analytical gradients. Since the dual solution is a by-product of the optimization, the exact gradients can be obtained ``for free''. The framework is demonstrated on generating trajectories in safe corridors for an unmanned aerial vehicle. Experiments demonstrate that bilevel optimization converges significantly more reliably than a standard NLP solver, and analytical gradients outperform finite differences in terms of computation speed and accuracy. With a 25,ms cutoff time, our approach achieves over 8 times better suboptimality than the current state-of-the-art.

Keywords: Optimization, Optimization algorithms, Lyapunov methods
Abstract: In this paper, we propose a new family of continuous-time optimization algorithms for time-varying, locally strongly convex cost functions, based on discontinuous second-order gradient optimization flows with provable finite-time convergence to local optima. To analyze our flows, we first extend a well-know Lyapunov inequality condition for finite-time stability, to the case of arbitrary time-varying differential inclusions, particularly of the Filippov type. We then prove the convergence of our proposed flows in finite time. We illustrate the performance of our proposed flows on a quadratic cost function to track a decaying sinusoid.

Keywords: Optimization, Distributed control, Networked control systems
Abstract: We consider a general distributed optimization problem, aiming to optimize the average of a set of local objectives that are Lipschitz continuous univariate functions, with the existence of same local constraint sets. To solve the problem, we propose a Chebyshev Proxy and Consensus-based Algorithm (CPCA). Compared with existing distributed optimization algorithms, CPCA is able to address the problem with non-convex Lipschitz objectives, and has low computational costs since it is free from gradient or projection calculations. These benefits result from 1) the idea of optimizing a Chebyshev polynomial approximation (i.e. a proxy) for the global objective to get (epsilon)-suboptimal solutions for any given precision (epsilon), and 2) the use of average consensus where the local proxies' coefficient vectors are gossiped to enable every agent to get such a global proxy. We provide comprehensive analysis of the accuracy and complexities of the proposed algorithm. Simulations are conducted to illustrate its effectiveness.

Keywords: Optimal control, Multivehicle systems, Nonholonomic systems
Abstract: This paper considers a two agent scenario containing an observer and a non-maneuvering target. The observer is maneuverable but is slower than the course-holding target. In this scenario, the observer is endowed with a nonzero radius of observation within which he strives at keeping the target for as long as possible. Using the calculus of variations, we pose and solve the optimal control problem, solving for the heading of the observer which maximizes the amount of time the target remains inside the radius of observation. Utilizing the optimal observer heading we compute the exposure time based upon the angle by which the target is initially captured. Presented, along with an example, are the zero-time of exposure heading, maximum time of observation heading, and proof that observation is persistent under optimal control.

Keywords: Optimization, Optimization algorithms, Machine learning
Abstract: The problem of finding the minimizer of a sum of convex functions is central to the field of optimization. Thus, it is of interest to understand how that minimizer is related to the properties of the individual functions in the sum. In this paper, we consider the scenario where one of the individual functions in the sum is not known completely. Instead, only a region containing the minimizer of the unknown function is known, along with some general characteristics (such as strong convexity parameters). Given this limited information about a portion of the overall function, we provide a necessary condition which can be used to construct an upper bound on the region containing the minimizer of the sum of known and unknown functions. We provide this necessary condition in both the general case where the uncertainty region of the minimizer of the unknown function is arbitrary, and in the specific case where the uncertainty region is a ball.

Keywords: Optimization algorithms
Abstract: We present a generic solver for unconstrained control problems (UCPs) whose objectives take the form of an integral functional of the controllers. The solver generalizes and improves upon our previously proposed algorithm for the Witsenhausen's counterexample, which provides the best-known results. In essence, we show that minimizing the objective implies minimizing the marginal cost functions almost everywhere, and we perform the latter task pointwisely by the adaptive minimization technique, which speeds up the computation. We implement single-threaded and parallelized versions of the proposed algorithm. Our implementation runs 30 times faster than our previous algorithm on the Witsenhausen's counterexample, and we demonstrate the applicability of the solver and discuss the possible generalization to constrained problems and multi-dimensional controllers through three more examples.

Keywords: Optimization algorithms, Robust control
Abstract: Iterative first-order methods such as gradient descent and its variants are widely used for solving optimization and machine learning problems. There has been recent interest in analytic or numerically efficient methods for computing worst-case performance bounds for such algorithms, for example over the class of strongly convex loss functions. A popular approach is to assume the algorithm has a fixed size (fixed dimension, or memory) and that its structure is parameterized by one or two hyperparameters, for example a learning rate and a momentum parameter. Then, a Lyapunov function is sought to certify robust stability and subsequent optimization can be performed to find optimal hyperparameter tunings. In the present work, we instead fix the constraints that characterize the loss function and apply techniques from robust control synthesis to directly search over algorithms. This approach yields stronger results than those previously available, since the bounds produced hold over algorithms with an arbitrary, but finite, amount of memory rather than just holding for algorithms with a prescribed structure.

Keywords: Optimization algorithms, Optimization, Machine learning
Abstract: Convex composition optimization is an emerging topic that covers a wide range of applications arising from stochastic optimal control, reinforcement learning and multi-stage stochastic programming. Existing algorithms suffer from unsatisfactory sample complexity and practical issues since they ignore the convexity structure in the algorithmic design. In this paper, we develop a new stochastic compositional variance-reduced gradient algorithm with the sample complexity of O((m+n)log(1/epsilon)+1/epsilon^3) where m+n is the total number of samples. Our algorithm is near-optimal as the dependence on m+n is optimal up to a logarithmic factor. Experimental results on real-world datasets demonstrate the effectiveness and efficiency of the new algorithm.

Keywords: Optimization algorithms, Robust control
Abstract: A classical method for risk-sensitive nonlinear control is the iterative linear exponential quadratic Gaussian algorithm. We present its convergence analysis from a first-order optimization viewpoint. We identify the objective that the algorithm actually minimizes and we show how the addition of a proximal term guarantees convergence to a stationary point.

Keywords: Optimization algorithms, Numerical algorithms, Optimization
Abstract: We consider a routing problem that arises in Persistent Intelligence Surveillance Reconnaissance missions, and pose it as a multiple depot traveling salesman problem, with an objective to minimize the maximum tour length, and an additional set of constraints on the revisit period of every node. We aim to satisfy the revisit period constraints by enforcing equivalent tour length constraints. We present a Mixed Integer Liner Programming (MILP) formulation for this proposed problem which is solved in a branch and cut framework. We also present a market-based solution, an iterative procedure, where each salesman is assigned prices in each iteration and the node assignments are updated accordingly. The quality of the solutions and computation time of the market-based solution are compared to a MILP solution using several numerical simulations. The MILP formulation does not scale well for larger instances with more than hundred nodes, however, the market based approach is able to produce quality solutions to these instances with only a few seconds of computation time.

Keywords: Optimization algorithms, Cooperative control, Emerging control applications
Abstract: This paper considers the problem of distributed beamforming using multi-agent coordination. Each agent is equipped with an antenna and the agents represent the individual elements in an antenna array. The agents are tasked to coordinate their relative location, phase offsets, and amplitude to construct a desired beam-pattern. As a prospective solution, we propose a two time-scale optimization algorithm that consists of a fast time-scale algorithm to solve for the amplitude and phase while a slow time-scale algorithm re-position the agents. In addition, the framework is further extended to the scenarios where the exact parameters of the model are unknown and a model-free optimization based on real-time feedback is proposed. The numerical results given here indicate the proposed approaches reconstruct the desired beam pattern.

Keywords: Optimization algorithms, Autonomous robots, Robotics
Abstract: Path planning for mobile robotics is a topic that has been studied for many decades, with many different formulations and goals. Considering obstacle avoidance with the very simple goal of minimizing the path distance from a start to end location, even this focused problem has attracted many solutions. The aspect of the problem studied in detail here is motivated by the question: what extent of the map needs to be considered by an algorithm to guarantee that the shortest path solution is within the considered extent? The algorithm presented in this paper examines this question in detail, revealing that the area of consideration can be calculated in stages of progress through a known map. Using this bound, the paper then proposes a method for guaranteeing the shortest path, while attempting to minimize the calculation time and memory requirements caused by consideration of map areas that would not admit the optimal path.

Keywords: Optimization algorithms, Optimization
Abstract: In this paper, we used a multi-objective optimization to increase the efficiency and decrease the torque ripple frequency of a BLDC motor by changing the dimensions of the stator sluts. Torque ripple usually happens when there is a variation in torque production and it causes unwanted vibrations and speed fluctuations. Simulations are used to demonstrate the effectiveness of each parameter in both efficiency and torque ripple as our objectives so we can have several design points and based on these points it is possible to reach to a mathematical formula in which describes the optimization problem. In order to get an explicit formula from true unknown responses, different meta-modeling methods were used and to solve the proposed optimization problem, Genetic Algorithm(GA) method was utilized in order to get Pareto frontiers and compare the results. Also in this paper, the accuracy of each model was checked by measuring its statistical parameters.

Keywords: Optimization algorithms, Delay systems, Adaptive control
Abstract: This paper develops a multivariable extremum-seeking control technique for static maps with constant and known input-output delays. The technique is based on the estimation of the gradient of the map as an unknown time-varying parameter using an infinite-dimensional observer. Unlike conventional extremum-seeking controllers, the proposed delay-compensated controller does not require the perturbation signal to be periodic, nor does it require any averaging analysis to guarantee closed-loop stability. Using Lyapunov stability analysis, we show that the observer-coupled closed-loop system exponentially converges to a small neighborhood of the origin. The effectiveness of the controller is demonstrated using a numerical simulation.

Keywords: Energy systems, Smart grid, Stochastic optimal control
Abstract: Undoubtedly, evolving wholesale electricity markets continue to provide new revenue opportunities for diverse generation, energy storage, and flexible demand technologies. In this paper, we quantitatively explore how price uncertainty impacts optimal market participation strategies and resulting revenues. Specifically, we benchmark 2-stage stochastic programming formulations for self-schedule and bidding market participation modes in a receding horizon model predictive control framework. To generate probabilistic price forecasts, we propose an autoregressive Gaussian process regression model and compare three sampling strategies. As an illustrative example, we study a price-taker generation company with six unique generation units using historical price data from CAISO (California market). We show that self-schedule is sensitive to the error in the forecast mean, whereas bidding requires price forecasts that cover extreme events (e.g., tails of the distribution). We benchmark realized market revenue against optimal bidding with perfect information and find static bid curve, time-varying bid curve, and self-schedule modes recovery 95.29%, 94.85%, and 84.87% of perfect information revenue, respectively.

Keywords: Adaptive control, Energy systems, Smart grid
Abstract: This work studies the grid-connected microgrid of a medical facility that is equipped with photovoltaic (PV) energy generation and a local battery energy storage system (BESS). Following a systematic approach in which power flow and state of charge (SoC) dynamics of the BESS is modeled and estimated from experiments, an optimal SoC reference profile is computed given electricity price, predicted solar generation, predicted load, and power and energy constraints. With SoC measurements subjected to errors, a Kalman filter is designed and used in a feedback controller that uses information on the Kalman filtered SoC, PV power and measured power flow at the PCC to compute the desired power demand signal for the DER to ensure the SoC of the BESS remains within operational conditions and tracks the optimal SoC reference profile. Experimental results on SoC scheduling and power control are provided to demonstrate the implementation of the approach.

Keywords: Nonlinear output feedback, Energy systems, Smart grid
Abstract: In this paper, a filter-based control scheme is developed for a single-phase grid-connected inverter system with local load. Improving the quality of the local load voltage in the grid-connected mode and injecting clean current to the grid at the same time is the main objective of the proposed scheme. Moreover, unity power factor at the grid side for different types of the local load is ensured by the proposed controller. Furthermore, this control approach ensures the seamless transfer between grid-connected and stand-alone operation modes without adjusting the controller structure and without any resynchronization scheme. The proposed scheme is validated by a Lyapunov stability analysis as well as via instantaneous dynamic circuit simulation in PLECS software.

Keywords: Smart grid, Stochastic systems, Power systems
Abstract: Over the last few decades, electricity markets around the world have adopted multi-settlement structures, allowing for balancing of supply and demand as more accurate forecast information becomes available. Given increasing uncertainty due to adoption of renewables, more recent market design work has focused on optimization of expectation of some quantity, e.g. social welfare. However, social planners and policy makers are often risk averse, so that such risk neutral formulations do not adequately reflect prevailing attitudes towards risk, nor explain the decisions that follow. Hence we incorporate the commonly used risk measure conditional value at risk (CVaR) into the central planning objective, and study how a two-stage market operates when the individual generators are risk neutral. Our primary result is to show existence (by construction) of a sequential competitive equilibrium (SCEq) in this risk-aware two-stage market. Given equilibrium prices, we design a market mechanism which achieves social cost minimization assuming that agents are non strategic.

Keywords: Smart grid, Energy systems, Lyapunov methods
Abstract: Three-phase inverters for photovoltaic grid-connected applications typically require some form of grid voltage phase detection in order to properly synchronize to the grid and control real and reactive power generation. A phase locked loop is used to determine this type of phase detection. However, in the present work, a method is proposed whereby the phase angle of the grid can be accurately identified solely via the grid current feedback. This phase observer is incorporated into a current controller which can manage the real and reactive power. Moreover, the maximum power point of the photovoltaic arrays is achieved without using a DC-DC converter. The design of this combined observer/controller system is motivated and validated via a Lyapunov stability analysis. Simulation results under various operating conditions are provided for further validation.

