Keywords: Energy systems
Abstract: Electrification of mobility and transport is a global megatrend that has been underway for decades. The mobility sector encompasses cars, trucks, busses and aircraft. These systems exhibit complex interactions of multiple modes of power flow. These modes can be thermal, fluid, electrical, or mechanical. A key challenge in working across various modes of power flow is the widely varying time scales of the subsystems which makes centralized control efforts challenging. This talk will present a particular distributed controller architecture for managing the flow of power based on on-line optimization. A hierarchical approach allows for systems operating on different time scales to be coordinated in a controllable manner. It also allows for different dynamic decision making tools to be used at different levels of the hierarchy based on the needs of the physical systems under control. Additional advantages include the modularity and scalability inherent in the hierarchy. Additional modules can be added or removed without changing the basic approach.

In addition to the hierarchical control, a particularly useful graph-based approach will be introduced for the purpose of modeling the system interactions and performing early stage design optimization. The graph approach, like the hierarchy, has benefits of modularity and scalability along with being an efficient framework for representing systems of different time scales. The graph allows design optimization tools to be implemented and optimize the physical system design for the purpose of control. Recent results will be presented representing both generic interconnected complex systems as well as specific examples from the aerospace and automotive application domains.

Keywords: Control applications, Optimal control, Mechanical systems/robotics
Abstract: As wind turbines in a wind farm are usually placed close together for economic reasons, wake effects play a big role in the power production of wind farms. Waked, downstream turbines can produce up to 70% less power than their upstream counterpart. Using Individual Pitch Control (IPC) in a clever way, it is possible to induce wake mixing behind a turbine to decrease the wake deficit that downstream turbines experience. By applying the Multi-Blade Coordinate (MBC) transformation, the direction of the wake can be manipulated dynamically. This approach is unique in the sense that it has a limited effect on the performance of the upstream turbine, while significantly improving the power capture of downstream machines. By applying a low-frequent sine signal on the tilt and yaw angles from the MBC transformation, the direction of the turbine thrust force can be manipulated. This results in a wake with a helical shape. High-fidelity Computational Fluid Dynamics (CFD) simulations in SOWFA show that IPC can increase the power production of a 2-turbine wind farm by up to 7.5%.

Keywords: Distributed control, Learning, Linear systems
Abstract: We study model-free learning methods for the output-feedback Linear Quadratic (LQ) control problem in finite-horizon subject to subspace constraints on the control policy. Subspace constraints naturally arise in the field of distributed control and present a significant challenge in the sense that standard model-based optimization and learning leads to intractable numerical programs in general. Building upon recent results in zeroth-order optimization, we establish model-free sample-complexity bounds for the class of distributed LQ problems where a local gradient dominance constant exists on any sublevel set of the cost function. We prove that a fundamental class of distributed control problems - commonly referred to as Quadratically Invariant (QI) problems - as well as others possess this property. To the best of our knowledge, our result is the first sample-complexity bound guarantee on learning globally optimal distributed output-feedback control policies.

Keywords: Networked control systems, Sensor networks, Linear systems
Abstract: The security problem in control systems essentially involves attacks manipulating the sensor measurements and/or actuator signals so as to cause varying degrees of damages, as has been evidenced in certain real-world events. The underlying principle behind such attacks involves corrupting the output signal so that it corresponds to those generated as a result of natural disturbance. Against this backdrop, the notion of security index quantifies the least effort involved in conducting such attacks. More precisely, the security index of a component, i, equals the minimum number of sensors and actuators that needs to be compromised so as to conduct a perfectly undetectable attack using component i. Thus, the larger the index the greater the effort, and vice-versa. The security index is typically defined with respect to the given plant model, i.e., fixed numerical entries in corresponding system matrices. In this work, in a broad sense, we are interested in notion(s) of security index that are not reliant on the specific numerical entries in the system matrices. Towards this end, we introduce a novel security index, called strong structural security index. This is defined as the security index of a component for all non-zero realizations. Subsequently, we provide graph-theoretic conditions, in terms of maximum size of uniquely restricted matching on suitably-defined bipartite graphs (resp. bipartite subgraphs), for computing lower (resp. upper) bounds on the strong structural security index of a component. Moreover, we also show that computing the said bounds is NP-hard.

Keywords: Adaptive control, Optimization, Optimization algorithms
Abstract: Conventional perturbation-based extremum seeking control (ESC) employ an exogenous time-dependent periodic signal to find an optimum of an unknown plant. In the case of periodic plants, to ensure stability of the overall system, the time-dependent signal must be slower than the plant such that there is sufficient time-scale separation between the plant and the ESC dynamics. This approach is suitable when the plant operates at a fixed time-scale. If the plant slows down during operation, the time-scale separation can be violated. As a result, the stability and performance of the overall system can no longer be guaranteed. Recently, we proposed an ESC for periodic systems, in which the external time-dependent dither signal in conventional ESC is replaced with a function of the periodic states of the plant, thereby making ESC time-invariant in nature. The advantage of using a state-based dither is that it inherently contains the information about the rate of the rhythmic task under control. Thus, in addition to maintaining time-scale separation at different plant speeds, the adaptation speed of a time-invariant ESC automatically changes, without changing the dither frequency.

Common approaches to prove stability of conventional ESC use either time-scale separation arguments coupled with averaging and singular perturbation methods or Lyapunov-based arguments by analyzing the stability of the origin by a change of variables. However, due to the autonomous nature of the overall system in our ESC scheme, both of these techniques cannot be used to prove its stability. In this poster, we will talk about a general method that can be used for proving the stability of time-invariant ESC in a rigorous manner.

Keywords: Robotics, Hybrid systems, Optimization
Abstract: Most current bipedal robots were modeled with an assumption that there is no slip between the stance foot and ground. This work relaxes the assumption and undertakes a comprehensive study of a two-link bipedal robot with foot slipping. It adopts minimal coordinates to construct the hybrid model dynamics and shows by simulation that the feasible gaits fail on slippery ground for two causes: falling backward or requiring negative contact force which cannot be provided by the ground. Further, the gaits falling backward have small cost of transport (CoT). To study the relationship between robustness (in the sense of preventing slipping/falling) and walking speed, step length, all selected gaits are optimized in terms of CoT. It is found that small walking speed can help prevent slipping/falling, especially for the gaits with large step length. In contrast, there is generally an optimal value in determining the step length that can best prevent slipping/falling for a specified walking speed. Moreover, the gaits with small step length and moderate speed (choppy strides) are robust and thus more preferable on slippery ground.

Keywords: Human-in-the-loop control, Markov processes, Computational methods
Abstract: In a Shared Mobility on Demand Service (SMoDS), dynamic pricing plays an important role in the form of an incentive for empowered passengers to decide on the ride offer. Strategies for determining the dynamic tariffs should be suitably designed so that the incurred demand and supply within the SMoDS platform are balanced and therefore economic efficiency is further achieved. In this manuscript, we formulate a discrete time Markov Decision Process (MDP) to determine the probability of acceptance of each empowered passenger that is desired by the SMoDS platform. The proposed MDP formulation is a versatile framework which is shown to explicitly accommodate passenger behavior and realize the desired system objective. Estimated Waiting Time (EWT) is utilized as a suitable metric to measure the balance between demand and supply, with the goal of regulating EWT around a target value. We propose the use of a Dynamic Programming algorithm to derive the optimal policy that achieves the regulation. Computational experiments are conducted to demonstrate effective regulation of EWT, through various scenarios.

Keywords: Information technology systems, Game theory, Stochastic systems
Abstract: Continuous authentication, as an extension to conventional authentication methods, is emerging as a promising technology for preventing and mitigating sophisticated identity theft and session hijacking attacks, which may cause catastrophic losses for individuals and companies and possibly disasters for critical infrastructures. In this study, adversarial attacks on continuous authentication security are considered, and the interaction between the attacker and an information technology (IT) manager (defender) is modeled as a dynamic stochastic game. In the considered model, the attacker, if it chooses to compromise the system (which has a cost for the attacker), observes and learns the traffic generated by the user, imitates legitimate user behavior and executes a rogue command on the resource. Meanwhile, the defender designs the systems parameters and security measures in order to detect suspicious behavior and to prevent unauthorized access while minimizing the monitoring expenses and the degradation of the user experience. Following common practice in game theoretic models of security, the attacker is assumed to be aware of the strategy of the defender whereas the defender is not aware of the attacker’s strategy. It is shown that the optimal attacker strategy exhibits a threshold structure, and consists of observing the user behavior to collect information at the beginning, and then attacking (rather than observing) after gathering enough data. From the defender's side, the optimal design of the security measures is provided. Numerical results are used to illustrate the intrinsic trade-off between monitoring cost and security risk. Even though continuous authentication can be effective in minimizing security risk, the results suggest that continuous authentication only is not enough to secure the system; the additional solutions, such as IDS, are essential for optimal security risk minimization.