Keywords: Smart grid, Optimization algorithms, Computational methods
Abstract: In recent years, it has become crucial to improve the resilience of electricity distribution networks (DNs) against storm-induced failures. Microgrids enabled by Distributed Energy Resources (DERs) can significantly help speed up re-energization of loads, particularly in the complete absence of bulk power supply. We describe an integrated approach which considers a pre-storm DER allocation problem under the uncertainty of failure scenarios as well as a post-storm dispatch problem in microgrids during the multi-period repair of the failed components. This problem is computationally challenging because the number of scenarios (resp. binary variables) increases exponentially (resp. quadratically) in the network size. Our overall solution approach for solving the resulting two-stage mixed-integer linear program (MILP) involves implementing the sample average approximation (SAA) method and Benders Decomposition. Additionally, we implement a greedy approach to reduce the computational time requirements of the post-storm repair scheduling and dispatch problem. The optimality of the resulting solution is evaluated on a modified IEEE 36-node network.

Keywords: Stochastic optimal control, Energy systems, Control applications
Abstract: We study the deadline scheduling problem for multiple deferrable jobs that arrive in a random manner and are to be processed before individual deadlines. The processing of the jobs is subject to a time-varying limit on the total processing rate at each stage. We formulate the scheduling problem as a restless multi-armed bandit (RMAB) problem. Relaxing the scheduling problem into multiple independent single-arm scheduling problems, we define the Lagrange index as the greatest tax under which it is optimal to activate the arm. For the formulated RMAB, we establish the indexability and the asymptotic optimality of the proposed randomized Lagrange index policy for large systems. Numerical results show that the proposed Lagrange index policy achieves 10%-83% higher average reward than the classical Whittle index policy (that does not take into account the processing rate limits).

Keywords: Energy systems, Adaptive control, Electrical machine control
Abstract: Blue economy industries, such as aquaculture or deep sea mining, are moving further offshore to take advantage of the vast scale of the ocean, but moving further offshore requires access to consistent, reliable power untethered to land-based power grids. With a high potential for low cost power generation in locations otherwise isolated from the grid, marine hydrokinetic turbines could serve to help meet this growing power demand. This paper presents a novel adaptive super-twisting sliding mode control strategy for permanent magnet synchronous generators (PMSG) driven by ocean current turbines (OCT). To ensure robustness and mitigate chattering during maximum power point tracking (MPPT), an adaptive gain adjustment technique is proposed for super-twisting sliding mode control. This technique does not require knowledge of the upper bounds of uncertainties, such as external marine environment variability or unmodeled dynamics. More specifically, the adaptive gain rate can vary with a sliding variable when system states are approaching or on the sliding mode, which constitutes the novelty of this paper. The adaptive dynamic gain enables the rapid establishment of the real 2-sliding mode, and this is accomplished without overestimating or underestimating the disturbance boundary. The Lyapunov function technique is used to analyze the finite time convergence of the closed-loop system. A numerical model of a 720-kW PMSG-based OCT is utilized for validating the effectiveness of the proposed control strategy, with simulated operating environmental conditions based on ocean current data collected from the Gulf Stream off Southeast Florida.

Keywords: Automotive systems, Estimation, Identification for control
Abstract: This paper focuses on the estimation of the maximum tire-road friction in the tire’s longitudinal direction. The proposed estimation scheme relies on readily available on-board sensor measurements, hence it does not require additional hardware. The estimation problem is divided over different subsystems. First, at the chassis motion level, the estimation of the vehicle’s longitudinal speed at its center of gravity takes place. Second, based on the wheels rotational dynamics, the longitudinal tire forces and slip ratio are estimated via a state observer design. Information on the longitudinal maximum friction coefficient is then extracted using a model-based identification technique, which relies on Pacejka’s Magic Formula tire model and the similarity method. Exponential convergence of the estimation error is guaranteed based on Lyapunov’s stability theorem. The implementation and validation of the proposed estimation scheme are carried out in a MATLAB/Simulink framework via co-simulation with CarSim. Accelerate-then-brake scenarios are investigated at constant and variable friction levels, ranging from low to high. The obtained results demonstrate the estimator’s ability to detect the maximum friction value, even at low tire slip values.

Keywords: Energy systems, Numerical algorithms, Predictive control for nonlinear systems
Abstract: This paper presents a millisecond-level look-ahead control algorithm for energy storage. The algorithm connects the optimal control with the Lagrangian multiplier associated with the state-of-charge constraint. It is compared to solving look-ahead control using a state-of-the-art convex optimization solver. The paper include discussions on sufficient conditions for including the non-convex simultaneous charging and discharging constraint, and provide upper and lower bounds for the primal and dual results under such conditions. Simulation results show that both methods obtain the same control result, while the proposed algorithm runs up to 100,000 times faster and solves most problems within one millisecond. The theoretical results from developing this algorithm also provide key insights into designing optimal energy storage control schemes at the centralized system level as well as under distributed settings.

Keywords: Energy systems, Machine learning, Identification
Abstract: A safe and stable lithium-ion battery is created by pairing a lithium-iron-phosphate (LFP) cathode with a lithium-titanate (LTO) anode. The open-circuit voltage of a LFP--LTO battery is flat over significant ranges of state of charge. The weak voltage state-of-charge relationship complicates state-of-charge estimation algorithms that require open-circuit voltage measurements to calibrate or initialize coulomb counting. An alternative to open-circuit voltage state-of-charge calibration is analyzing the small-signal frequency response, which is state-of-charge dependent. The present paper estimates the electrochemical impedance spectra (EIS) using system-identification methods. The battery EIS is extracted from on/off current perturbations via balancing resistors. A LASSO regularization method is applied to the extracted EIS data for the purpose of obtaining frequencies of interest that are state-of-charge dependent. Extracted frequencies of interest are then utilized to create a state-of-charge predictor. The resulting method is validated using experimental results.

Keywords: Energy systems, Machine learning, Randomized algorithms
Abstract: The available resource of offshore wind could technically provide all the needed renewable electricity to the U.S. grid. However, the cost of energy from current floating offshore wind turbines (FOWTs) designs are not economical due to inefficiencies and maintenance costs associated with transitioning terrestrial wind technology offshore. By co-designing lighter less expensive FOWTs with individual pitch control (IPC) of each blade, efficiencies could increase, and costs could decrease to make offshore wind economically viable. However, the nonlinear dynamics and breadth of nonstationary wind and wave loading present challenges to designing effective and robust IPC.

This manuscript presents the development, design, and simulation of machine learning control (MLC) for IPC of FOWTs. MLC has been shown effective for many complex nonlinear fluid-structure interaction problems. This project investigates scaling up these component-level control problems to the system level control of the NREL 5MW OC3 FOWT. A massively parallel genetic program (GP) is developed using MATLAB Simulink and OpenFAST that efficiently evaluates new individuals and selectively tests fitness of each generation in the most challenging design load case. The proposed controller was compared to a baseline PID controller using a cost function that captured the value of annual energy production with maintenance costs correlated to ultimate loads and harmonic fatigue. The proposed controller achieved 67% of the cost of the baseline PID controller, resulting in 4th place in the ARPA-E ATLAS Offshore competition for IPC of the OC3 FOWT for the given design load cases.

Keywords: Energy systems, Fault detection, Identification
Abstract: Li-ion batteries (LIB) are used in many applications because of their high-power/energy density, long life cycle, and low self-discharge rate. Li-ion batteries are susceptible to mechanical damages which may lead to an electrical short, thermal runaway, and possibly explosions or fires. However, in some situations such as in a mild car accident, battery packs can get moderate deformations without an electrical short or immediate thermal runaway. Currently, there is no reliable battery characterization method to determine the safety of the batteries for future use after sustaining mechanical damage. In this study, we investigated the connection between the mechanical indentation of Li-ion cells to their impedance spectra. After the initial characterization of four Li-ion pouch cells, the first part of the study included indenting a cell and monitoring their Electrochemical Impedance Spectroscopy (EIS) and charge/discharge cycling data and comparing them to the ones from the control (intact) samples. The second part of the study included incremental indentation of the cell and collecting EIS measurements at specific times after each increment. The control group went through the same EIS measurements and the time series data were compared to the indented cell. Our results showed that while some properties of batteries such as ohmic resistance remain relatively constant when the battery is subjected to incremental mechanical damage, changes in low-frequency impedance were observed with the mechanical loading, suggesting a criterion to measure the safety of pouch cells.

Keywords: Autonomous systems, Transportation networks, Smart grid
Abstract: This paper presents an algorithmic framework to optimize the operation of an Autonomous Mobility-on-Demand system whereby a centrally controlled fleet of electric self-driving vehicles provides on-demand mobility. In particular, we first present a mixed-integer linear program that captures the joint vehicle coordination and charge scheduling problem, accounting for the battery level of the single vehicles and the energy availability in the power grid. Second, we devise a heuristic algorithm to compute near-optimal solutions in polynomial time. Finally, we apply our algorithm to realistic case studies for Newport Beach, CA. Our results validate the near optimality of our method with respect to the global optimum, whilst suggesting that through vehicle-to-grid operation we can enable a 100% penetration of renewable energy sources and still provide a high-quality mobility service.

Keywords: Automotive control, Automotive systems
Abstract: Wheel slip regulation algorithms constitute a significant part of active vehicle safety systems in heavy commercial road vehicles. Any wheel slip regulation algorithm designed for set-point tracking requires a reference operating wheel slip irrespective of its control strategy. This reference slip is unique to each tire-road interface and changes with tire normal load. The present work proposes the optimal reference slip as one that minimizes braking distance, and introduces a recursive algorithm for its estimation. It further proposes a framework for reference slip estimation and wheel slip control. The credibility of the framework was established by testing in a Hardware-in-Loop setup consisting of a pneumatic braking system and IPG TruckMaker® a high fidelity vehicle dynamics simulation environment. The algorithm estimated the reference slip corresponding to different tire-road interfaces and resulted in stable braking maneuvers with 7-17 % reduction in braking distance.

Keywords: Automotive control, Control software, Sensor fusion
Abstract: This paper covers some of the design details of adapting an 850 horsepower NASCAR cup car that it can be driven by a person with quadriplegia.

Keywords: Automotive control, Optimal control, Optimization algorithms
Abstract: Under real-world driving conditions, connected and automated vehicles (CAVs) must plan and follow an energy-optimal and collision-free speed trajectory with a high updating rate, based on available information limited by its communication range. This paper presents a speed planner using analytical closed-form optimal solutions. Using the simplest vehicle model, we derive closed-form solutions as functions of boundary conditions (BCs) and summarize them without and with pure state variable inequality constraints imposed by speed limits and the preceding vehicle. Then we introduce multiple driving modes (e.g., eco-approach to a traffic signal) for responding to dynamically changing situations and show how to set BCs for each mode while retaining the nature of globally optimal solutions. Finally, we perform a large-scale simulation study to identify and quantify the energy impacts of CAVs for real-world driving routes. A simple but effective planner based on closed-form solutions shows a significant energy saving potential compared with human-driven vehicles and adaptive cruise controlled vehicles with connectivity.

Keywords: Automotive control, Automotive systems
Abstract: Gasoline engines remain the most dominant power source for light-duty to medium-duty ground vehicles. Improving gasoline engine efficiency is critical for reducing greenhouse gas emissions and fuel consumption. Cylinder deactivation has proved to be an effective measure in improving fuel efficiency. Nevertheless, the market share of cylinder deactivation remains limited, mostly for the engines that lack of valve deactivation hardware. This paper presents an alternative control framework which is capable of enabling cylinder deactivation without using the current valve deactivation mechanisms. The new cylinder deactivation control framework mainly consists of closed-loop lambda controllers, a dual-stage adaptive engine speed controller and a two-stage emissions control. Experimental results demonstrated that, in idling condition, the proposed control framework was capable of reducing fuel consumption by up to 25.33%, while maintaining smooth engine speed profile with minimal engine speed variations. The feasibility of emissions control using two-stage TWC control is also discussed and validated using experimental data.

Keywords: Automotive control, Automotive systems, Variable-structure/sliding-mode control
Abstract: With increasingly strict regulations on Diesel engine NOx emission, urea-based selective catalytic reduction (SCR) systems with high NOx conversion efficiency and low tailpipe ammonia (NH3) slip, are much needed. One of the major challenges in SCR operation is the intrinsic tradeoff between tailpipe NOx emissions and NH3 slip. A SCR system with high NH3 loading in the front and low NH3 loading in the rear, has demonstrated the potential to achieve low tailpipe NOx and NH3 emissions. Due to various uncertainties and highly dynamic exhaust conditions, the NH3 storage distribution along the axial direction can be much different from the desired NH3 storage distribution, which may result in high tailpipe NOx or NH3 emissions. The main issue with the existing SCR system is that it may take a significant amount of time for the NH3 storage distribution to recover from the undesired one to the desired one. To address this issue, this paper proposed a novel two-cell SCR system with a bypass valve over the upstream cell, and sliding mode control (SMC)-based control algorithms for each cell in the system to quickly recover the NH3 storage distribution from an undesired one to the target one. Simulation results over US06 cycle demonstrated that, the presented SCR system is able to reduce the time cost of profile transition of NH3 storage level, and achieve 95% or higher NOx conversion efficiency and less than 10 ppm NH3 slip after the desired NH3 storage profile is accomplished. Such a novel SCR architecture and advanced control algorithms can collectively improve SCR performance in highly dynamic operating conditions.