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: 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: 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: Control applications, Lyapunov methods
Abstract: One of the pressing concerns for next-generation manufacturing is the development of techniques for guaranteeing that a control system is cyberattack-resilient in the sense that even if a cyberattack is successful at breaking information technology-based defenses (e.g., it succeeds at providing a false sensor measurement to the controller), closed-loop stability is still maintained. Our prior work has provided a nonlinear systems definition for cyberattacks. This work explores how a nonlinear systems perspective on cyberattack-resilience for false sensor measurements provided to controllers may allow an economic model predictive control (EMPC) formulation known as Lyapunov-based EMPC (LEMPC) to be designed such that if a cyberattack occurs at a sampling time, the closed-loop state will not leave a region where a known feedback control law exists that can stabilize the origin of the closed-loop system if the cyberattack is detected and non-falsified state measurements are then provided within that sampling period.

Keywords: Building and facility automation, Optimal control, Simulation
Abstract: In this paper, a model predictive control (MPC) framework for building energy system setpoint optimization is developed and tested. The performance of the MPC framework is presented in comparison to a baseline case, where a fixed setpoint schedule is used. To simulate the MPC procedure, an EnergyPlus building model is used to represent the actual building that the optimal setpoints are applied to, and a Gaussian process (GP) regression meta-model is used in the MPC procedure that generates the optimal setpoints. The performance outputs that are used for evaluation are total heating, ventilation and air conditioning (HVAC) energy usage and the Fanger predicted mean vote (PMV) thermal comfort measure. The inputs for the GP regression meta-models are selected to be representative of data points that could be obtained by modern supervisory control and data acquisition (SCADA) systems to support data-driven building models. The supervisory MPC framework is capable of reducing the total energy usage with minor adjustments in thermal comfort.

Keywords: Chemical process control, Robust control, Process Control
Abstract: In this article, a novel distributionally robust optimization (DRO) based multistage model predictive control (MPC) framework is proposed to hedge against the uncertainty in control problems. Without a priori knowledge of the exact uncertainty distribution, an ambiguity set, constructed based on principal component analysis, incorporates the first-order moment information instead. A linear performance measure is chosen so that the worst-case expected problem can be exactly dualized by adopting the affine decision rule. By considering input and state constraints robustly with respect to a support set, which specifies the domain of the uncertainty, the DRO-based MPC model is developed as a robust optimization problem. The proposed framework is illustrated on a two-mass-spring system and results show the improved control performance compared to traditional control strategies.

Keywords: Networked control systems, Communication networks, Predictive control for nonlinear systems
Abstract: Rollout approaches are an effective tool to address the problem of co-designing the transmission schedule and the corresponding input values, when the controller is connected to the plant via a resource-constrained communication network. These approaches typically employ an MPC, activated at multiples of the period length of a base transmission schedule. Using multi-step invariant terminal regions and a prediction horizon longer than the base period, stability can be ensured. This strategy, however, suffers from intrinsic shortcomings, such as a high computational complexity and low robustness. We aim to resolve these drawbacks by proposing an MPC with periodically time-varying terminal region and cost for the rollout setup, where the controller is activated at each time instant and features an arbitrary but fixed prediction horizon. We consider in more detail two specific setups from the literature on Networked Control Systems, namely the token bucket network and actuator scheduling. For both setups, we provide conditions for which convergence under application of the time-varying MPC can be guaranteed. In a numerical example, we demonstrate the benefits of the proposed method.

Keywords: Nonlinear output feedback, Predictive control for nonlinear systems, Stability of nonlinear systems
Abstract: This paper presents an output feedback model predictive control for a class of nonlinear systems in a multi-rate scheme, where the control sampling period is larger than the estimation sampling period. With a small sampling period, the observer is designed to be faster than the dynamics of the closed-loop system under state feedback. Stabilization is achieved by a separation approach in which the control is designed first using state feedback and practical stabilization is achieved by output feedback.

Keywords: Predictive control for linear systems, Constrained control, LMIs
Abstract: The purpose of this paper is to simplify the treatment of major issues of Model Predictive Control (MPC) like stability, feasibility and computational burden. To this purpose a new MPC policy is developed using a two Degrees of Freedom (2DoF) control scheme where the output r(k) of the feedforward Input Estimator (IE) is used as input forcing the closed-loop system Sigma_f . f is the feedback connection of an LTI plant p with an LTI feedback controller g. The task of g is to guarantee the stability of Sigma_f , as well as the fulfillment of hard constraints on some physical variables for any input r(k) satisfying an ”a priori” determined admissibility condition. The input r(k) is computed by the feedforward IE through the on-line minimization of a suitably definite finite-horizon quadratic cost functional and is applied to f according to the usual receding horizon strategy. To improve numerical efficiency, r(k) is assumed to be given by a B-spline function. This greatly decreases the number of decision variables of the on-line optimization procedure. It is shown that stability and recursive feasibility of the adopted MPC strategy are guaranteed in advance, regardless the chosen prediction horizon.

Keywords: Predictive control for linear systems, Model/Controller reduction, Manufacturing systems
Abstract: Model predictive control (MPC) has been used in a variety of industrial processes. Due to its large computation time application to fast and complex processes is limited. Therefore, reducing complexity has been focus of intense research. In this paper, a complexity reduction strategy for a linear MPC is developed. It is used for the control of an overactuated roll gap with two different actuator types in a cold rolling mill. One of the actuators is able to accept new set points only at a much slower sample time than the fast actuator. In order to coordinate the actuators, a moving manipulation point scheme is introduced where a single time-varying optimization variable of the slow actuator moves through the control horizon of the fast actuator. The fast actuator ensures reference tracking regarding the roll gap when the slow one is idle. The proposed scheme reduces the number of optimization variables as well as constraints and thus enables control of faster processes. Simulation results show the proof of concept and an enhancement in computation time. Moreover, a real-time implementation is demonstrated on a cold rolling mill.

Keywords: Predictive control for linear systems, Stochastic optimal control, Distributed control
Abstract: In this paper, we present a novel stochastic output-feedback MPC scheme for distributed systems with additive process and measurement noise. The chance constraints are treated with the concept of probabilistic reachable sets, which, under a unimodality assumption on the disturbance distributions are guaranteed to be satisfied in closed-loop. By conditioning the initial state of the optimization problem on feasibility, the fundamental property of recursive feasibility is ensured. Closed-loop chance constraint satisfaction, recursive feasibility and convergence to an asymptotic average cost bound are proven. An example of three interconnected subsystems is carried out.

Keywords: Process Control, Predictive control for linear systems
Abstract: We propose a stochastic model predictive control scheme for linear discrete-time and time-invariant systems with chance constraints and additive normal distributed disturbance with known covariance. The proposed approach allows to balance between performance optimization and robustness by adjusting the probability p of the chance constraints. The Controllability Gramian and the Riccati inequality are used to determine a set that contains all disturbances with a probability p. The proposed approach guarantees asymptotic stability of the nominal system, probabilistic convergence of the real system as well as constraint satisfaction and recursive feasibility in terms of probability. A proof of concept example considers the control of a wind turbine, where the wind acts as a normal distributed disturbance.

Keywords: Predictive control for linear systems, Robust control, Uncertain systems
Abstract: Classical tube based MPC utilizes a fixed linear feedback to counteract the effect of the disturbance in the prediction. This allows to bound the error between the predictions and the real system in form of sets; i.e. the real state evolution is bounded to sets called tubes. The choice of the corresponding feedback determines the size and shape of the tube and thus influences the domain of attraction and the control performance. Choosing the feedback often requires to compromise different objectives. We propose to compose the tube online based on multiple tubes, each of which is determined by a different linear gain to reduce the conservatism and to improve the controller performance. The underlying optimization problem remains convex and the computational demand increases only marginally. We discuss the resulting closed loop properties such as constraint satisfaction, robust recursive feasibility and stability. Moreover an adaption of the cost function is investigated, which allows to improve the closed loop performance. An example illustrates the proposed approach and its benefits.