Keywords: Automotive systems, Modeling, Optimal control
Abstract: In this paper, we present a modelling approach to vehicle energy management based on Port-Hamiltonian systems representations. We consider a network of interconnected port-Hamiltonian systems that describes the powertrain components and auxiliaries in the vehicle. This description is suitable to obtain a systematic approach to formulate a decomposable optimal control problem for Complete Vehicle Energy Management. A cost function that describe total energy consumption of the vehicle is proposed in terms of internal energy and losses of each system connected system in the network, which has provided an insightful physical interpretation. Taking advantage of the modularity of the formulation proposed, we present a distributed optimization algorithm to find solutions to the energy management problem. To illustrate this modelling methodology, a case study to optimize the energy consumption of a battery electric vehicle is proposed.

Keywords: Automotive systems, Computational methods, Stochastic systems
Abstract: Online simulations conducted in vehicles can enable predictive control of automotive systems. This capability can be especially valuable for complex propulsion systems to manage performance, safety, and efficiency under changing drive conditions. Reliable online simulations require accurate models. However, modeling errors are unavoidable, and the inputs from the driver and environment are subject to uncertainty and generally unknown a priori, rendering the system stochastic. Furthermore, limited computing resources in a vehicle can prohibit solving stochastic systems, posing a major challenge. This paper seeks to alleviate these computational bottlenecks by utilizing generalized Polynomial Chaos to efficiently propagate and quantify uncertainty without loss of accuracy for online propulsion system simulations. To demonstrate the effectiveness of this method, uncertainty quantification is performed for simulations of vehicle launch where both model and input uncertainties are considered. A standard Monte Carlo method is used as a baseline for comparison. It is shown that, for the same accuracy, the proposed method is more than two orders of magnitude faster than a Monte Carlo method. A variance-based sensitivity analysis is also used to quantify the statistical contribution from each uncertainty source to the output. The outcome suggests that the proposed method is well-suited to automotive applications where fast and accurate on-board simulation capabilities are required.

Keywords: Automotive systems, Automotive control, Estimation
Abstract: Driver models are invaluable for planning in autonomous vehicles as well as validating their safety in simulation. Highly parameterized black-box driver models are very expressive, and can capture nuanced behavior. However, they usually lack interpretability and sometimes exhibit unrealistic-even dangerous-behavior. Rule-based models are interpretable, and can be designed to guarantee safe behavior, but are less expressive due to their low number of parameters. In this article, we show that online parameter estimation applied to the Intelligent Driver Model captures nuanced individual driving behavior while providing collision free trajectories. We solve the online parameter estimation problem using particle filtering, and benchmark performance against rule-based and black-box driver models on two real world driving data sets. We evaluate the closeness of our driver model to ground truth data demonstration and also assess the safety of the resulting emergent driving behavior.

Keywords:
Abstract: The National Science Foundation (NSF) offers a number of funding opportunities for investigators working in the field of controls, both within the disciplinary programs in Engineering and other directorates, and through cross-cutting initiatives that are foundation-wide. This presentation will describe opportunities that are relevant to the robotics, dynamics and controls communities. The presentation will also describe programs targeted toward junior investigators, as well as guidelines for proposal preparation and NSF’s Intellectual Merit and Broader Impacts criteria. Question-and-answer session will follow the presentation.

Speakers: Dr. Kishan Baheti, Dr. Jordan Berg, Dr. Irina Dolinskaya, Dr. Robert G. Landers, and Dr. Eduardo Misawa

Keywords:
Abstract: Mitsubishi Electric Research Laboratories (MERL) is a leading research organization located in Cambridge, Massachusetts, USA that conducts fundamental research for industrially-motivated problems. MERL is a subsidiary of Mitsubishi Electric Corporation, a 41B global manufacturer of a wide range of products including industrial robots, automotive electronics and equipment, HVAC (heating, ventilation, and air conditioning) systems, factory automation equipment, electrical power systems, elevators, satellites, and information visualization systems. MERL is an active and collaborative member of both the academic and industrial communities. MERL researchers collaborate with corporate laboratories and business units in Japan, as well as academic partners from around the world to develop novel solutions to challenging problems. In particular, several researchers at MERL develops new theoretical results in control and systems theory and apply them to a wide variety of products and applications.

In this talk we will present an overview of research activities at MERL, including fundamental controls research and the application of state-of-the-art control techniques to a variety of real-world systems. We will focus on fundamental research subjects including model predictive control and the control of constrained systems, estimation and motion planning for autonomous systems, and learning for control. In addition, we will describe how these fundamental research areas impact applications such as autonomous vehicles, spacecraft guidance and control, GNSS-based positioning, energy-efficient HVAC systems, high-precision manufacturing.

We encourage students, researchers and faculty interested in collaborating with MERL to attend this talk

Keywords:
Abstract: The IEEE CSS Women in Control committee is responsible for, but not limited to, promoting membership, gathering and disseminating appropriate information about women in IEEE CSS and the profession, and facilitating the development of mentoring and programs to promote the retention, recruitment, and growth of women IEEE CSS members. The IEEE WiC invites all ACC women attendees to join us for an online luncheon with interesting speakers on the first day of the conference, Wednesday, July 1st, 2020.

Keywords:
Abstract: ASME DSCD TC Chairs Meeting Xiaobo Tan is inviting you to a scheduled Zoom meeting.

Topic: ASME DSCD TC Chairs Meeting Time: Jul 1, 2020 12:00 PM Mountain Time (US and Canada) Session WeLuT4 Join Zoom Meeting https://msu.zoom.us/j/92010609422

Contact meeting organizer for password.

One tap mobile +13126266799,,92010609422# US (Chicago) +16468769923,,92010609422# US (New York)

Dial by your location +1 312 626 6799 US (Chicago) +1 646 876 9923 US (New York) +1 301 715 8592 US (Germantown) +1 346 248 7799 US (Houston) +1 669 900 6833 US (San Jose) +1 253 215 8782 US (Tacoma) Meeting ID: 920 1060 9422 Find your local number: https://msu.zoom.us/u/acowsl6Y5p

Join by SIP 162.255.36.11 US East 162.255.37.11 US West 92010609422@vip2.zoomcrc.com

Join by H.323 162.255.36.11 (US East) 162.255.37.11 (US West) 221.122.88.195 (China) 115.114.131.7 (India) 213.19.144.110 (EMEA) 202.177.207.158 (Australia) 209.9.211.110 (Hong Kong) 64.211.144.160 (Brazil) 69.174.57.160 (Canada) Meeting ID: 920 1060 9422 Password: 486763

Join by Skype for Business https://msu.zoom.us/skype/92010609422

Keywords:
Abstract: TC Smart Cities Wed: 12:00PM to 1:30 PM Session: WeLuT4 https://udel.zoom.us/j/97541636655 Use 2020 ACC conference password

Keywords:
Abstract: ASME DSCD Vibration TC Meeting

Wed: 12:00 PM to 1:00 PM Session: WeLuT4

https://buffalo.zoom.us/j/98441521787 Use 2020 ACC conference password

Keywords:
Abstract: Organized by Junmin Wang jwang@austin.utexas.edu

Contact Organizer for Meeting Details

Keywords: Optimal control, Multivehicle systems, Intelligent systems
Abstract: In this paper, we propose a reinforcement learning (RL) approach to the coordinated control of multi-agent systems with a special emphasis on intelligent transportation systems. As a result, for a network of autonomous vehicles, a novel distributed predictive cruise control (PCC) algorithm based on RL is proposed to reduce idle time and maintain an adjustable speed depending on the traffic signals. Under the proposed distributed PCC law, given the signal phase and timing (SPaT) message from upcoming traffic intersections, autonomous vehicles can cross intersections without stopping. The effectiveness of the proposed approach has been validated through Paramics microscopic traffic simulations by choosing a scenario in Statesboro, Georgia. For different traffic demands, the travel time and fuel consumption rate of vehicles are compared between non-PCC and PCC algorithms. Microscopic traffic simulation results show that the proposed PCC algorithm is able to reduce both fuel consumption rate and travel time.

Keywords: Optimization, Machine learning, Learning
Abstract: Online convex optimization against dynamic comparators in the literature is limited to linear models. In this work, we relax this requirement and propose a memory-efficient emph{online} universal function approximator based on compressed kernel methods. Our approach hinges upon viewing non-stationary learning as online convex optimization with dynamic comparators, for which performance is quantified by dynamic regret. Prior works control dynamic regret growth only for linear models. In contrast, we hypothesize actions belong to reproducing kernel Hilbert spaces (RKHS). We propose a functional variant of online gradient descent (OGD) operating in tandem with greedy subspace projections. Projections are necessary to surmount the fact that RKHS functions have complexity proportional to time. For this scheme, we establish sublinear dynamic regret growth in terms of the functional path length, and that the memory of the function sequence remains moderate. Experiments demonstrate the usefulness of the proposed technique for online nonlinear regression and classification problems with non-stationary data.

Keywords: Optimization algorithms, Learning, Distributed control
Abstract: This paper considers distributed bandit online optimization with time-varying coupled inequality constraints. The global cost and the coupled constraint functions are the summations of local convex cost and constraint functions, respectively. The local cost and constraint functions are held privately and only at the end of each period the constraint functions are fully revealed, while only the values of cost functions at queried points are revealed, i.e., in a so called bandit manner. A distributed bandit online primal-dual algorithm with two queries for the cost functions per period is proposed. The performance of the algorithm is evaluated using its expected regret, i.e., the expected difference between the outcome of the algorithm and the optimal choice in hindsight, as well as its constraint violation. We show that mathcal{O}(T^{c}) expected regret and mathcal{O}(T^{1-c/2}) constraint violation are achieved by the red{proposed} algorithm, where T is the total number of iterations and cin[0.5,1) is a user-defined trade-off parameter. Assuming Slater's condition, we show that mathcal{O}(sqrt{T}) expected regret and mathcal{O}(sqrt{T}) constraint violation are achieved. The theoretical results are illustrated by numerical simulations.

Keywords: Mean field games, Game theory, Large-scale systems
Abstract: While the topic of mean-field games (MFGs) has a relatively long history, heretofore there has been limited work concerning algorithms for the computation of equilibrium control policies. In this paper, we develop a computable policy iteration algorithm for approximating the mean-field equilibrium in linear-quadratic MFGs with discounted cost. Given the mean field, each agent faces a linear-quadratic tracking problem, the solution of which involves a dynamical system evolving in retrograde time. This makes the development of forward-in-time algorithm updates challenging. By identifying a structural property of the mean-field update operator, namely that it preserves sequences of a particular form, we develop a forward-in-time equilibrium computation algorithm. Bounds that quantify the accuracy of the computed mean-field equilibrium as a function of the algorithms stopping condition are provided. The optimality of the computed equilibrium is validated numerically. In contrast to the most recent/concurrent results, our algorithm appears to be the first to study infinite-horizon MFGs with non-stationary mean-field equilibria, though with focus on the linear quadratic setting.

Keywords: Iterative learning control, Large-scale systems, Hierarchical control
Abstract: We propose a new reinforcement learning based approach to designing hierarchical linear quadratic regulator (LQR) controllers for heterogeneous linear multi-agent systems with unknown state-space models and separated control objectives. The separation arises from grouping the agents into multiple non-overlapping groups, and defining the control goal as two distinct objectives. The first objective aims to minimize a group-wise block-decentralized LQR function that models group-level mission. The second objective, on the other hand, tries to minimize an LQR function between the average states (centroids) of the groups. Exploiting this separation, we redefine the weighting matrices of the LQR functions in a way that they allow us to decouple their respective algebraic Riccati equations. Thereafter, we develop a reinforcement learning strategy that uses online measurements of the agent states and the average states to learn the respective controllers based on the approximate Riccati equations. Since the first controller is block-decentralized and, therefore, can be learned in parallel, while the second controller is reduced-dimensional due to averaging, the overall design enjoys a significantly reduced learning time compared to centralized reinforcement learning.

Keywords: Optimization, Identification, Estimation
Abstract: The quality of data plays an important role in determining the accuracy of battery parameter identification/estimation, and hence data optimization for estimation has received increasing attention lately. The main idea is to design input excitation that is most sensitive to the target parameter(s) under estimation. However, most existing studies are hampered by the complexity of sensitivity computation/optimization, and need to impose heuristic patterns on the input to facilitate the optimization. Consequently, they are not capable of finding the ultimate optimal input pattern, and explaining the underlying mechanism. This is especially the case for the parameters of the first-principle electrochemical battery model. This paper aims at performing direct optimization of input excitation with no imposed pattern to find the ultimate optimal profile for estimating battery electrochemical parameters. The practice is enabled by the analytic expressions of sensitivity derived in our previous work. Based on the optimization results, we will explore the features/patterns of optimal profiles for different parameters by correlating to the analytic sensitivity expressions.

Keywords: Identification, Estimation, Energy systems
Abstract: We address the problem of identifying the parameters of a polymer electrolyte membrane (PEM) fuel cell model when no prior parameter estimates are available. To this end, we build upon a recently developed parameter identification framework and use an extended local sensitivity analysis to obtain a more global picture of parameter sensitivities. The extended analysis consists of local analyses carried out at multiple sampled points in the parameter space. The results from this extended analysis are then used to optimally select a subset of parameters for identification. Particularly, the selected subset optimizes the expected value of the well-known D-criterion over the parameter space. Being derived from the extended analysis, the selected subset is robust to initial assumptions about the nominal parameter values. Similar procedures are then used for robust optimal experimental design (OED) for the purpose of parameter identification. The robust OED approach is benchmarked against another experimental design method based on Latin Hypercube Sampling (LHS). The effectiveness of the proposed methods is investigated by identifying model parameters using synthetic data. The results demonstrate the utility of the robust parameter subset selection and OED procedures in enabling accurate parameter identification.