Keywords: Predictive control for nonlinear systems, Automotive control, Control applications
Abstract: A Non-linear Model Predictive Control (NMPC) scheme is proposed for the optimal power-spit problem of a hybrid diesel-electric marine propulsion plant. The NMPC controller directly regulates the torque setpoints of the internal combustion engine and the electric motor/generator and ensures that certain constraints concerning the engine overloading are applied. In this way, fuel consumption and NOx emissions can be reduced. The modeling for the controller design was based on experimental data gathered from the hybrid plant and on first principles for the diesel engine behavior and battery charging. The controller was experimentally tested in real-time operation. Results showed that the controller rejected successfully load disturbances and maintained the rotational speed reference of the powertrain as well as the desirable state of charge in battery, in respect to the limits of the system.

Keywords: Predictive control for nonlinear systems, Optimal control, Identification
Abstract: In this paper a real-time capable hierarchical nonlinear model predictive control framework for the energy management of a hybrid electric vehicle is presented. As high-level energy management, nonlinear model predictive control is employed. Therefore, a nonlinear optimal control problem is formulated using a control-oriented internal model derived from high-fidelity models and experimental data. The multiple shooting algorithm and an Euler backward scheme are used to discretize the optimal control problem. The resulting nonlinear problem is solved in real-time using Sequential Quadratic Programming. A rule-based gear choice and engine on / off strategy is added. Actuator dynamics and drivability are adressed in a fast low-level linear time-variant model predictive ontroller. The result is analyzed and compared to the optimal solution obtained by dynamic programming using a simplified model of the vehicle.

Keywords: Predictive control for linear systems, Estimation, Chemical process control
Abstract: The paper is concerned with the multi-criteria and multivariate Control Performance Assessment (CPA) of Multiple-Input Multiple-Output Model Predictive Control. The task of control quality estimation in case of the multi-variable MPC poses serious challenges. The main difficulty is the fact that the engineer has to cope with multiple signals with no single measure that might be representative. Thus, there is a need for an approach which could compare and take into account various assessment aspects. Proposed method combines different integral, Gaussian and non-Gaussian measures and shows robustness to real-life disturbances.

Keywords: Predictive control for nonlinear systems, Constrained control, Chemical process control
Abstract: In this paper, a composite model predictive control (MPC) structure for the class of standard singularly perturbed systems is developed. To control and optimize economic performance of the slow dynamics, an economic MPC is developed and to stabilize the fast dynamics, a fast MPC is designed. Both MPC strategies do not rely on explicit stabilizing and/or terminal constraints. Working within a sampled-data setting, closed-loop stability in the sense of asymptotic stability of the economically optimal steady state is discussed. A chemical process example is given to demonstrate the control architecture.

Keywords: Predictive control for nonlinear systems, Process Control, Stability of nonlinear systems
Abstract: A method for integrating optimization and control during on-line process operation is known as economic model predictive control (EMPC). EMPC optimizes a general cost function which reflects process economics subject to a model of the process. One formulation of EMPC which can maintain closed-loop stability in the presence of sufficiently small disturbances is Lyapunov-based EMPC (LEMPC). In this work, we make precise connections between closed-loop stability considerations under LEMPC and numerical approximations (via Taylor series) of the solution of the nonlinear dynamic model of the process used in the controller. A chemical process example is utilized to demonstrate the concepts developed.

Keywords: Process Control, Uncertain systems, Chemical process control
Abstract: Appropriate greenhouse temperature should be maintained to ensure crop production while minimizing energy consumption. Even though weather forecasts could provide a certain amount of information to improve control performance, it is not perfect and forecast error may cause the temperature to deviate from the acceptable range. To inherent uncertainty in weather that affects control accuracy, this paper develops a data-driven robust model predictive control (DDRMPC) approach for greenhouse temperature control. The dynamic model is obtained from thermal resistance-capacitance modeling derived by the Building Resistance-Capacitance Modeling (BRCM) toolbox. Uncertainty sets of ambient temperature and solar radiation are captured by support vector clustering technique, and they are further tuned for better quality by training-calibration procedure. A case study shows that the DDRMPC has better control performance compared to rule-based control, certainty equivalent MPC, and robust MPC. DDRMPC approach ends up with 12% less total energy consumption than rule-based control strategy.

Keywords: Process Control, Optimization, Robust control
Abstract: Irrigation takes considerable amount of water; however, in many cases up to half of it is wasted. Improving the efficiency of irrigation control, therefore, is an important task for sustainable water management. Most existing irrigation control systems are based on soil moisture level. In this work, we explore theoretically the use of continuous values stem water potential (SWP) as a basis for control. SWP is a more direct measure of plant water status than the soil moisture level. After linearizing and discretizing a nonlinear dynamic model of water dynamics in a plant, we develop a model predictive control (MPC) framework for regulating SWP. To prevent plants from suffering water stress, data-driven robust MPC (DDRMPC) which captures the uncertainty of weather forecast error is implemented. A case study based on almond tree is presented to characterize the effectiveness of the DDRMPC strategy relative to on-off control. Sensitivity analysis on the prediction horizon and penalty weights were performed to investigate the varying irrigation control decisions. For the case characterized, the analysis shows that controlling tree trunk water potential through DDRMPC can save 2.5% amount of water comparing to on-off control while maintaining zero violation.

Keywords: Predictive control for nonlinear systems, Flexible structures, Distributed parameter systems
Abstract: Distributed parameter systems (DPS) are formulated as partial differential equations (PDE). Especially, under time-varying boundary conditions, PDE introduce force coupling. In the case of the flexible stacker crane (STC), nonlinear coupling is introduced. Accordingly, online trajectory planning and tracking can be addressed using a nonlinear model predictive control (NMPC). However, due to the high computational demands of a NMPC, this paper discusses a possibility of embedding nonlinearities inside a linear parameter varying (LPV) system and thus make a use of a numerically low-demanding linear MPC. The resulting mismatches are treated as parametric and additive uncertainties in the context of robust tube-based MPC (TMPC). For the proposed approach, most of the computations are carried out offline. Only a simple convex quadratic program (QP) is conducted online. Additionally a soft-constrained extension was briefly proposed. Simulation results are used to illustrate the good performance, closed-loop stability and recursive feasibility of the proposed approach despite uncertainties.

Keywords: Predictive control for nonlinear systems, Optimization algorithms, Identification for control
Abstract: We consider the problem of data-driven, on-the-fly, constrained control of unknown nonlinear dynamics from a single finite-horizon trajectory. We are motivated by applications with severe limits on data availability and require online controller synthesis. Using a one-step receding horizon control framework, we pose the data-driven optimal control problem as a contextual optimization problem. Specifically, we seek to iteratively minimize a smooth (black-box) objective function over iteration-dependent constraints. We propose C2Opt, a data-driven, convex-optimization-based approach to solve this problem. We utilize correct-by-construction function approximation bounds on the objective function to guide the decision process. Our bounds are synthesized from data and smoothness assumptions, and accommodates incorporation of side information like convexity, monotonicity, and loose bounds on the function. We then use tractable convex optimization problems to solve for the iterates, given the context and past data. We demonstrate C2Opt in successful navigation of a unicycle to a target destination, using only the data collected from a single trajectory under control constraints.

Keywords: Predictive control for nonlinear systems, Nonholonomic systems, Autonomous systems
Abstract: This paper introduces an accurate nonlinear model predictive control-based algorithm for trajectory following. The algorithm incorporates both the planned state and control trajectories into its cost functional - generally, following algorithms do not incorporate control trajectories into their cost functionals. Comparisons are made against two trajectory following algorithms, where the trajectory planning problem is to safely follow a person using an automated ATV with control delays in a dynamic environment while simultaneously optimizing speed and steering, minimizing control effort, and minimizing the time-to-person. Results indicate that the proposed algorithm reduces collisions, tracking error, orientation error, and time-to-goal. Therefore, including the control trajectories into the trajectory following algorithm helps the vehicle follow planned state trajectories more accurately, which ultimately improves safety, especially in dynamic environments.

Keywords: Predictive control for nonlinear systems, Lyapunov methods, Human-in-the-loop control
Abstract: Lyapunov-based economic model predictive control (LEMPC) is an optimization-based control design that computes economically-optimal control actions for a process while maintaining the closed-loop state within a bounded region of state-space; however, it may be difficult to design in practice without closed-loop simulations, as it requires an auxiliary stabilizing controller, Lyapunov function, and a number of sets to be developed to ensure closed-loop stability. Practical application of this method could benefit from methods which make it more likely that, without simulations to identify aspects of the control design that would provide stability, controller parameters can be selected that would maintain stability. In this work, we propose a method to seek to enhance tractability of LEMPC by providing initial suggestions for reducing the likelihood that textit{ad hoc} selection of a value for one of its parameters would be problematic for closed-loop stability.

Keywords: Predictive control for nonlinear systems, Autonomous systems, Learning
Abstract: A task decomposition method for iterative learning model predictive control is presented. We consider a constrained nonlinear dynamical system and assume the availability of state-input pair datasets which solve a task T1. Our objective is to find a feasible model predictive control policy for a second task, T2, using stored data from T1. Our approach applies to tasks T2 which are composed of subtasks contained in T1. In this paper we formally define subtasks and the task decomposition problem, and provide proofs of feasibility and iteration cost improvement over simple initializations. We demonstrate the effectiveness of the proposed method on autonomous racing and robotic manipulation experiments.

Keywords: Robotics, Networked control systems, Large-scale systems
Abstract: Multi-agent coverage control is used as a mechanism to influence the behavior of a group of robots by introducing time-varying domains. The coverage optimization problem is modified to adopt time-varying domains, and the proposed control law possesses an exponential convergence characteristic. Complex multi-agent control is simplified by specifying the desired distribution and behavior of the robot team as a whole. In the proposed approach, design of the inputs to the multi-agent system, i.e., time-varying density and time-varying domain, are agnostic to the size of the system. Analytic expressions of surface and line integrals present in the control law are obtained under uniform density. The scalability of the proposed control strategy is analyzed and verified via numerical simulation. Experiments on real robots are used to test the proposed control law.

Keywords: Autonomous robots, Linear systems, Autonomous systems
Abstract: Malicious attacks on modern autonomous cyber-physical systems (CPSs) can leverage information about the system dynamics and noise characteristics to hide while hijacking the system toward undesired states. Given attacks attempting to hide within the system noise profile to remain undetected, an attacker with the intent to hijack a system will alter sensor measurements, contradicting with what is expected by the system's model. To deal with this problem, in this paper we present a framework to detect non-randomness in sensor measurements on CPSs under the effect of sensor attacks. Specifically, we propose a run-time monitor that leverages two statistical tests, the Wilcoxon Signed-Rank test and Serial Independence Runs test to detect inconsistent patterns in the measurement data. For the proposed statistical tests we provide formal guarantees and bounds for attack detection. We validate our approach through simulations and experiments on an unmanned ground vehicle (UGV) under stealthy attacks and compare our framework with other anomaly detectors.

Keywords: Autonomous systems, Robotics
Abstract: The Rimon-Koditscheck potential is known to be a navigation function for sufficiently curved worlds. With global information, a diffeomorphism can be constructed to convert complex star worlds into sphere worlds so that the point agent can follow the negative gradient of the navigation function potential. Without global information, convergence guarantees only exist for sufficiently curved worlds or ellipsoidal worlds. This work closes the gap between local information and global information by extending convergence guarantees with local information to star worlds in general. Convergence to the goal and obstacle avoidance is established from every initial position in the free space. Results are numerically verified with a discretized version of the proposed dynamic flow.

Keywords: Control applications, Mechanical systems/robotics, Fluid power control
Abstract: The applications, modeling and control of continuum manipulators have been strongly developing over the past years. In this contribution, we present a model-based control concept for the Bionic Motion Robot – a quasi-continuum robot with six actuator degrees of freedom driven by compressed air. The model is based on a constant curvature approach for the mechanics also being used for the feed-back control, and a dynamic representation of the bellows pressures. The control law itself consists of a cascaded concept, where an underlying pressure control compensates for the compressibility of the driving fluid. With measurements, the working principle of the proposed methods are illustrated and the effectiveness of the control is demonstrated.

Keywords: Control applications, Mechanical systems/robotics, PID control
Abstract: In some applications of control, the objective is to optimize the constant asymptotic response of the system by moving the state of the system from one forced equilibrium to another. Since suppression of the transient response is not the main objective, the feedback control law can operate quasi-statically, that is, extremely slowly relative to the open-loop dynamics. Although integral control can be used to achieve the desired setpoint, three issues must be addressed, namely, nonlinearity, uncertainty, and multistability, where multistability refers to the fact that multiple locally stable equilibria may exist for the same constant input; multistability is the mechanism underlying hysteresis. The present paper applies an adaptive digital PID controller to achieve quasi-static control of systems that are nonlinear, uncertain, and multistable. The approach is demonstrated on multistable systems involving unmodeled cubic and backlash nonlinearities.

Keywords: Cooperative control, Autonomous robots, Networked control systems
Abstract: We consider the problem of manipulating an unknown payload using multiple, non-communicating agents. The objective is to find an optimal input force of each agent so that linear and angular velocity of a rigid object tracks a reference velocity. We show that the primal–dual gradient dynamics for the associated optimization program can be completely decoupled into local dynamics that each agent can implement using only their own measurements. We prove that the proposed optimization dynamics converge locally at an exponential rate, and provide numerical simulations that demonstrate their performance on practical problems.

Keywords: Decentralized control, Autonomous robots, Adaptive control
Abstract: We present a decentralized adaptive control strategy for collective payload transport by differential-drive robots with manipulator arms. The controllers only require robots' measurements of their own heading and velocity and their manipulator angle and angular velocity, and the only information provided to the robots is the target speed and direction of transport. The control strategy does not rely on inter-robot communication, prior information about the load dynamics and geometry, or knowledge of the number of robots and their distribution around the payload. We first design the desired manifolds of motion for the entire system such that they are compatible with the holonomic constraints between the robots and the payload. Then, we design adaptive controllers for a team of differential-drive robots that initially grasp a payload in an arbitrary configuration. We also analytically establish the stability and convergence of the system trajectories to the desired payload motion. We demonstrate the effectiveness of the proposed controllers through 3D physics simulations with realistic dynamics.

Keywords: Estimation, Agents-based systems, Kalman filtering
Abstract: With the rise in use for ground robots, there is a need for efficient detection and tracking to improve SLAM performed by each such agent. This paper analyses the effectiveness of the Dynamic Joint Probabilistic Data Association (DJPDA) framework for target tracking in dense environments by creating a test bed with a fleet of ground robots. DJPDA is a framework that utilizes tensor decomposition, a commonly used technique to tackle “curse of dimensionality”, to handle the exponential growth in the binary non-competing joint association events (or feasible events) in Joint Probabilistic Data Association (JPDA) filter. The number of feasible events in JPDA is reduced by utilizing the “core” tensor, a result of the tensor decomposition, as a surrogate for the input of JPDA instead of the complete set of measurements. The test bed created for this experiment consists of 5 Kobuki ground robots. The laser scan data from the on-board Xbox Kinect sensor is collected for each time step by using one of these robots as the observer. Finally, the collected point cloud data is passed to the DJPDA framework for offline computation of the tracks to compare these predicted tracks with the true tracks obtained from the odometry data for each ground robot.

Keywords: Mechanical systems/robotics, Robotics, Optimal control
Abstract: A capture point approach is applied for a humanoid to push carts. When friction compensation was added, such material-handling was more effective. By contrast, without such compensation, external disturbances yielded unstable walking. As such, tuning dynamic models of walking coupled with arm compliance yielded better performance. This is important to ensure balance and minimize the disturbance effects.

Keywords: Mechanical systems/robotics, Mechatronics, Control laboratories
Abstract: Real-time Python refers to using Python in real-time feedback control experiments by combining an Arduino microcontroller with a computer. This paper uses a Raspberry Pi to improve upon a previous method that combined a laptop and an Arduino. The primary improvement is switching from serial to i2c for communication between the Arduino and Python, which significantly reduces the latency in communications. The reduction in latency allows the digital control frequency to increase from 200 Hz to 500 Hz. The latency improvements are verified by oscilloscope measurements. The new i2c based approach is applied to vibration suppression control for a 3D printed beam.