Keywords: Energy systems, Identification, Modeling
Abstract: In this paper, a lumped fractional order model (FOM) is developed for a lithium-ion battery (LIB). Electrochemical impedance spectroscopy (EIS) is used to obtain the cell impedance at different temperatures and SOCs, and at a wide range of frequencies. This data is used for the model parameterization which is a complex nonlinear least squares problem. A modified Levenberg-Marquardt algorithm is adopted to solve this problem. Two statistical measures, including sum of squares of residual and coefficient of determination are employed to incrementally build up on the integer order model (IOM) and obtain the final FOM. A derivative-based parameter sensitivity analysis is performed to study the influence of parameters on the model by considering their interactions. The terminal voltage of the FOM is simulated by mathematical approximation of the fractional derivative to study the model accuracy under temporal duty cycles. A comparative analysis between the fractional and integer order models is performed to gain insight on pros and cons of each model.

Keywords: Modeling, Energy systems
Abstract: In this paper, a novel physics-based modeling framework is developed for lithium ion battery packs. To address a gap in the literature for pack-level simulation, we establish a high fidelity physics-based model that incorporates electrochemical-thermal-aging behavior for each cell and pack-level electrical and thermal interactions, as well as cell heterogeneity. Governing equations in the form of Partial Differential Equations (PDEs) are discretized into a system of Ordinary Differential Equations (ODEs) using Finite Difference and Finite Volume methods and reformulated into state-space models for both cell and pack dynamics. Computational time studies are conducted to demonstrate the effects of spatial discretization fidelity and pack size on simulation time. Pack model predictive capabilities are exercised through sensitivity analysis of cell design parameters. Effects of parameter perturbation are shown for pack voltage and energy responses. The goal for this modeling framework is to provide a computationally-feasible and easily scalable framework for high-fidelity offline simulation and optimization without compromising the integrity of cell dynamics across multiple time scales.

Keywords: Energy systems, Nonlinear systems identification, Identification for control
Abstract: This paper focuses on the problem of online parameter estimation in an electrochemical Li-ion battery model. Online parameter estimation is necessary to account for model mismatch, environmental disturbances, and cycle-induced aging in Li-ion battery models. Sensitivity analysis can improve parameter estimation by identifying which data the parameters are most sensitive to. However, computing parameter sensitivity in full-order electrochemical models is typically intractable for online applications. Using a reduced-order model can lower the computational burden and, as we demonstrate, approximates well the sensitivity of the higher-order model. To provide further insight into the parameter estimation challenge, we analyze the effect that identifying parameters according to voltage RMSE data has on internal state errors. We perform a simulation study which demonstrates that parameter estimation approaches based on this paradigm are not sufficient for safe battery operation or other control objectives that require accurate estimates of these states.

Keywords: Identification, Optimization, Energy systems
Abstract: This paper investigates input trajectory optimization for parameter identifiability in a lithium-ion battery temperature cycling experiment. Such optimal experimental design can improve battery parameterization speeds and accuracies significantly. These potential improvements are well-established in the literature for thermal, electrochemical, and multi-physics battery models. However, to the best of our knowledge, the fundamental structure of the resulting optimal test trajectories is relatively less-explored. We examine the problem of optimizing the trajectory of thermal chamber temperature versus time in a lithium-ion battery temperature cycling test. We pose this as a Pareto-optimal control problem, with the competing objectives being the maximization of the Fisher identifiability of the battery’s thermal time constant versus the minimization of the L_2 norm of the control input. Pontryagin analysis reveals that the optimal trajectory is a switching trajectory constrained within battery cell temperature bounds, where the rate at which the solution proceeds from one bound to another is governed by the Pareto weight. Solving this problem numerically, using dynamic programming, supports these insights from Pontryagin analysis.

Keywords: Multivehicle systems, Autonomous robots, Fault detection
Abstract: This paper investigates fault detection, localization, and mitigation of autonomous vehicles platoons. The platoon is a network of autonomous vehicles that communicate together to move in a desired way. A fault in an autonomous vehicles platoon is a failure in either a physical component of a vehicle or a communication link between two vehicles in the platoon. This failure may lead to damage in one or more of the autonomous vehicles. Model-based health monitoring of a network of vehicles requires knowledge of a model of the system and the excitation signal, and thus may not be applicable. In this paper, we use measurements from available sensors in the platoon to identify sensor-to-sensor models that can be used for health monitoring, fault localization, and fault mitigation in the platoon. The dynamics of the network and the vehicles and the excitation signal that acts on the platoon are assumed to be unknown. We apply the proposed approach to a model of a platoon of autonomous vehicles and an experimental setup consisting of a platoon of three autonomous robots.

Keywords: Multivehicle systems, Cooperative control, Optimization
Abstract: This paper studies a cooperative air-ground routing problem, in which an unmanned aerial vehicle (UAV) and a ground vehicle coordinate with one another to visit a set of targets. The UAV, subject to limits on storage capacity and maximum flight distance, must rendezvous with the ground vehicle periodically, to upload the collected data and recharge. To solve such a cooperative routing problem, we formulate a chance-constrained mixed-integer linear program, where the chance constraint captures uncertainties in the model, such as uncertain speed or travel time of the two vehicles. Consequently, the obtained solution is less conservative compared to those found using robust methods, since only the worst case realization of uncertainties is considered in the later methods. Also novel to our approach, is that air ground rendezvous locations are not restricted to occur only at targets. Instead, the two vehicles are free to meet at any point on the UAV's path, which significantly increases the feasible solution space and leads to better solutions. We corroborate the effectiveness of the proposed approach through a set of numerical simulations on randomly generated instances.

Keywords: Multivehicle systems, Autonomous systems, Estimation
Abstract: In this paper, we investigate the motion prediction of human-driven vehicles in traffic flow mixed with connected autonomous vehicles (CAVs). Here, CAVs indicate the self-driving vehicles that can exchange messages with other vehicles through wireless vehicle-to-vehicle (V2V) communication, while human-driven vehicles are not equipped with V2V communication devices. The recursive least square (RLS) method is applied to approximate the dynamic model of the human-driven vehicles immediately ahead of CAVs, and the identified model is further used to predict the future motion of human-driven vehicles. To improve the prediction accuracy, triggering time and physical constraints are included in the predictor. Numerical simulations are utilized to validate the predictor. The effects of uncertain human reaction delays, time-varying parameters of human-driven vehicles, and stochastic packet drops in V2V communication on the accuracy of RLS prediction are also investigated.

Keywords: Multivehicle systems, Autonomous systems, Delay systems
Abstract: On a connected vehicle model with sensing, communication and human reaction delays, here we present an approach to investigate the stability of the equilibrium flow dynamics on the plane of model parameters. The approach is next utilized to reveal how the number of vehicles as well as penetration rates of sensors and communication lines in the arising vehicle network could influence the stability regions with respect to model parameters.

Keywords: Multivehicle systems, Traffic control, Stochastic systems
Abstract: The behavior of the optimal velocity model is investigated in this paper. Both deterministic and stochastic perturbations are considered in the Optimal velocity model and the behavior of the dynamical systems and their convergence to their associated averaged problems is studied in detail.

Keywords: Multivehicle systems, Traffic control, Autonomous systems
Abstract: Traffic flow models have been the subject of extensive studies for decades. The interest in these models is both as the result of their important applications as well as their complex behavior which makes them theoretically challenging. In this paper, an optimal velocity dynamical model is considered and analyzed. We consider a dynamical model in the presence of perturbation and show that not only such a perturbed system converges to an averaged problem, but also we can show its order of convergence. Such understanding is important from different aspects, and in particular, it shows how well we can approximate a perturbed system with its associated averaged problem.

Keywords: Cooperative control, Multivehicle systems, Optimal control
Abstract: The proliferation of automatic control systems and their connectivity accents the performance of their interactions. In particular, connected and automated vehicles could become high-impact examples thanks to their energy use and productivity effects. Ideally, a controller might collectively optimize all agents' control moves for a given objective. However, limits on computational complexity, incomplete knowledge of the central planner, and risks of a single point of failure make distributed control attractive as well. This paper proposes a collaborative heuristic to approach the performance of centralized control with decentralized-like computational effort. The related centralized controller is also described in detail and evaluated as a baseline. A collaboration-intensive obstacle avoidance scenario involving electric vehicles is simulated to demonstrate benefits over fully decentralized control. While centralized optimization performed best with an 8.6% energy reduction, the collaborative decentralized scheme reached a favorable computation-performance tradeoff with a 6.7% energy reduction.

Keywords: Automotive control, Variable-structure/sliding-mode control, Control applications
Abstract: In this paper, we propose a scheme of lane change control for automated vehicle without path planning and its tracking. It is not easy to make path planning for lane change although there are several method for the path planning such as using a combination of sinusoidal functions and high order polynomial functions. In this paper, time-varying hyperplane is utilized to cope with the problem of path planning and control for lateral motion in lane change control. Designing a sliding hyperplane in terms of lateral position and velocity is presented so that the lateral error converges uniformly and smoothly during lane changing irrespective of the amount of lateral offset. The simulation-based optimization approach is utilized to obtain the optimal time-varying sliding hyperplane. The stability of the closed-loop system is proved with the analysis of a discrete time-varying system. The effectiveness of the proposed method is validated with numerical simulation showing the uniform settling of lateral tracking error no matter what the desired lateral offset is commanded.

Keywords: Constrained control, Optimal control, Autonomous systems
Abstract: Connected and automated vehicles (CAVs) provide the most intriguing opportunity to reduce pollution, energy consumption, and travel delays. In earlier work, we addressed the optimal coordination of CAVs using Hamiltonian analysis. In this paper, we investigate the nature of the unconstrained problem and provide conditions under which the state and control constraints become active. We derive a closed-form analytical solution of the constrained optimization problem and evaluate the solution using numerical simulation.

Keywords: Automotive systems, Sensor fusion, Learning
Abstract: Radars, LiDARs and cameras have been widely adopted in autonomous driving applications due to their complementary capabilities of environment perception. However, one problem lies in how to effectively improve the cross-area tracking accuracy with massive data from multiple sensors. This paper proposes a novel tracking solution that is composed of a reinforcement-learning-based Gaussian mixture model (GMM), submodel center realignment, and data-driven trajectory association. Specifically, developed with a Q-learning-based cluster number, an improved GMM-EM algorithm is firstly investigated to cluster the dense short-range radar data points. Subsequently, an innovative kinetic-energy-aware approach is presented to realign the Q-learning GMM cluster centers for position error mitigation. In addition to Q-learning GMM clustering, a weight-scheduled method is presented to associate the data from a long-range radar and cameras for cross-area object extraction and trajectory fusion. Eighteen experiments for training and one experiment for verification were conducted on a fully-instrumented autonomous vehicle. Experimental results demonstrate that a better tracking performance in crossing detection areas can be achieved by the proposed method.

Keywords: Automotive systems, Estimation
Abstract: Real-time traffic prediction is crucial for optimization and control of connected and autonomous vehicles (CAVs). The knowledge of preceding vehicle’s future trajectory determines the bounds of car-following distance for the target vehicle. A key challenge of traffic prediction is to handle mixed traffic scenarios where both CAVs and non-CAVs co-exist. In this work, a traffic prediction framework is developed based on traffic flow model. Information from connected vehicles (CVs) provides partial measurement of traffic states (traffic speed and traffic density). The unknown traffic states can be estimated using a state observer. Once the full traffic states are known, future traffic states can be predicted by propagating the traffic flow model forward in time. With on-board perception sensors, CVs are capable of detecting locations and speeds of adjacent vehicles. This additional information can potentially improve the performance of the traffic prediction. The proposed traffic prediction framework is comprehensively evaluated for a signalized roadway for various penetration rates of connectivity and locations of CVs. Simulation results show that additional information from perception sensors can help reduce the prediction error by 25%.

Keywords: Automotive systems, Autonomous systems, Formal verification/synthesis
Abstract: We propose a decision-making system for automated driving with formal guarantees, synthesized from Signal Temporal Logic (STL) specifications. STL formulae specifying overall and intermediate driving goals and the traffic rules are encoded as mixed-integer inequalities and combined with a simplified vehicle motion model, resulting in a mixed-integer optimization problem. The specification satisfaction for the actual vehicle motion is guaranteed by imposing constraints on the quantitative semantics of STL. For reducing the computational burden, we propose an STL encoding that results in a block-sparse structure. The same STL formulae are used for monitoring faults due to imperfect prediction on the vehicle and environment. We demonstrate our method on an urban scenario with intersections, obstacles, and no-pass zones.

Keywords: Modeling, Predictive control for nonlinear systems, Manufacturing systems
Abstract: This paper presents a learning-based approach to modeling and control of inkjet 3D printing. First, we propose and experimentally validate a learning-based model for inkjet 3D printing. The proposed model uses a physics-based model paradigm that has been reformulated into a neural-network-like structure. This formulation enables back-propagation and the associated benefits of data-driven model identification while retaining physical interpretation of the learned model itself. Next, we propose and demonstrate a predictive control algorithm that leverages the neural-network-like structure of the model. Back-propagation is used for efficient gradient calculations to determine optimal control inputs, namely droplet patterns for subsequent layer(s), to optimize a quadratic cost function.

Keywords: Identification for control, Reduced order modeling, Control of metal processing
Abstract: This paper addresses process modeling for the selective laser melting (SLM) process. We experimentally investigate the response of the SLM process output (measured by a coaxial near-infrared camera) to changing input laser power. We determined that first and second order models can be used to capture this input-output behavior. Next, we studied the dependency of this transfer function on laser scan speed and other process variables that evolve over a typical part build, such as thermal properties of surrounding medium (bulk powder, build plate, or solidified part) or layer number. The transfer function was found to strongly depend on the material environment (solidified material or bulk powder). Further, transfer function also depended on the layer number, exhibiting transient behavior. We report identified 1st order transfer functions for different scan speeds, locations on the build plate, and different layer numbers. Identified models and quantification of their variability will serve as foundational work for the future implementation of advanced real-time process control algorithms.

Keywords: Modeling, Manufacturing systems, Model Validation
Abstract: Additive manufacturing (AM) is a digital manufacturing technology that manufactures a 3D object in a bottom-up and layer-by-layer fashion. Fused deposition modeling(FDM), also known as desktop 3D printing, is one of the most commonly used AM technologies with numerous applications in academia and industry. Some of the greatest challenges with FDM include poor repeatability and reliability of the process, leading to mid-process failures or out-of-spec final products. Closed-loop control applications for FDM have been proposed as a means of mitigating mid-process failures. However, no models currently exist to enable control of the bead cross-sectional dimensions for the extruded material. This work presents a control-oriented model describing the effect of process parameters on cross-sectional dimensions of the deposited beads in FDM. A geometric model is presented and a procedure to evaluate the unknown machine and material-specific parameters in the model is provided by leveraging design of experiments. The proposed model is experimentally validated and the accuracy of the results is presented. The results show that the proposed model accurately represents the bead cross-sectional geometry and is suitable for closed-loop control applications.

Keywords: Manufacturing systems, Iterative learning control, Modeling
Abstract: Selective Laser Melting (SLM) is a common additive manufacturing technique which uses a scanning laser source to fuse metal powder layer by layer. Although complex geometries can be produced, quality and repeatability of parts are still two challenges due to complex physical transformations of the metal powder and highly dynamic temperature fields. Finite Element Models (FEMs) have been developed by researchers in order to predict melt pool behavior. However, simulations on FEM software are too computationally intensive for real-time control applications. Thus, there arises the need for a control-oriented model of SLM processes. In this paper, a state-space control-oriented layer-to-layer model based on the general heat conduction equation is developed. The layer-to-layer model is constructed to step one from layer’s thermal feature measurement to the next, thus reducing computational complexity to a level suitable for control. To validate the model, an experiment of a rectangular thin part was conducted, and the simulation described the experimental thermal measurements 5% error in the output.

Keywords: Control of metal processing, Iterative learning control, Manufacturing systems
Abstract: Additive manufacturing processes fabricate parts in a layer-by-layer fashion, depositing material along a pre-defined path before incrementing to the next layer. Although the thickness of any given layer is bounded, in-layer dynamics can couple with layer-to-layer dynamics such that height defects amplify from one layer to the next. This is considered instability in the layer domain. By considering each layer as an iteration, additive processes can be categorized as repetitive processes. Although Repetitive Process Control (RPC) algorithms exist that can stabilize the process and converge to desired reference, it is typically assumed that the reference and disturbance are constant from layer to layer. In this paper, the problem of tracking references (layer thicknesses) that change from layer to layer is considered. The bandwidth of the changing references is considered bounded in both the spatial and layer domains. A loop-shaping design process is then considered, in which the bounds are mapped to a bound on the two-dimensional sensitivity function and projected onto weighting filters in an LQR control formulation. The layer-to-layer controller is then constructed from traditional LQR methods. The controller is demonstrated on a simulation of laser metal deposition for a wavy wall build having frequency content in both the spatial and layer domains.

Keywords: Agents-based systems, Manufacturing systems, Machine learning
Abstract: Adaptive manufacturing has revolutionized desktop prototyping and the production of physical models for non-load bearing or stress inducing applications. Many extrusion-based printers are available for purchase by entrepreneurial enthusiasts or businesses with manufacturing space limitations. These low-cost printers allow for quick prototyping but are not designed or intended for high quality production or high-cycle production, requiring extensive user tuning and upkeep to maintain the printer in usable condition. In a quest to apply modern deep learning and reinforcement learning based models, this work focuses on the development of control systems and infrastructure needed to resolve many of these intrinsic limitations of desktop 3D printers. A series of real-time agents were designed and deployed to actively monitor the printing of every layer and make continuous corrections in the printing parameters and G-code commands to reduce the variance in the tensile strength of homogeneous parts printed in a large batch.

Keywords: Energy systems, Transportation networks, Smart grid
Abstract: Major Transportation Network Companies (TNCs), including Uber and Lyft, are in the process of electrifying their fleets. These electrified fleets can provide electricity service in addition to transportation service. This paper examines the problem of optimal spatial pricing for an electrified TNC jointly providing transportation and electricity services. When the system demand is spatially balanced, we establish a tight characterization of the optimal pricing policy and the platform profit. Furthermore, we highlight the role electricity service may provide in reducing the transportation spatial imbalance and characterize the resulting synergetic value.

Keywords: Smart grid, Delay systems, Networked control systems
Abstract: A novel framework is presented for design of wide-area-control (WAC) in smart grid power systems. The problem setup regards scenarios where control suffers from noisy measurement and time-lags. We evaluate the risk of phase incoherence in interconnected power machines. We outline the effect of systemic parameters, information flow and disturbances on risk of phase deviations. Our results suggest limitations of delay-aware control in presence of multiple sources of noise. We develop algorithmic network design tools to enhance and, in some cases, optimize coherency and robustness of the closed-loop system.

Keywords: Smart grid, Distributed control, Markov processes
Abstract: There is enormous flexibility potential in the power consumption of the majority of electric loads. This flexibility can be harnessed to obtain services for managing the grid: with carefully designed decision rules in place, power consumption for the population of loads can be ramped up and down, just like charging and discharging a battery, without any significant impact to consumers' needs. The concept is called Demand Dispatch, and the grid resource obtained from this design virtual energy storage (VES). In order to deploy VES, a balancing authority is faced with two challenges: 1. how to design local decision rules for each load given the target aggregate power consumption (distributed control problem), and 2. how to coordinate a portfolio of resources to maintain grid balance, given a forecast of net-load (resource allocation problem). Rather than separating resource allocation and distributed control, in this paper the two problems are solved simultaneously using a single convex program. The joint optimization model is cast as a finite-horizon optimal control problem in a mean-field setting, based on the new KLQ optimal control approach proposed recently by the authors. The simplicity of the proposed control architecture is remarkable: With a large portfolio of heterogeneous flexible resources, including loads such as residential water heaters, commercial water heaters, irrigation, and utility-scale batteries, the control architecture leads to a single scalar control signal broadcast to every resource in the domain of the balancing authority.

Keywords: Smart grid, Power systems, Agents-based systems
Abstract: Owning to the capacity constraints and the uneven distribution of resources, energy management problem in energy internet is a major concern. To cope with the variations and complexity of large scale energy management, a distributed actor-critic reinforcement learning based method is proposed for optimal energy management. First, the intelligent action is decided in the distributed agent to alleviate the pressure on centralized intelligent computing. The distributed action of each agent is based on its neighbour information, and an actor-critic reinforcement learning algorithm is applied for dealing with the continuous action space. Then, aiming at the supply-demand balance, the action is adjusted based on global information exchange. After action adjustment, the corresponding rewards are sent to each agent. Finally, the modified action is executed in each agent. And received rewards are utilized to update each agent. Simulation driven by Pecan Street Inc.’s Dataport demonstrates that the proposed distributed method is effective.

Keywords: Smart grid, Power systems
Abstract: Thermostatically Controlled Loads (TCLs), e.g., A/Cs and water heaters, are a source of flexible power demand for the power grid: many different power consumption trajectories exist that can maintain consumers' quality of service (QoS). Extensive research has shown that flexible loads can provide valuable grid services. Quantifying the flexibility capacity of a collection of TCLs is a well-studied problem. However, many studies consider temperature constraints alone, while most TCLs are on/off loads that have cycling (or lock-out) constraints. Studies that have considered lock-out constraints have proposed quantifications that depend on the control algorithm used to coordinate loads to provide grid services.

In this work, we present a characterization of the capacity of a collection of TCLs that considers not only temperature, but also cycling and total energy constraints. Our characterization is independent of the algorithm used to control the TCLs; it depends only on the QoS constraints on the individual TCLs. The proposed characterization can be used for planning a feasible power deviation trajectory for a collection of TCLs by solving a convex optimization problem.

Keywords: Smart grid
Abstract: Flexible loads are a valuable resource for the power grid of the future to help with balancing demand and generation. A balancing authority (BA) needs to know how much flexibility a load has, meaning what type of power deviation (from the baseline demand) signals are feasible for the load. In this work we present a characterization of capacity for a flexible load in terms of the power spectral density of the power deviation. We then show how this characterization can be used for resource allocation for the grid by determining what portion of the grid's needs can be met by a collection of such loads. The key difference with prior work on flexibility characterization is that ours is posed in terms of the statistical properties grid's net load and load's demand deviation, not on specific instances of these signals. The proposed characterization can thus be used for long term planning.

Keywords: Biomedical, Modeling, Nonlinear systems identification
Abstract: A method for estimating the instantaneous heart rate (HR) using the morphological features of one electrocardiogram (ECG) cycle (beat) at a time is proposed. This work is not aimed at introducing an alternative way for HR estimation, but rather illustrates the utility of model-based ECG analysis in online individualized monitoring of the heart function. The HR estimation problem is reduced to fitting one parameter, whose value is related to the nine parameters of a realistic nonlinear model of the ECG and estimated from data by nonlinear least-squares optimization. The method feasibility is evaluated on synthetic ECG signals as well as signals acquired from MIT-BIH databases at Physionet website. Moreover, the performance of the method was tested under realistic free-moving conditions using a wearable ECG and HR monitor with encouraging results.

Keywords: Lyapunov methods, Adaptive control, Robotics
Abstract: Hybrid exoskeletons combine two increasingly common rehabilitative therapies, functional electrical stimulation (FES) and robotic therapy, for use on individuals with neuromuscular disorders. As hybrid exoskeletons increase in popularity and complexity, it remains an ever-important issue to not only assist people in performing rehabilitation, but also to guarantee their safety while coupled to the exoskeleton. In this paper, a novel adaptive controller for hybrid exoskeletons is developed to regulate an admittance error system using the exoskeleton’s motors while simultaneously regulating a position error system using the operator’s muscles, stimulated through FES. The stability of the controller is rigorously analyzed using a combined Lyapunov-passivity approach and while the hybrid exoskeleton is proven to be energetically passive, the admittance error system is proven to demonstrate global exponential convergence to a uniform ultimate bound. Simulations were performed on a two degree-of-freedom lower-limb hybrid exoskeleton to demonstrate the efficacy of the controller. Results show the controller achieves an average admittance tracking error of 0.0+/-0.08 rad and 0.00+/-0.08 rad/s for joint one (the knee joint), and 0.01+/-0.11 rad and 0.01+/-0.12 rad/s for joint two (the ankle joint), while simultaneously applying FES to the operator’s muscles for rehabilitation.

Keywords: Biomedical, Variable-structure/sliding-mode control, Fluid power control
Abstract: This paper presents the design and kinematic analysis of a Haptic Glove for medical elastographic imaging virtual palpation. Of the 12 degrees of freedom present in the index finger, middle finger, and thumb of the hand, the design constrains 5 and controls 6 with pneumatic air cylinder actuators, allowing uncontrolled, but measured motion in the remaining degree of freedom. Nearly linear bijective transfer functions between the actuator positions and joint angles are found in closed form for all 6 actuated joints. A nonlinear force and stiffness controller based on backstepping-sliding mode algorithm was designed and implemented.

Keywords: Biomedical, Modeling, Computational methods
Abstract: This paper explores the gait learning and coordination through physical human-human interaction. The interaction and coordination are modeled as a two-step process: 1) encoding the human gait as a periodic process and 2) adjustment of the periodic gait cycle based on the external forces due to physical interactions. Three-legged walking experiments are conducted with two human dyads. Magnitude and direction of the interaction force, as well as the knee joint angles and ground reaction forces of the tied legs are collected. The knee joint trajectory of the two participants is modeled using dynamic movement primitives (DMP) coupled with force feedback through iterative learning. Gait coordination is modeled as a learning process based on kinematics from the last gait cycle and real-time interaction force feedback. The proposed method is compared with a popular baseline DMP model, which performs batch regression based on data from the previous gait cycle. The proposed model performed better in modeling one pair in the cooperative experiment compared to the baseline algorithm. The results and approaches for improving the algorithm are further discussed.

Keywords: Biomedical, Predictive control for nonlinear systems, Control applications
Abstract: Previous studies have proposed closed-loop methods to determine one of the vagal nerve stimulation (VNS) parameters based on the feedback of heart rate but never consider the blood pressure. This study develops a nonlinear model predictive controller (NMPC) for a multi-location VNS system, which controls heart rate (HR) and mean arterial pressure (MAP) by independently manipulating multiple stimulation parameters in three different locations. Pulsatile and non-pulsatile models of the integrated cardiovascular system and baroreflex regulation are developed to predict the dynamic response of the MAP and HR. The two models are compared and elicit similar dynamics. The NMPC algorithm incorporates the non-pulsatile model and is evaluated by the pulsatile model. Three set point tracking cases are presented and discussed. The performance of NMPC shows the feasibility and usefulness of the proposed control algorithm for the regulation of HR and MAP during multi-location VNS.

Keywords: Robotics, Human-in-the-loop control, Optimization
Abstract: Kinematic control approaches for exoskeletons follow specified trajectories, which overly constrain individuals who have partial or full volitional control of their lower-limbs. In our prior work, we proposed a general matching framework for underactuated energy shaping to provide task-invariant, energetic exoskeletal assistance. While the proposed shaping strategies demonstrated benefits such as reduced human torques during walking, it remains unclear how the parameters of these shaping strategies are related to the torque reduction. Meanwhile, research indicates that customizing assistance via optimization techniques can substantially improve exoskeleton's performance for each individual. Motivated by this fact, we combine derivative-free, sample-efficient optimization algorithms with our energy shaping strategies to propose a task-invariant learning framework for lower-limb exoskeletons. Through rapid online optimization, this framework enables exoskeletons to adjust shaping parameters for minimizing human joint torques across users and tasks. Simulation results show that shaping strategies with optimal parameters effectively reduce human joint torques and metabolic cost during simulated walking. In addition, the optimal exoskeleton torques calculated using able-bodied subjects' kinematic data closely match the real human joint torques across locomotor tasks.