Keywords: Mechanical systems/robotics, Robotics, Estimation
Abstract: This paper presents an optimized approach to planning energy-aware paths for skid-steer vehicles during elevation changes. Specifically, this work expands upon a previously presented power model by including the effect of elevation changes on the energy usage of the robot. The total power needed to travel to a goal location is then combined with an instantaneous center of rotation (ICR) kinematic model to plan energy-aware paths using a Sampling Based Model Predictive Optimization (SBMPO) algorithm. This method is demonstrated using a simulated environment with a wide range of varying scenarios representative of real-world usages. The results show that, in some hilly cases, it is more energy efficient to take a longer path when operating skid-steer robots on mixed terrain. These results are intended to improve the accuracy of energy consumption models for robotics.

Keywords: Mechanical systems/robotics, Autonomous robots, Robotics
Abstract: The ability to track a general walking path with specific timing is crucial to the operational safety and reliability of bipedal robots for avoiding dynamic obstacles, such as pedestrians, in complex environments. This paper introduces an online, full-body motion planner that generates the desired impact-aware motion for fully-actuated bipedal robotic walking. The main novelty of the proposed planner lies in its capability of producing desired motions in real-time that respect the discrete impact dynamics and the desired impact timing. To derive the proposed planner, a full-order hybrid dynamic model of fully-actuated bipedal robotic walking is presented, including both continuous dynamics and discrete lading impacts. Next, the proposed impact-aware online motion planner is introduced. Finally, simulation results of a 3-D bipedal robot are provided to confirm the effectiveness of the proposed online impact-aware planner. The online planner is capable of generating full-body motion of one walking step within 0.6 second, which is shorter than a typical bipedal walking step.

Keywords: Nonholonomic systems, Lyapunov methods, Autonomous robots
Abstract: We present a partial-consensus controller for a network of nonholonomic mobile robots in the presence of time-varying communication delays. The control problem that we address is that of having a group of vehicles converging to a relatively common non-specified Cartesian position value, but each having an imposed desired orientation. The vehicles are assumed to communicate over a network with a connected, undirected graph topology. The proposed decentralized control law is smooth time-varying, of the delta−Persistently-Exciting class. We establish uniform global asymptotic stability for the closed-loop system.

Keywords: Autonomous robots, Robust control, Game theory
Abstract: Imperfect weather forecasts complicate robot planning. A conservative motion planning algorithm is developed to address uncertainty in autonomous boat missions. Dynamic Programming generates an optimal action for every location and game theory strategically handles uncertainty. Experimental results using Massachusetts Bay forecasts show the robot is able minimize the cost of worst-case weather.

Keywords: Robotics, Adaptive control, Uncertain systems
Abstract: This paper considers formation control of multiple wave-adaptive-modular vessels (WAM-Vs) with the aid of neighbors' information when there are parametric uncertainty and non-parametric uncertainty in the dynamics of each WAM-V. With the aid of backstepping techniques, distributed robust tracking controllers are proposed. To avoid calculation of the derivative of signals, distributed command filtered controllers are also proposed. Simulation results show the effectiveness of the proposed controllers.

Keywords: Robotics, Human-in-the-loop control
Abstract: The field of soft robotics has evinced considerable interest recently due to its importance in several practical applications. Teleoperation of a soft manipulator to do a multitude of tasks in a remote environment is one such promising application. The dexterity and conformity of a soft robot can be constructively utilized for enhanced motion planning and manipulability in cluttered environments. To that end, this paper investigates an adaptive task space bilateral teleoperation framework for soft robots with dynamic uncertainties assuming a non-redundant rigid master manipulator and a redundant soft slave manipulator under the piecewise constant curvature hypothesis. First, the dynamics of the soft robot are approximated as a rigid link manipulator with elastic joints using an existing augmented formulation in the literature. The task space adaptive bilateral teleoperation framework is then introduced based on this rigid-robot-like formulation. The null space velocity of the soft robot is also exploited to achieve sub-task objectives. Finally, the proposed control algorithms are experimentally investigated on a planar soft robot and the results are discussed pointing out the important observations.

Keywords: Robotics, Hybrid systems
Abstract: In this paper, we present a continuous norm-regulation-based limit cycle control for vertical hoppers, inspired by the phased-locked controller for an anchored spring-mass-damper system. Our approach provides continuous real-time norm regulation during the stance phase, leading to faster convergent rate and larger disturbance rejection capability as compared to the conventional impulsive or continuous stance phase control approaches. We analytically proved that the proposed controller can asymptotically stabilize the vertical hopper to a desired limit cycle with an intermittent transition between the stance and flight phases. In addition, we also proved that the convergence of the controller can be preserved even with model parameter uncertainty. Simulations and experiments were conducted to demonstrate the effectiveness of the proposed controller.

Keywords: Robotics
Abstract: This article establishes an Explicit Model Predictive Control (EMPC) scheme for controlling the adhesion of a climbing Vortex Robot (VR). The VR utilizes an Electric Ducted Fan (EDF) as the Vortex Actuator (VA), where the dynamics have been identified via an AutoRegressive-Moving-Average, with eXternal input (ARMAX) identification scheme. An explicit controller via the use of a Constraint Finite Time Optimal Control (CFTOC) approach is designed in an offline manner and implemented for the case of the VR, where the adhesion reference is provided by a static force model. The presented approach results in a lookup table realization that ensures overall system stability in all state transitions, while being able to accurately control the adhesion force for arbitrary setup orientations. The efficacy of the proposed control scheme is demonstrated through experimental results involving a moving test surface under random inclinations and robot orientations.

Keywords: Robotics, Optimization, Autonomous robots
Abstract: Mobile manipulators, constructed by mobile platforms and manipulators, have become a promising solution to future factories for introducing flexibility to manufacturing. This paper presents a method, hierarchical receding horizon control algorithm (HRHC), to assure safety and achieve higher time and space efficiency in robots surrounded by time-varying environments. HRHC contains an optimization-based motion planning module that takes account of both the mobile platform and the manipulator to utilize the kinematic redundancy, and a low-level safety controller to deal with fast changes in the environment. With this method, we verify the performance through experiments. The result shows that space efficiency is increased and the HRHC can guarantee local safety in dynamic environments.

Keywords: Robotics, Lyapunov methods, Autonomous robots
Abstract: In this paper, we investigate the task of approaching a rigid tumbling satellite (Target) with a fully-actuated manipulator-equipped spacecraft (Servicer). We consider a Servicer with an end-effector-mounted exteroceptive sensor for feedback of Target motion. This sensor, however, provides only a noisy relative pose (position and orientation) of the tumbling Target's grasping frame. For this time-varying scenario, we propose a novel method, which is a cascade interconnection of a geometric Extended Kalman Filter (EKF) observer and a geometric controller. The key idea is to estimate the unforced Target's full state-space with the proposed EKF, and then use these estimates in feed-forward and feedback terms of the control law, while exploiting the fully-actuated Servicer. This results in a cascade interconnection, for which we prove the Local Asymptotic Stability (LAS) property. Furthermore, the effectiveness of the proposed method for the approach task is demonstrated through simulation.

Keywords: Robotics, Mechanical systems/robotics
Abstract: This paper proposes a method to generate feasible trajectories for robotic systems with predefined sequences of switched contacts. The proposed trajectory generation method relies on sampling-based methods, optimal control, and reachability analysis. In particular, the proposed method is able to quickly test whether a simplified model-based planner, such as the Time-to-Velocity-Reversal planner, provides a reachable contact location based on reachability analysis of the multi-body robot system. When the contact location is reachable, we generate a feasible trajectory to change the contact mode of the robotic system smoothly. To perform reachability analysis efficiently, we devise a method to compute forward and backward reachable sets based on element-wise optimization over a finite time horizon. Then, we compute robot trajectories by employing optimal control. The main contributions of this study are the following. Firstly, we guarantee whether planned contact locations via simplified models are feasible by the robot system. Secondly, we generate optimal trajectories subject to various constraints given a feasible contact sequence. Lastly, we improve the efficiency of computing reachable sets for a class of constrained nonlinear systems by incorporating bi-directional propagation (forward and backward). To validate our methods we perform numerical simulations applied to a humanoid robot walking.