Keywords: Control applications, Estimation, Filtering
Abstract: Atomic force microscopy (AFM) is a powerful instrument that has the ability to characterize sample topography on nanoscale resolution. AFM is widely used in different fields, such as nanotechnology, semiconductor, Microelectromechanical Systems (MEMS), bioscience. In the case of obtaining 3D topography of a large range sample, we need to know the relative position of the AFM probe to the sample. The scanning range of an AFM generally is much smaller than the sample size. Therefore, it is hard to localize the AFM tip position without other auxiliary microscopes such as optical microscope. Moreover, the AFM scanned images on a MEMS sample typically involve only simple geometries with sparse features which usually leads to the difficulty of localization. Besides, the system uncertainties including piezoelectric scanner hysteresis, thermal drift, and coarse dual stage would affect positioning accuracy. In this paper, we propose an AFM tip localization method using particle filter referring to macro robot Simultaneous localization and mapping (SLAM). We take the AFM scanned image as the unique sensor and the sample layout as the map. The sensor model of the particle filter is based on a feature extraction algorithm. To verify the efficacy of the proposed methods, both simulations and experiments are conducted, and the proposed tip localization method is highly promising.

Keywords: Optimal control, Learning, Robotics
Abstract: We study the problem of taking simulations to the real world (RW) for autonomous robotic systems with dynamic uncertainties and unknown disturbances while maintaining the optimal performance and stability of the designed controller designed in simulation. In general, an optimal and robust controller that is designed through simulation often does not perform similarly when deployed in the RW. We focus on using simulations to generate an optimal control policy utilizing the Memetic algorithm (MA) iteratively. The simulation-to-RW performance and stability are realized by using an adaptive fuzzy system to learn the uncertain part of the dynamic model, disturbance and noises. We demonstrate experimentally that this method permits the development of optimal control design in simulations and integrates adaptive learning rules to enable precise and repetitive trajectory tracking for the wheeled mobile robot (WMR) with disturbances and uncertainties.

Keywords: Mechatronics, Observers for nonlinear systems, Fluid power control
Abstract: Position feedback in a solenoid actuated system typically requires a position sensing device such as a Linear Variable Differential Transformer (LVDT). The goal of self-sensing is to obtain position information directly from the electrical signals to the solenoid actuators, thus obviating the additional cost and footprint of a LVDT or another displacement sensing device. Such measurement is possible due to the position dependence of electrical inductance in the solenoids. This paper proposes a finite-dimensional nonlinear observer for the magnetic flux linkage for the solenoids. Once the flux linkage has been identified, the solenoid position can be determined via the position-inductance relationship. The algorithm has been adapted for actual solenoids modeled as a third-order system that includes two eddy current modes accurate up to 1024 Hz. Implementation on commercially low-cost solenoids (with 5mm stroke) has demonstrated RMS position accuracy up to 0.061mm. The ability to self-sense accurately can enable solenoids to be deployed at low-cost for many motion control applications besides hydraulic valves.

Keywords: MEMs and Nano systems, Mechatronics
Abstract: This paper presents a software-hardware integrated approach to high-speed large-range dynamic mode imaging of atomic force microscope (AFM). High speed AFM imaging is needed to interrogate dynamic processes at nanoscale such as chemical reactions. High-speed dynamic-modes such as tapping-mode AFM imaging are challenging as the probe tapping motion is highly sensitive to the highly nonlinear probe-sample interaction during the imaging process. The existing hardware-based approach via bandwidth enlargement, however, results in a substantial loss of imaging area that can be covered. Contrarily, software-based approach, for example, the recently developed adaptive multiloop mode (AMLM) technique has demonstrated its efficacy in increasing the tapping-mode imaging speed without loss of imaging size. However, further improvement has been limited by the the hareware bandwidth and the online signal processing speed and computation complexity.Thus, in this paper, the AMLM technique is furhter enhanced and integrated with the FPGA platform to further increase the imaging speed without loss of quality and imaging range. Issues in FPGA implementation in AFM imaging experiment is presented and discussed to illustrate this integrated approach.

Keywords: Mechatronics, Control applications, Networked control systems
Abstract: Atomic force microscope (AFM) is able to perform high resolution 3-D topography image at a nanometer resolution. This paper demonstrates the amplitude-detection mode atomic force microscopy (AM-AFM) with the proposed modified cycloid trajectory (MCT) for lateral-axes. Besides, the internal model principle-based neural network complementary sliding mode control (IMP-based NNCSMC) approach of designing controller is implemented for the xy-piezoelectric scanner to overcome some non-linear uncertainties or disturbance. Furthermore, the MCT method is a smooth and cycloid-like scan pattern that can avoid the containing frequency in fast axis signal beyond mechanical bandwidth of the scanner such that AFM can achieve higher scan speed than a raster scan. Finally, some comparison results between MCT and raster scan will be provided.

Keywords: Adaptive control, Adaptive systems, Delay systems
Abstract: In this paper, we address the design and analysis of multivariable extremum seeking for unknown static maps subject to arbitrarily long time delays. Gradient-based method is considered. Multi-input systems with different time delays in each individual input channel are dealt with. The phase compensation of the dither signals and the inclusion of predictor feedback with a perturbation-based (averaging-based) estimate of the Hessian allow to obtain local exponential convergence results to a small neighborhood of the optimal point, even in the presence of delays. Unlike previous publications considering multiparameter extremum seeking and delays, the stability analysis is carried without using backstepping transformation, which also eliminates the complexity of the controller. In a nutshell, a simpler implementation scheme and direct analysis without invoking successive backstepping transformation can be assured. In particular, the delays in our approach are independent of the dither frequency and system's dimension such that fast convergence rates are still guaranteed. A numerical example illustrates the performance of the new delay-compensated extremum seeking scheme and its simplicity.

Keywords: Adaptive control, Adaptive systems, Direct adaptive control
Abstract: This paper describes the development of a model reference adaptive controller suitable for quadrotors equipped with manipulators capable of grasping and transporting objects of unknown mass. Through a Lyapunov argument, this estimation and control algorithm is shown to be asymptotically stable when tracking a desired reference trajectory. The adaptive controller is then demonstrated on a quadrotor of unknown mass, showing that it can track a reference trajectory with sufficient accuracy to pick up a payload and maintain stable flight, despite the sudden change in vehicle parameters induced by the payload. This new adaptive controller is slower to respond to changes in vehicle mass than control strategies that use more conventional mass estimation methods, but it is subject to less noise since it does not make use of accelerometer data, resulting in a smoother control trajectory.

Keywords: Adaptive control, Direct adaptive control, Delay systems
Abstract: The paper deals with the problem of simultaneous adaptive compensation of external disturbances and asymptotic reference tracking in linear time-invariant multi-input multi-output square plants with input delay and unmeasureble state. The plant to be controlled can be unstable. The external signals --- disturbance to be compensated and reference to be tracked --- are modeled as multi-harmonic signals with unknown frequencies, phases and amplitudes. The solution is based on suitable external signals parameterization and prediction as well as on special tracking error augmentation allowing one to overcome the problem of input delay in adaptation algorithm. Two adaptation algorithms using special augmented error are proposed: gradient-based algorithm and an algorithm with improved parametric convergence based on application of Kreisselmeier's scheme. No {em a priori} assumptions about knowledge of external signal parameters as well as about persistent excitation (PE) condition are needed for asymptotic reference tracking and disturbance rejection in the closed-loop system.

Keywords: Adaptive control, Direct adaptive control, Uncertain systems
Abstract: The paper addresses the problem of adaptive compensation of unmatched disturbance in nonlinear parametrically uncertain systems presented in parametric-strict-feedback form. The disturbance is modeled as a vector of unmeasurable multisinusoidal functions with a priori unknown amplitudes, frequencies and phases. The proposed solution is based on observer-based disturbance parameterization, modular backstepping design and special adaptation algorithm (identifier) with memory regressor extension. New adaptation algorithm proposed has two important properties. First, it has improved parametric convergence achieved by regressor recording over past period of time. Recording is provided by involving a linear SISO filter into the structure of the algorithm. Second, inclusion of this filter allows us to calculate the high-order time derivatives of the adjustable parameters used directly in virtual and actual control laws.

Keywords: Adaptive control, Fault tolerant systems, Stochastic systems
Abstract: In this paper, we develop distributed output feedback adaptive consensus control protocols for addressing networked multiagent systems subject to exogenous stochastic disturbances and sensor and actuators attacks. Specifically, for a class of linear leader-follower multiagent systems with an undirected communication graph topology we develop an output feedback adaptive control design protocol for each follower agent to address malicious attacks on the actuator signals of the follower agents as well as sensor attacks on the output neighborhood synchronization errors measurements. The proposed adaptive controllers involve an indirect adaptive architecure that estimates and compensates for the malicious attacks while guaranteeing uniform ultimate boundedness of the state tracking error for each agent in a mean-square sense.

Keywords: Adaptive control, Neural networks, Uncertain systems
Abstract: We introduced a {it working memory} augmented adaptive controller in our recent work. The controller uses attention to read from and write to the working memory. Attention allows the controller to read specific information that is relevant and update its working memory with information based on its relevance, similar to how humans pick relevant information from the enormous amount of information that is received through various senses. The retrieved information is used to modify the final control input computed by the controller. We showed that this modification speeds up learning.

In the above work, we used a soft-attention mechanism for the adaptive controller. Controllers that use soft attention update and read information from all memory locations at all the times, the extent of which is determined by their relevance. But, for the same reason, the information stored in the memory can be lost. In contrast, hard attention updates and reads from only one location at any point of time, which allows the memory to retain information stored in other locations. The downside is that the controller can fail to shift attention when the information in the current location becomes less relevant.

We propose an attention mechanism that comprises of (i) a hard attention mechanism and additionally (ii) an attention reallocation mechanism. %that effectively retains the benefit of hard attention and at the same time overcomes its limitation by reallocating attention whenever required. The attention reallocation enables the controller to reallocate attention to a different location when the relevance of the location it is reading from diminishes. The reallocation also ensures that the information stored in the memory before the shift in attention is retained which can be lost in both soft and hard attention mechanisms. %We illustrate through simulations that the memory that uses the proposed attention mechanism stores a more accurate representation of the variations in the hidden layer values of the neural network (NN). Through detailed simulations of various scenarios for two link robot robot arm systems we illustrate the effectiveness of the proposed attention mechanism.

Keywords: Autonomous robots, Automotive control, Machine learning
Abstract: Autonomous vehicle driving systems face the challenge of providing safe, feasible and human-like driving policy quickly and efficiently. The traditional approach usually involves a search or optimization-based planning followed by a model-based controller. This may prove to be inadequate in some driving scenarios due to disturbance, uncertainties and limited computation time. The more recent end-to-end approaches aim at overcoming these issues by learning a policy to map from sensor data to controls using machine learning techniques. Although being attractive for its simplicity, they also show some drawbacks such as sample inefficiency and difficulties in validation and interpretability. This work presents an approach that attempts to exploit both worlds, combining machine learning-based and model-based control into an imitation learning framework that mimic expert driving behavior while obtaining safe and smooth driving. The dataset is generated from high-fidelity simulations of vehicle dynamics and model predictive control (MPC). A smooth spline-based motion planning represents the policy provided by a constrained neural network exploiting the convex hull property of B-splines. The policy network is trained with few dataset aggregations coming from its induced distribution of states. The learned policy is used as guidance for model-based feedback control and tested on a 15DOF high fidelity vehicle model.

Keywords: Autonomous robots, Autonomous systems, Control applications
Abstract: In this work, we address the problem of real-time control and localization of an autonomous art exhibit designed using Mecanum wheeled omnidirectional mobile robots. Currently, the exhibit is animated using open-loop control as there are no practical means of implementing a localization system in art galleries. This paper proposes a nonlinear model predictive controller (NMPC) supported by a Decawave localization system for trajectory tracking of the robots. A Robot Operating System (ROS) based NMPC is implemented to control the motion of the robot for a smooth and drift-free trajectory tracking. The Automatic Control and Dynamic Optimization (ACADO) toolkit is used to find the optimal control action while considering all constraints on system states and inputs. A Local Positioning System (LPS) is implemented using Decawave 1001 modules to provide position feedback with an Optitrack motion capture system providing ground truth information for validation. The proposed control approach is first evaluated using the Virtual Robot Experimentation Platform (V-REP), which additionally provides means of trajectory design and simulation of the robot in the desired exhibition space. Thereafter, laboratory experiments are conducted to evaluate the performance of the proposed control and localization systems. The results show a smooth and drift-free performance of the system with less than 10% error of the robot size, which can be deployed at gallery spaces with minimal setup requirements.

Keywords: Autonomous robots, Control applications, Robotics
Abstract: This paper presents a framework for Simultaneous Localization and Mapping (SLAM) by combining a novel method for object discovery and localization from a monocular camera with depth information provided by Light Detection and Ranging (LiDAR). One major challenge in vision is discovering unknown objects without prior training/supervision, in the wild, and on-the-fly. In our framework, no training samples are available prior to the deployment. We develop an efficient proposal-matching method to discover object temporal saliency, and then fine- tune these frequently matched object proposals according to tracking information. Detected features of the objects are used as landmark features, and are merged with the LiDAR data in the proposed LIV-LAM (LiDAR and Visual Localization and Mapping). Compared to most visual SLAM or LiDAR-based SLAM, the novelty of this method is the computationally-efficient object detection and localization for feature set-and-match, in order to increase the accuracy of the generated map. The results show that the presented method is superior in both accuracy and efficiency of the maps generated by LiDAR.

Keywords: Autonomous robots, Control applications, Optimal control
Abstract: A nonlinear optimal (H-infinity) control approach is developed for the model of the ballbot. This robotic system consists of a rolling sphere with a rigid body, in the form of an inverted pendulum, mounted on top of the sphere. Because of the nonlinearities and underactuation that are due to the dynamics of both the rolling sphere and of the rotational motion of the rigid body, control of the ballbot is a non-trivial problem. In the proposed control method, the dynamic model of the ballbot undergoes first approximate linearization around a temporary operating point which is updated at each iteration of the control algorithm. The linearization process makes use of first-order Taylor series expansion and relies also on the computation of the Jacobian matrices of the state-space model of the robotic system. A stabilizing H-infinity feedback controller is designed for the approximately linearized description of the ballbot. An algebraic Riccati equation is solved at each time-step of the control method so as to compute the controller's feedback gains. Through Lyapunov analysis the stability properties of the control scheme are proven. Besides, through the article's results it is demonstrated that the control method retains the advantages of linear optimal control, that is fast and accurate tracking of the reference setpoints under moderate variations of the control inputs.

Keywords: Autonomous robots, Control applications, Simulation
Abstract: Mid-water localization is challenging for autonomous underwater vehicles (AUVs) due to limited communications and geo-referencing capabilities in the underwater environment, coupled with unknown complex and dynamic surroundings. Existing solutions typically utilize expensive sensors that may not be available to all AUVs. In this paper, we consider an AUV descending through the water column and propose an approach for mid-water localization using inertial and depth sensors only. During a descent of the vehicle, we leverage spiral motion, which allows for exploitation of vehicle dynamics along with associated inertial sensor measurements for localization. The spiral motion enables us to observe and estimate the influence of environmental flow (i.e., ocean currents) on the vehicle motion, thereby enhancing the understanding of the environment through active perception. The estimated flow together with inertial and depth sensor measurements are integrated in the vehicle motion model for localization. Comparing our approach with conventional dead-reckoning, the simulation results demonstrate its promising potential.

Keywords: Autonomous robots, Cooperative control, Transportation networks
Abstract: A control strategy for a leader-follower system without interagent communication is developed for cooperative transportation in this work. Two cascaded UKFs (unscented Kalman filters) are developed to estimate the external force of the leader as to eliminate the need of force sensors. Compared to most existing results, performance of the developed force estimators is invariant to lighting conditions since the developed UKFs do not require measurement from vision systems. To enhance robustness of the control scheme, a switching controller along with a triggering condition is developed for the follower, so that the impact to the performance of the closed-loop system caused by the disturbances can be minimized. Experiments are conducted to evaluate the control performance. Additionally, interesting phenomena are observed from the experiments and discussed, which can facilitate the improvement of the next generation cable-based transportation systems.

Keywords: Agents-based systems, Autonomous systems, Cooperative control
Abstract: Flocking behavior has attracted considerable attention in multi-agent systems. The structure of flocking has been predominantly studied through the application of artificial potential fields coupled with velocity consensus. These approaches, however, do not consider the energy cost of the agents during flocking, which is especially important in large-scale robot swarms. This paper introduces an optimal control framework to induce flocking in a group of agents. Guarantees of energy minimization and safety are provided, along with a decentralized algorithm that satisfies the optimality conditions and can be realized in real time. The efficacy of the proposed control algorithm is evaluated through simulation in both MATLAB and Gazebo.

Keywords: Agents-based systems, Adaptive control, Automotive systems
Abstract: This paper derives a principled framework for efficiently and safely navigating partially observable multi-agent systems using an optimal deterministic planner. This is accomplished by decoupling the navigation problem into independent collision avoidance and guidance problems and by providing mechanisms for solving both efficiently. While the framework does forgo global optimality in order to compute solutions, we argue that such optimality is unattainable in practice due to intractability, so nothing is actually sacrificed. An example solution is demonstrated for a novel graph traversal problem using a deterministic, single-agent velocity profile planner in a partially observable, multi-agent setting.

Keywords: Agents-based systems, Autonomous systems, Cooperative control
Abstract: To improve the flocking convergence speed for completing some emergent tasks, e.g. search and rescue in an accident, this paper proposes a novel fast convergent flocking control method for multi-agent systems with switching communication topology. First, a new distributed per-step convergence factor (D-PCF) is defined to indicate flocking convergence speed of each agent, whose summation indicates the overall convergence speed of multi-agent systems. Then, to decrease the D-PCF and thus increase the flocking convergence speed of the overall multi-agent system, a new distributed fast synchronization (DFS) algorithm is developed to determine position-dependent and velocity-dependent adjacency weights on every communication channel of an ad hoc network with switching topologies. The simulation results show that compared with a conventional flocking control, the flocking control using the DFS algorithm with switching communication topology can significantly increase the flocking convergence speed without requiring larger control inputs.

Keywords: Agents-based systems, Cooperative control, Distributed control
Abstract: This paper presents a novel approach for handling packet loss in first-order consensus protocols based on a Taylor series expansion. We propose a dynamic memory approach which depends on a locally measured loss rate at each agent. The quasi-Taylor method assumes that each agent is storing not only the past received value of its neighbours in a memory, but the last nu received states of all neighbours in order to predict the future trajectory with the quasi-Taylor estimation. In addition, we use the so-called importance measure to label the most important information received at each time step. Then, depending on the measured loss rate the past data points of a neighbour are used to predict the future trajectory. Hereby, the trajectory is determined as a combination of different orders of the quasi-Taylor estimation. In order to minimise the distance of the consensus value to the actual average, we propose to use an adaptive step size for predicting the future trajectory of the neighbours. We give an upper bound on the convergence rate when uniform packet loss is assumed and show that the proposed approach outperforms existing methods from the literature with the help of simulation studies.

Keywords: Agents-based systems, Cooperative control, LMIs
Abstract: This paper studies formation control in second-order multi-agent systems where communication between the neighbouring agents is based on fulfillment of dynamic event-triggering (DET) conditions. A novel co-design optimization is proposed to simultaneously design all required control and event-triggering parameters. We use the flexibility of the proposed co-design method to optimize the inter-event interval for a predefined formation convergence rate. The optimization is based on the scalarization (weighted sum) approach which provides a structured trade-off between the formation convergence rate and intensity of the event-triggerings. The computational complexity of the proposed optimization is independent of the network size. Simulations for multi-agent systems consisting a class of unmanned aerial vehicles (UAV) quantify the effectiveness of the proposed approach.

Keywords: Agents-based systems, Cooperative control, Sensor networks
Abstract: In this paper, we consider a synchronization problem for wireless sensor networks using pulse-based communication in the presence of faulty or malicious nodes. Such nodes have the effect of disturbing the normal nodes. We propose a resilient distributed algorithm for synchronization among the normal nodes by assuming worst-case attacks. The algorithm functions even under a sparse network structure, which is important for scalability of this type of application. The approach is based on mean subsequence reduced (MSR) type algorithms from the area of multi-agent consensus. We characterize a detection method for finding malicious nodes that transmit pulses irregularly. Then, we show that as long as the malicious ones send pulses in a way they can remain undetected, the normal nodes can reach synchronization by ignoring suspicious pulses. To illustrate the results, numerical examples are provided.

Keywords: Estimation, Adaptive systems, Kalman filtering
Abstract: In many applications of state estimation, it is efficient to confine the output-error injection to a prescribed subspace of the state space. This paper considers this problem by applying the unscented Kalman filter and retrospective cost state estimator (RCSE) to linear and nonlinear systems with subspace-constrained state correction. As an application of these techniques, parameter estimation is considered for linear and nonlinear systems with unknown parameters, where the output-error injection is confined to the subspace corresponding to the states representing the unknown parameters.

Keywords: Estimation, Chemical process control, Sensor fusion
Abstract: In the last few decades, various spectroscopic soft sensors that predict sample properties from its spectroscopic readings have been reported. To improve prediction performance, variable selection that aims to eliminate irrelevant wavelengths is often performed prior to soft sensor model building. However, due to the data-driven nature of many variable selection methods, they can be sensitive to the choice of the training data, and oftentimes the selected wavelengths show little connection to the underlying chemical bonds or function groups that determine the property of the sample. To address these limitations, we proposed a new variable selection method, namely consistency enhanced evolution for variable selection (CEEVS), which focuses on identifying the variables that are consistently selected from different training dataset. To demonstrate the effectiveness and robustness of CEEVS, we compared it with three representative variable selection methods using two published NIR datasets. We show that by identifying variables with high selection consistency, CEEVS not only achieves improved soft sensor performance, but also identifies key chemical information from spectroscopic data.

Keywords: Estimation, Closed-loop identification
Abstract: It is shown how the kernel approach to joint parameter and state estimation can be improved to handle large measurement noise. High accuracy of estimation results from combining the powers of the kernel representation of the differential invariance in the system, a feasible recursive version of the generalized least squares with covariance weighting to eliminate regression dilution and suitable choices of instrumental variables to compensate for the error-in-the variable in the stochastic regression formulation.

Keywords: Estimation, Decentralized control, Machine learning
Abstract: This paper addresses distributed parameter estimation in randomized one-hidden-layer neural networks. A group of agents sequentially receive measurements of an unknown parameter that is only partially observable to them. In this paper, we present a fully distributed estimation algorithm where agents exchange local estimates with their neighbors to collectively identify the true value of the parameter. We prove that this distributed update provides an asymptotically unbiased estimator of the unknown parameter, i.e., the first moment of the expected global error converges to zero asymptotically. We further analyze the efficiency of the proposed estimation scheme by establishing an asymptotic upper bound on the variance of the global error. Applying our method to a real-world dataset related to appliances energy prediction, we observe that our empirical findings verify the theoretical results.

Keywords: Estimation, Differential-algebraic systems, Process Control
Abstract: The Kalman filter and its variants have been developed for state estimation in semi-explicit, index 1 DAE systems in current literature. In this work, we develop a method for state estimation in non--linear fully implicit, index 1 differential algebraic equation (DAEs) systems. In order to extend the Kalman filtering techniques for the fully-implicit index-1 DAE systems, in the correction step we convert the fully--implicit DAE into a system of ordinary differential equations (ODEs). This is achieved by the index reduction of DAE using the method of successive differentiation of algebraic equations. This is a challenging problem as the fully implicit DAE system does not contain explicit algebraic states unlike in the semi--explicit case. In this work, we propose a linear transformation of the mass matrix which enables us to find candidate algebraic states for the system. This transformation on the mass matrix is relatively simpler than constructing the transformation matrices in the Weierstrass-Kronecker canonical form. We illustrate our proposed method with two examples, a linear and a nonlinear fully implicit index 1 DAE system.

Keywords: Estimation, Emerging control applications, Fault detection
Abstract: This paper addresses the problem of secure state estimation in the presence of attacks on sensor measurements of a linear time invariant (LTI) systems. We assume that the system is equipped with a common l0-based attack-resilient state estimator and a sound anomaly detector. We introduce the notion of perfect attackability (PA) for LTI systems with bounded noise, when the attacker may introduce an unbounded estimation error while remaining undetected by the anomaly detector. Finally, necessary and sufficient conditions for perfectly attackable systems are provided, and illustrated on examples.

Keywords: H-infinity control, Robust control, Linear systems
Abstract: In this paper, we develop a covariance control problem to address a tradeoff between H_2 performance and H_oo disturbance attenuation. In particular, we formulate a mixed-norm H_2/H_oo and entropy covariance control problem that guarantees that the state covariance of an uncertain dynamical system driven by white noise is upper bounded in the sense of the cone of nonnegative definite matrices by a given threshold matrix via state feedback control. This is accomplished by combining covariance control theory and mixed norm H_2/H_oo control theory. By using suitable transformations involving dynamic weighting on the complimentary sensitivity system transfer function, the formulation is applicable to robustness problems. The proposed formulation allows for solutions via semidefinite programming. Finally, an illustrative numerical example is provided to show the efficacy of the proposed approach.

Keywords: Networked control systems, Agents-based systems, Robust control
Abstract: A distributed event-triggered controller is developed for formation control and leader tracking (FCLT) with robustness to adversarial Byzantine agents for a class of heterogeneous multi-agent systems (MASs). A reputation-based strategy is developed for each agent to detect Byzantine agent behaviors within their neighbor set and then selectively disregard Byzantine state information. Selectively ignoring Byzantine agents results in time-varying discontinuous changes to the network topology. Nonsmooth dynamics also result from the use of the event-triggered strategy enabling intermittent communication. Nonsmooth Lyapunov methods are used to prove stability and FCLT of the MAS consisting of the remaining cooperative agents.

Keywords: Optimal control, Robust control, LMIs
Abstract: The Conic Sector Theorem can be employed for controller synthesis to ensure input-output stability. This work develops a synthesis method for conic, observer based controllers by minimizing an upper-bound on the closed-loop H2-norm. The proposed method can be seen as the dual of an existing optimal synthesis method, but with an alternative initialization to expand the set of plants for which it is feasible. Moreover, this results in better performance in some examples and therefore provides a useful alternative tool for robust and optimal control.