Keywords: Control applications, Lyapunov methods
Abstract: One of the pressing concerns for next-generation manufacturing is the development of techniques for guaranteeing that a control system is cyberattack-resilient in the sense that even if a cyberattack is successful at breaking information technology-based defenses (e.g., it succeeds at providing a false sensor measurement to the controller), closed-loop stability is still maintained. Our prior work has provided a nonlinear systems definition for cyberattacks. This work explores how a nonlinear systems perspective on cyberattack-resilience for false sensor measurements provided to controllers may allow an economic model predictive control (EMPC) formulation known as Lyapunov-based EMPC (LEMPC) to be designed such that if a cyberattack occurs at a sampling time, the closed-loop state will not leave a region where a known feedback control law exists that can stabilize the origin of the closed-loop system if the cyberattack is detected and non-falsified state measurements are then provided within that sampling period.

Keywords: Building and facility automation, Optimal control, Simulation
Abstract: In this paper, a model predictive control (MPC) framework for building energy system setpoint optimization is developed and tested. The performance of the MPC framework is presented in comparison to a baseline case, where a fixed setpoint schedule is used. To simulate the MPC procedure, an EnergyPlus building model is used to represent the actual building that the optimal setpoints are applied to, and a Gaussian process (GP) regression meta-model is used in the MPC procedure that generates the optimal setpoints. The performance outputs that are used for evaluation are total heating, ventilation and air conditioning (HVAC) energy usage and the Fanger predicted mean vote (PMV) thermal comfort measure. The inputs for the GP regression meta-models are selected to be representative of data points that could be obtained by modern supervisory control and data acquisition (SCADA) systems to support data-driven building models. The supervisory MPC framework is capable of reducing the total energy usage with minor adjustments in thermal comfort.

Keywords: Chemical process control, Robust control, Process Control
Abstract: In this article, a novel distributionally robust optimization (DRO) based multistage model predictive control (MPC) framework is proposed to hedge against the uncertainty in control problems. Without a priori knowledge of the exact uncertainty distribution, an ambiguity set, constructed based on principal component analysis, incorporates the first-order moment information instead. A linear performance measure is chosen so that the worst-case expected problem can be exactly dualized by adopting the affine decision rule. By considering input and state constraints robustly with respect to a support set, which specifies the domain of the uncertainty, the DRO-based MPC model is developed as a robust optimization problem. The proposed framework is illustrated on a two-mass-spring system and results show the improved control performance compared to traditional control strategies.

Keywords: Networked control systems, Communication networks, Predictive control for nonlinear systems
Abstract: Rollout approaches are an effective tool to address the problem of co-designing the transmission schedule and the corresponding input values, when the controller is connected to the plant via a resource-constrained communication network. These approaches typically employ an MPC, activated at multiples of the period length of a base transmission schedule. Using multi-step invariant terminal regions and a prediction horizon longer than the base period, stability can be ensured. This strategy, however, suffers from intrinsic shortcomings, such as a high computational complexity and low robustness. We aim to resolve these drawbacks by proposing an MPC with periodically time-varying terminal region and cost for the rollout setup, where the controller is activated at each time instant and features an arbitrary but fixed prediction horizon. We consider in more detail two specific setups from the literature on Networked Control Systems, namely the token bucket network and actuator scheduling. For both setups, we provide conditions for which convergence under application of the time-varying MPC can be guaranteed. In a numerical example, we demonstrate the benefits of the proposed method.

Keywords: Nonlinear output feedback, Predictive control for nonlinear systems, Stability of nonlinear systems
Abstract: This paper presents an output feedback model predictive control for a class of nonlinear systems in a multi-rate scheme, where the control sampling period is larger than the estimation sampling period. With a small sampling period, the observer is designed to be faster than the dynamics of the closed-loop system under state feedback. Stabilization is achieved by a separation approach in which the control is designed first using state feedback and practical stabilization is achieved by output feedback.

Keywords: Predictive control for linear systems, Constrained control, LMIs
Abstract: The purpose of this paper is to simplify the treatment of major issues of Model Predictive Control (MPC) like stability, feasibility and computational burden. To this purpose a new MPC policy is developed using a two Degrees of Freedom (2DoF) control scheme where the output r(k) of the feedforward Input Estimator (IE) is used as input forcing the closed-loop system Sigma_f . f is the feedback connection of an LTI plant p with an LTI feedback controller g. The task of g is to guarantee the stability of Sigma_f , as well as the fulfillment of hard constraints on some physical variables for any input r(k) satisfying an ”a priori” determined admissibility condition. The input r(k) is computed by the feedforward IE through the on-line minimization of a suitably definite finite-horizon quadratic cost functional and is applied to f according to the usual receding horizon strategy. To improve numerical efficiency, r(k) is assumed to be given by a B-spline function. This greatly decreases the number of decision variables of the on-line optimization procedure. It is shown that stability and recursive feasibility of the adopted MPC strategy are guaranteed in advance, regardless the chosen prediction horizon.

Keywords: Predictive control for linear systems, Model/Controller reduction, Manufacturing systems
Abstract: Model predictive control (MPC) has been used in a variety of industrial processes. Due to its large computation time application to fast and complex processes is limited. Therefore, reducing complexity has been focus of intense research. In this paper, a complexity reduction strategy for a linear MPC is developed. It is used for the control of an overactuated roll gap with two different actuator types in a cold rolling mill. One of the actuators is able to accept new set points only at a much slower sample time than the fast actuator. In order to coordinate the actuators, a moving manipulation point scheme is introduced where a single time-varying optimization variable of the slow actuator moves through the control horizon of the fast actuator. The fast actuator ensures reference tracking regarding the roll gap when the slow one is idle. The proposed scheme reduces the number of optimization variables as well as constraints and thus enables control of faster processes. Simulation results show the proof of concept and an enhancement in computation time. Moreover, a real-time implementation is demonstrated on a cold rolling mill.

Keywords: Predictive control for linear systems, Stochastic optimal control, Distributed control
Abstract: In this paper, we present a novel stochastic output-feedback MPC scheme for distributed systems with additive process and measurement noise. The chance constraints are treated with the concept of probabilistic reachable sets, which, under a unimodality assumption on the disturbance distributions are guaranteed to be satisfied in closed-loop. By conditioning the initial state of the optimization problem on feasibility, the fundamental property of recursive feasibility is ensured. Closed-loop chance constraint satisfaction, recursive feasibility and convergence to an asymptotic average cost bound are proven. An example of three interconnected subsystems is carried out.

Keywords: Process Control, Predictive control for linear systems
Abstract: We propose a stochastic model predictive control scheme for linear discrete-time and time-invariant systems with chance constraints and additive normal distributed disturbance with known covariance. The proposed approach allows to balance between performance optimization and robustness by adjusting the probability p of the chance constraints. The Controllability Gramian and the Riccati inequality are used to determine a set that contains all disturbances with a probability p. The proposed approach guarantees asymptotic stability of the nominal system, probabilistic convergence of the real system as well as constraint satisfaction and recursive feasibility in terms of probability. A proof of concept example considers the control of a wind turbine, where the wind acts as a normal distributed disturbance.

Keywords: Predictive control for linear systems, Robust control, Uncertain systems
Abstract: Classical tube based MPC utilizes a fixed linear feedback to counteract the effect of the disturbance in the prediction. This allows to bound the error between the predictions and the real system in form of sets; i.e. the real state evolution is bounded to sets called tubes. The choice of the corresponding feedback determines the size and shape of the tube and thus influences the domain of attraction and the control performance. Choosing the feedback often requires to compromise different objectives. We propose to compose the tube online based on multiple tubes, each of which is determined by a different linear gain to reduce the conservatism and to improve the controller performance. The underlying optimization problem remains convex and the computational demand increases only marginally. We discuss the resulting closed loop properties such as constraint satisfaction, robust recursive feasibility and stability. Moreover an adaption of the cost function is investigated, which allows to improve the closed loop performance. An example illustrates the proposed approach and its benefits.