Keywords: Optimization algorithms, Robust adaptive control, Lyapunov methods
Abstract: We study the uniform asymptotic stability properties of the heavy-ball optimization dynamics with general time-varying damping. Unlike existing results in the literature, which have focused mainly on standard convergence results, we study a stronger limiting notion called uniform asymptotic stability, which is instrumental for the design of feedback-based algorithms. Given that recent results in the literature have shown that a class of heavy-ball optimization dynamics with vanishing damping fails to satisfy this limiting notion, we study sufficient and necessary conditions on the time-varying coefficients such that uniform asymptotic stability for the set of minimizers of the cost function is achieved. Our main results show that such conditions are related to the notion of persistence of excitation, which is commonly used in adaptive control and system identification. Finally, we show that the persistence of excitation condition is not necessary for a class of high-resolution accelerated optimization dynamics with Hessian-driven damping. Our results are established by using a nested Matrosov's theorem that has not been used before in the analysis of accelerated optimization algorithms.

Keywords: Robust control, Distributed control, Optimal control
Abstract: We generalize the system level synthesis framework to systems defined by bounded causal linear operators, and use this parameterization to make connections between robust system level synthesis and the robust control literature. In particular, by leveraging results from L1 robust control, we show that necessary and sufficient conditions for robust performance with respect to causal bounded linear uncertainty in the system dynamics can be translated into convex constraints on the system responses. We exploit this connection to show that these conditions naturally allow for the incorporation of delay, sparsity, and locality constraints on the system responses and resulting controller implementation, allowing these methods to be applied to large-scale distributed control problems -- to the best of our knowledge, these are the first such robust performance guarantees for distributed control systems.

Keywords: Optimal control, Machine learning, Computational methods
Abstract: Computing optimal feedback controls for nonlinear systems generally requires solving Hamilton-Jacobi-Bellman (HJB) equations, which, in high dimensions, are notoriously difficult. Existing strategies often rely on specific, restrictive problem structures, or are valid only locally around some nominal trajectory. In this paper, we propose a data-driven method to approximate semi-global solutions to HJB equations for general high-dimensional nonlinear systems and compute optimal feedback controls in real-time. To accomplish this, we model solutions to HJB equations with neural networks (NNs) trained on data generated without discretizing the state space. Training is made more effective and data-efficient by leveraging the known problem structure and using the partially-trained NN to aid in further data generation. We demonstrate the effectiveness of our method by learning solutions to HJB equations for nonlinear systems of dimension up to 30 arising from the stabilization of a Burgers'-type partial differential equation. The trained NNs are then used for real-time optimal feedback control of these systems.

Keywords: Optimal control, Distributed parameter systems
Abstract: By exploiting min-plus linearity, semiconcavity, and semigroup properties of dynamic programming, a fundamental solution semigroup for a class of approximate finite horizon linear infinite dimensional optimal control problems is constructed. Elements of this fundamental solution semigroup are parameterized by the time horizon, and can be used to approximate the solution of the corresponding finite horizon optimal control problem for any terminal cost. They can also be composed to compute approximations on longer horizons. The value function approximation provided takes the form of a min-plus convolution of a kernel with the terminal cost. A general construction for this kernel is provided, along with a spectral representation for a restricted class of sub-problems.

Keywords: Distributed parameter systems, Robust control, Uncertain systems
Abstract: In this paper, we address a two output tracking problem under the guidance of the internal model principle for an Euler-Bernoulli beam equation, which is motivated from a recent paper [SIAM J. Control Optim., 57(2019), 1890-1928] where an adaptive control approach was used to estimate in real time all coefficients of the external harmonic disturbance. Compared with the aforementioned paper, we improve the results significantly from several aspects: a)~ both position and velocity are regulated simultaneously instead of a single position; b)~ the convergence is uniformly exponentially instead of asymptotically; c)~ the control contains essentially the internal model and is therefore conditionally robust; d)~ a man-made condition has been removed with a necessary assumption only. Our approach is systematic and straightforward in dealing with other PDEs with the same kind.

Keywords: Computational methods, Distributed parameter systems
Abstract: In this paper we present an optimal error estimate for hp-C0 finite element approximations of distributed parameter LQR control problems. We show that if the control weighting operators are sufficiently smooth, then the solutions of the corresponding infinite dimensional Operator Riccati Equations that define optimal feedback gain operators retain the same order of smoothness. In this case, one can extend convergence results previously obtained for parabolic control problems to a general class of distributed parameter LQR control problems. Convergence and optimal error estimates are provided for a general class of control systems when the system operator generates a compact analytic semigroup In addition, these new results eliminates the ``log'' term appearing in previous papers and applies to parabolic PDE systems and hyperbolic problems with sufficient damping. The basic theory is presented in an abstract space framework and then applied to systems modeled by parabolic and damped hyperbolic PDEs. Examples are provided to illustrate the theory.

Keywords: Distributed parameter systems, Linear systems
Abstract: We study boundary feedback stabilization of a pair of coupled reaction-diffusion equations with distinct diffusivities, where attractivity to the origin is required to occur in a finite time which is prescribed independently of initial conditions. Our approach is twofold: we first develop control laws which render the system to a cascade form; we then utilize two different time-varying control laws which ensure prescribed-time stabilization of each equation. To achieve the desired stabilization while ensuring that the resulting boundary feedback controllers remain bounded, it is necessary to prescribe two different rates of attractivity to the origin. We achieve this by using distinct time-varying gains with either square or cubic “blow-up” functions, which diverge at the prescribed terminal time. The ensuing control laws achieve stabilization of the plant in a prescribed terminal time.

Keywords: Distributed parameter systems, Computational methods, Stability of nonlinear systems
Abstract: Feedback control problems involving autonomous quadratic systems are prevalent, yet very few software tools are available for approximating their solution. This paper represents a step forward in the special case where both the state equation and the control costs are quadratic. As it represents the natural extension of the linear-quadratic regulator (LQR) problem, we describe this setting as the quadratic-quadratic regulator (QQR) problem. We describe an algorithm that exploits the structure of the QQR problem that arises when implementing Al'Brekht's method. As we show, this well-known algorithm has an elegant formulation when written using Kronecker products and produces linear systems with a special structure that can take advantage of modern tensor-based linear solvers. We demonstrate this formulation on a random test problem then apply it to a discretized distributed parameter control problem that fits the QQR framework. This approach is amenable to low degree polynomial feedback laws in systems with modest model dimensions, for example, systems produced by modern model reduction methods. Comparisons to linear feedback control laws show a benefit in using the QQR formulation.

Keywords: Algebraic/geometric methods, Lyapunov methods, Stability of nonlinear systems
Abstract: This paper presents a finite-time stable (FTS) attitude tracking control scheme in discrete time for an unmanned vehicle. The attitude tracking control scheme guarantees discrete-time stability of the feedback system in finite time. This scheme is developed in discrete time as it is more convenient for onboard computer implementation and guarantees stability irrespective of sampling period. Finite-time stability analysis of the discrete-time tracking control is carried out using discrete Lyapunov analysis. This tracking control scheme ensures stable convergence of attitude tracking errors to the desired trajectory in finite time. The advantages of finite-time stabilization in discrete time over finite-time stabilization of a sampled continuous time tracking control system is addressed in this paper through a numerical comparison. This comparison is performed using numerical simulations on continuous and discrete FTS tracking control schemes applied to an unmanned vehicle model.

Keywords: Algebraic/geometric methods, Optimization, Stability of nonlinear systems
Abstract: Consensus on nonlinear spaces is of use in many control applications. This paper proposes a gradient descent flow algorithm for consensus on hypersurfaces. We show that if an inequality holds, then the system converges for almost all initial conditions and all connected graphs. The inequality involves the hypersurface Gauss map and the gradient and Hessian of the implicit equation. Moreover, for the inequality to hold, it is necessary that the manifold is the boundary of a convex set. The literature already contains an algorithm for consensus on hypersurfaces. That algorithm on any ellipsoid is equivalent to our algorithm on the unit sphere. In particular, that algorithm achieves almost global synchronization on ellipsoids. These findings suggest that strong convergence results for consensus seeking gradient descent flows may be established on manifolds that are the boundaries of convex sets.

Keywords: Algebraic/geometric methods, Stability of nonlinear systems, Numerical algorithms
Abstract: An approach for the introduction of transverse coordinates in a vicinity of a periodic trajectory is presented. The approach allows finding by numerical integration periodic normalized mutually-orthogonal vector-functions that form a continuously differentiable basis on moving Poincar'e sections for a given periodic solution of a nonlinear dynamical system.

The found moving frame is used to define new local (transverse) coordinates for an associated affine nonlinear control system in a neighborhood of the trajectory, and to proceed with orbital stability analysis and/or synthesis of a stabilizing feedback control law.

As a demonstrating example of the approach, the problem of orbital stabilization of a trajectory of a multibody car system is considered. The results of computer simulations of the system are presented.

Keywords: Biologically-inspired methods, Robust control, Stability of nonlinear systems
Abstract: In recent years diverse computational models of emotional learning observed in the mammalian brain have inspired a number of self-learning control approaches. These architectures are promising in terms of their learning ability and low computational cost. However, the lack of rigorous stability analysis and mathematical proofs of stability and performance has limited the proliferation of these controllers. To address this drawback, this paper proposes a modified brain emotional neural network structure using a radial basis function inside the Thalamus and an emotional signal based on an integral action structure to increase performance. Mathematical stability proofs are provided, together with numerical simulations, demonstrating the superior performance obtained with the new modifications proposed to the emoional learning-inspired control.

Keywords: Chaotic systems, Robust control, Stability of nonlinear systems
Abstract: The Chameleon hidden chaotic system is a chaotic system with exciting and particular properties. One of its specific features is that by changing its constant parameters, the flow exhibits three classes of hidden attractors containing one stable equilibrium, line of equilibria, or no equilibria, and self-excited attractors. In this paper, a new Chameleon hidden hyperchaotic flow is proposed and the corresponding hidden attractors and self-excited attractor are evaluated. Then, a new super-twisting fast terminal adaptive sliding mode control technique is suggested for finite-time stability of Chameleon hidden hyperchaotic flows. Dynamics of this chaotic system with different values of constant parameters has been studied using phase portraits, stability analysis and bifurcation diagrams. Descriptive simulations on the Chameleon hidden hyperchaotic system with external disturbances and parametric uncertainties are presented to approve the effectiveness of the proposed method.

Keywords: Constrained control, Lyapunov methods, Stability of nonlinear systems
Abstract: This paper presents a novel design method for barrier avoidance control (or state constrained control) in a state space. This is achieved through a diffeomorphic transformation, which projects the constrained region in the original space into a radially large region in the new space. By this transformation, the state constrained control problem is converted into a non-constrained problem, and thus significantly increases the flexibility of the control design options compared with existing state constrained control methods. A case study on sliding mode control under barrier avoidance scheme is given to demonstrate the effectiveness of the proposed method.

Keywords: Cooperative control, Aerospace, Optimal control
Abstract: We study a multi-body asset-guarding game in missile defense where teams of interceptor missiles collaborate to defend a non-manuevering asset against a group of threat missiles. We approach the problem in two steps. We first formulate an assignment problem where we optimally assign subsets of collaborating interceptors to each threat so that all threats are intercepted as far away from the asset as possible. We assume that each interceptor is controlled by a collaborative guidance law derived from linear quadratic dynamic games. Our results include a 6-DOF simulation of a 5-interceptor versus 3-threat missile engagement where each agent is modeled as a missile airframe controlled by an autopilot. Despite the assumption of linear dynamics in our collaborative guidance law and the unmodeled dynamics in the simulation environment (e.g., varying density and gravity), we show that the simulated trajectories match well with those predicted by our approach. Furthermore, we show that a more agile threat, with greater speed and acceleration, can be intercepted by inferior interceptors when they collaborate. We believe the concepts introduced in this paper may be applied in asymmetric missile defense scenarios, including defense against advanced cruise missiles and hypersonic vehicles.

Keywords: Cooperative control, Agents-based systems, Distributed control
Abstract: In this paper, we consider scalable output and regulated output synchronization problems for heterogeneous networks of right-invertible linear agents based on localized information exchange where in the case of regulated output synchronization, the reference trajectory is generated by a so-called exosystem. We assume that all the agents are introspective, meaning that they have access to their own local measurements. We propose a scale-free linear protocol for each agent to achieve output and regulated output synchronizations. These protocols are designed solely based on agent models and they need no information about communication graph and the number of agents or other agent models information.

Keywords: Cooperative control, Autonomous robots, Aerospace
Abstract: We present a formation control algorithm for double-integrator agents, where the formation is time varying and the agents' controls satisfy a priori bounds (e.g., the controls accommodate actuator saturation). We assume that each agent has relative-position-and-velocity feedback of its neighbor agents, where the communication structure is a strongly connected graph, and at least one agent has a measurement of its position and velocity relative to the leader (if applicable). The main analytic results provide sufficient conditions such that all agents have a priori bounded controls and converge to the desired time-varying relative positions with one another and the leader. The results are global for an undirected communication structure, and local for the directed case. We also demonstrate the algorithm in rotorcraft experiments.

Keywords: Cooperative control, Networked control systems, Decentralized control
Abstract: In this paper, finite-time attitude consensus control laws for multi-agent rigid body systems are presented using rotation matrices. The control objective is to stabilize the relative configurations in a finite convergence time. First, the control design is done on the kinematic level where the angular velocities are the control signals. Next, the design is conducted on the dynamic level in the framework of the tangent bundle TSO(3) associated with SO(3), where the torques implement the feedback control of relative attitudes and angular velocities. The Lyapunov-based almost global finite-time stability of the consensus subspace is demonstrated for both cases. Finally, numerical simulations are provided to verify the effectiveness of the proposed consensus control algorithms.