Keywords: Predictive control for nonlinear systems, Automotive control, Control applications
Abstract: A Non-linear Model Predictive Control (NMPC) scheme is proposed for the optimal power-spit problem of a hybrid diesel-electric marine propulsion plant. The NMPC controller directly regulates the torque setpoints of the internal combustion engine and the electric motor/generator and ensures that certain constraints concerning the engine overloading are applied. In this way, fuel consumption and NOx emissions can be reduced. The modeling for the controller design was based on experimental data gathered from the hybrid plant and on first principles for the diesel engine behavior and battery charging. The controller was experimentally tested in real-time operation. Results showed that the controller rejected successfully load disturbances and maintained the rotational speed reference of the powertrain as well as the desirable state of charge in battery, in respect to the limits of the system.

Keywords: Predictive control for nonlinear systems, Optimal control, Identification
Abstract: In this paper a real-time capable hierarchical nonlinear model predictive control framework for the energy management of a hybrid electric vehicle is presented. As high-level energy management, nonlinear model predictive control is employed. Therefore, a nonlinear optimal control problem is formulated using a control-oriented internal model derived from high-fidelity models and experimental data. The multiple shooting algorithm and an Euler backward scheme are used to discretize the optimal control problem. The resulting nonlinear problem is solved in real-time using Sequential Quadratic Programming. A rule-based gear choice and engine on / off strategy is added. Actuator dynamics and drivability are adressed in a fast low-level linear time-variant model predictive ontroller. The result is analyzed and compared to the optimal solution obtained by dynamic programming using a simplified model of the vehicle.

Keywords: Predictive control for linear systems, Estimation, Chemical process control
Abstract: The paper is concerned with the multi-criteria and multivariate Control Performance Assessment (CPA) of Multiple-Input Multiple-Output Model Predictive Control. The task of control quality estimation in case of the multi-variable MPC poses serious challenges. The main difficulty is the fact that the engineer has to cope with multiple signals with no single measure that might be representative. Thus, there is a need for an approach which could compare and take into account various assessment aspects. Proposed method combines different integral, Gaussian and non-Gaussian measures and shows robustness to real-life disturbances.

Keywords: Predictive control for nonlinear systems, Constrained control, Chemical process control
Abstract: In this paper, a composite model predictive control (MPC) structure for the class of standard singularly perturbed systems is developed. To control and optimize economic performance of the slow dynamics, an economic MPC is developed and to stabilize the fast dynamics, a fast MPC is designed. Both MPC strategies do not rely on explicit stabilizing and/or terminal constraints. Working within a sampled-data setting, closed-loop stability in the sense of asymptotic stability of the economically optimal steady state is discussed. A chemical process example is given to demonstrate the control architecture.

Keywords: Predictive control for nonlinear systems, Process Control, Stability of nonlinear systems
Abstract: A method for integrating optimization and control during on-line process operation is known as economic model predictive control (EMPC). EMPC optimizes a general cost function which reflects process economics subject to a model of the process. One formulation of EMPC which can maintain closed-loop stability in the presence of sufficiently small disturbances is Lyapunov-based EMPC (LEMPC). In this work, we make precise connections between closed-loop stability considerations under LEMPC and numerical approximations (via Taylor series) of the solution of the nonlinear dynamic model of the process used in the controller. A chemical process example is utilized to demonstrate the concepts developed.

Keywords: Process Control, Uncertain systems, Chemical process control
Abstract: Appropriate greenhouse temperature should be maintained to ensure crop production while minimizing energy consumption. Even though weather forecasts could provide a certain amount of information to improve control performance, it is not perfect and forecast error may cause the temperature to deviate from the acceptable range. To inherent uncertainty in weather that affects control accuracy, this paper develops a data-driven robust model predictive control (DDRMPC) approach for greenhouse temperature control. The dynamic model is obtained from thermal resistance-capacitance modeling derived by the Building Resistance-Capacitance Modeling (BRCM) toolbox. Uncertainty sets of ambient temperature and solar radiation are captured by support vector clustering technique, and they are further tuned for better quality by training-calibration procedure. A case study shows that the DDRMPC has better control performance compared to rule-based control, certainty equivalent MPC, and robust MPC. DDRMPC approach ends up with 12% less total energy consumption than rule-based control strategy.

Keywords: Process Control, Optimization, Robust control
Abstract: Irrigation takes considerable amount of water; however, in many cases up to half of it is wasted. Improving the efficiency of irrigation control, therefore, is an important task for sustainable water management. Most existing irrigation control systems are based on soil moisture level. In this work, we explore theoretically the use of continuous values stem water potential (SWP) as a basis for control. SWP is a more direct measure of plant water status than the soil moisture level. After linearizing and discretizing a nonlinear dynamic model of water dynamics in a plant, we develop a model predictive control (MPC) framework for regulating SWP. To prevent plants from suffering water stress, data-driven robust MPC (DDRMPC) which captures the uncertainty of weather forecast error is implemented. A case study based on almond tree is presented to characterize the effectiveness of the DDRMPC strategy relative to on-off control. Sensitivity analysis on the prediction horizon and penalty weights were performed to investigate the varying irrigation control decisions. For the case characterized, the analysis shows that controlling tree trunk water potential through DDRMPC can save 2.5% amount of water comparing to on-off control while maintaining zero violation.

Keywords: Predictive control for nonlinear systems, Flexible structures, Distributed parameter systems
Abstract: Distributed parameter systems (DPS) are formulated as partial differential equations (PDE). Especially, under time-varying boundary conditions, PDE introduce force coupling. In the case of the flexible stacker crane (STC), nonlinear coupling is introduced. Accordingly, online trajectory planning and tracking can be addressed using a nonlinear model predictive control (NMPC). However, due to the high computational demands of a NMPC, this paper discusses a possibility of embedding nonlinearities inside a linear parameter varying (LPV) system and thus make a use of a numerically low-demanding linear MPC. The resulting mismatches are treated as parametric and additive uncertainties in the context of robust tube-based MPC (TMPC). For the proposed approach, most of the computations are carried out offline. Only a simple convex quadratic program (QP) is conducted online. Additionally a soft-constrained extension was briefly proposed. Simulation results are used to illustrate the good performance, closed-loop stability and recursive feasibility of the proposed approach despite uncertainties.

Keywords: Predictive control for nonlinear systems, Optimization algorithms, Identification for control
Abstract: We consider the problem of data-driven, on-the-fly, constrained control of unknown nonlinear dynamics from a single finite-horizon trajectory. We are motivated by applications with severe limits on data availability and require online controller synthesis. Using a one-step receding horizon control framework, we pose the data-driven optimal control problem as a contextual optimization problem. Specifically, we seek to iteratively minimize a smooth (black-box) objective function over iteration-dependent constraints. We propose C2Opt, a data-driven, convex-optimization-based approach to solve this problem. We utilize correct-by-construction function approximation bounds on the objective function to guide the decision process. Our bounds are synthesized from data and smoothness assumptions, and accommodates incorporation of side information like convexity, monotonicity, and loose bounds on the function. We then use tractable convex optimization problems to solve for the iterates, given the context and past data. We demonstrate C2Opt in successful navigation of a unicycle to a target destination, using only the data collected from a single trajectory under control constraints.

Keywords: Predictive control for nonlinear systems, Nonholonomic systems, Autonomous systems
Abstract: This paper introduces an accurate nonlinear model predictive control-based algorithm for trajectory following. The algorithm incorporates both the planned state and control trajectories into its cost functional - generally, following algorithms do not incorporate control trajectories into their cost functionals. Comparisons are made against two trajectory following algorithms, where the trajectory planning problem is to safely follow a person using an automated ATV with control delays in a dynamic environment while simultaneously optimizing speed and steering, minimizing control effort, and minimizing the time-to-person. Results indicate that the proposed algorithm reduces collisions, tracking error, orientation error, and time-to-goal. Therefore, including the control trajectories into the trajectory following algorithm helps the vehicle follow planned state trajectories more accurately, which ultimately improves safety, especially in dynamic environments.

Keywords: Predictive control for nonlinear systems, Lyapunov methods, Human-in-the-loop control
Abstract: Lyapunov-based economic model predictive control (LEMPC) is an optimization-based control design that computes economically-optimal control actions for a process while maintaining the closed-loop state within a bounded region of state-space; however, it may be difficult to design in practice without closed-loop simulations, as it requires an auxiliary stabilizing controller, Lyapunov function, and a number of sets to be developed to ensure closed-loop stability. Practical application of this method could benefit from methods which make it more likely that, without simulations to identify aspects of the control design that would provide stability, controller parameters can be selected that would maintain stability. In this work, we propose a method to seek to enhance tractability of LEMPC by providing initial suggestions for reducing the likelihood that textit{ad hoc} selection of a value for one of its parameters would be problematic for closed-loop stability.

Keywords: Predictive control for nonlinear systems, Autonomous systems, Learning
Abstract: A task decomposition method for iterative learning model predictive control is presented. We consider a constrained nonlinear dynamical system and assume the availability of state-input pair datasets which solve a task T1. Our objective is to find a feasible model predictive control policy for a second task, T2, using stored data from T1. Our approach applies to tasks T2 which are composed of subtasks contained in T1. In this paper we formally define subtasks and the task decomposition problem, and provide proofs of feasibility and iteration cost improvement over simple initializations. We demonstrate the effectiveness of the proposed method on autonomous racing and robotic manipulation experiments.

Keywords: Robotics, Networked control systems, Large-scale systems
Abstract: Multi-agent coverage control is used as a mechanism to influence the behavior of a group of robots by introducing time-varying domains. The coverage optimization problem is modified to adopt time-varying domains, and the proposed control law possesses an exponential convergence characteristic. Complex multi-agent control is simplified by specifying the desired distribution and behavior of the robot team as a whole. In the proposed approach, design of the inputs to the multi-agent system, i.e., time-varying density and time-varying domain, are agnostic to the size of the system. Analytic expressions of surface and line integrals present in the control law are obtained under uniform density. The scalability of the proposed control strategy is analyzed and verified via numerical simulation. Experiments on real robots are used to test the proposed control law.

Keywords: Autonomous robots, Linear systems, Autonomous systems
Abstract: Malicious attacks on modern autonomous cyber-physical systems (CPSs) can leverage information about the system dynamics and noise characteristics to hide while hijacking the system toward undesired states. Given attacks attempting to hide within the system noise profile to remain undetected, an attacker with the intent to hijack a system will alter sensor measurements, contradicting with what is expected by the system's model. To deal with this problem, in this paper we present a framework to detect non-randomness in sensor measurements on CPSs under the effect of sensor attacks. Specifically, we propose a run-time monitor that leverages two statistical tests, the Wilcoxon Signed-Rank test and Serial Independence Runs test to detect inconsistent patterns in the measurement data. For the proposed statistical tests we provide formal guarantees and bounds for attack detection. We validate our approach through simulations and experiments on an unmanned ground vehicle (UGV) under stealthy attacks and compare our framework with other anomaly detectors.

Keywords: Autonomous systems, Robotics
Abstract: The Rimon-Koditscheck potential is known to be a navigation function for sufficiently curved worlds. With global information, a diffeomorphism can be constructed to convert complex star worlds into sphere worlds so that the point agent can follow the negative gradient of the navigation function potential. Without global information, convergence guarantees only exist for sufficiently curved worlds or ellipsoidal worlds. This work closes the gap between local information and global information by extending convergence guarantees with local information to star worlds in general. Convergence to the goal and obstacle avoidance is established from every initial position in the free space. Results are numerically verified with a discretized version of the proposed dynamic flow.

Keywords: Control applications, Mechanical systems/robotics, Fluid power control
Abstract: The applications, modeling and control of continuum manipulators have been strongly developing over the past years. In this contribution, we present a model-based control concept for the Bionic Motion Robot – a quasi-continuum robot with six actuator degrees of freedom driven by compressed air. The model is based on a constant curvature approach for the mechanics also being used for the feed-back control, and a dynamic representation of the bellows pressures. The control law itself consists of a cascaded concept, where an underlying pressure control compensates for the compressibility of the driving fluid. With measurements, the working principle of the proposed methods are illustrated and the effectiveness of the control is demonstrated.

Keywords: Control applications, Mechanical systems/robotics, PID control
Abstract: In some applications of control, the objective is to optimize the constant asymptotic response of the system by moving the state of the system from one forced equilibrium to another. Since suppression of the transient response is not the main objective, the feedback control law can operate quasi-statically, that is, extremely slowly relative to the open-loop dynamics. Although integral control can be used to achieve the desired setpoint, three issues must be addressed, namely, nonlinearity, uncertainty, and multistability, where multistability refers to the fact that multiple locally stable equilibria may exist for the same constant input; multistability is the mechanism underlying hysteresis. The present paper applies an adaptive digital PID controller to achieve quasi-static control of systems that are nonlinear, uncertain, and multistable. The approach is demonstrated on multistable systems involving unmodeled cubic and backlash nonlinearities.

Keywords: Cooperative control, Autonomous robots, Networked control systems
Abstract: We consider the problem of manipulating an unknown payload using multiple, non-communicating agents. The objective is to find an optimal input force of each agent so that linear and angular velocity of a rigid object tracks a reference velocity. We show that the primal–dual gradient dynamics for the associated optimization program can be completely decoupled into local dynamics that each agent can implement using only their own measurements. We prove that the proposed optimization dynamics converge locally at an exponential rate, and provide numerical simulations that demonstrate their performance on practical problems.

Keywords: Decentralized control, Autonomous robots, Adaptive control
Abstract: We present a decentralized adaptive control strategy for collective payload transport by differential-drive robots with manipulator arms. The controllers only require robots' measurements of their own heading and velocity and their manipulator angle and angular velocity, and the only information provided to the robots is the target speed and direction of transport. The control strategy does not rely on inter-robot communication, prior information about the load dynamics and geometry, or knowledge of the number of robots and their distribution around the payload. We first design the desired manifolds of motion for the entire system such that they are compatible with the holonomic constraints between the robots and the payload. Then, we design adaptive controllers for a team of differential-drive robots that initially grasp a payload in an arbitrary configuration. We also analytically establish the stability and convergence of the system trajectories to the desired payload motion. We demonstrate the effectiveness of the proposed controllers through 3D physics simulations with realistic dynamics.

Keywords: Estimation, Agents-based systems, Kalman filtering
Abstract: With the rise in use for ground robots, there is a need for efficient detection and tracking to improve SLAM performed by each such agent. This paper analyses the effectiveness of the Dynamic Joint Probabilistic Data Association (DJPDA) framework for target tracking in dense environments by creating a test bed with a fleet of ground robots. DJPDA is a framework that utilizes tensor decomposition, a commonly used technique to tackle “curse of dimensionality”, to handle the exponential growth in the binary non-competing joint association events (or feasible events) in Joint Probabilistic Data Association (JPDA) filter. The number of feasible events in JPDA is reduced by utilizing the “core” tensor, a result of the tensor decomposition, as a surrogate for the input of JPDA instead of the complete set of measurements. The test bed created for this experiment consists of 5 Kobuki ground robots. The laser scan data from the on-board Xbox Kinect sensor is collected for each time step by using one of these robots as the observer. Finally, the collected point cloud data is passed to the DJPDA framework for offline computation of the tracks to compare these predicted tracks with the true tracks obtained from the odometry data for each ground robot.

Keywords: Mechanical systems/robotics, Robotics, Optimal control
Abstract: A capture point approach is applied for a humanoid to push carts. When friction compensation was added, such material-handling was more effective. By contrast, without such compensation, external disturbances yielded unstable walking. As such, tuning dynamic models of walking coupled with arm compliance yielded better performance. This is important to ensure balance and minimize the disturbance effects.

Keywords: Mechanical systems/robotics, Mechatronics, Control laboratories
Abstract: Real-time Python refers to using Python in real-time feedback control experiments by combining an Arduino microcontroller with a computer. This paper uses a Raspberry Pi to improve upon a previous method that combined a laptop and an Arduino. The primary improvement is switching from serial to i2c for communication between the Arduino and Python, which significantly reduces the latency in communications. The reduction in latency allows the digital control frequency to increase from 200 Hz to 500 Hz. The latency improvements are verified by oscilloscope measurements. The new i2c based approach is applied to vibration suppression control for a 3D printed beam.

Keywords: Mechanical systems/robotics, Robotics, Estimation
Abstract: This paper presents an optimized approach to planning energy-aware paths for skid-steer vehicles during elevation changes. Specifically, this work expands upon a previously presented power model by including the effect of elevation changes on the energy usage of the robot. The total power needed to travel to a goal location is then combined with an instantaneous center of rotation (ICR) kinematic model to plan energy-aware paths using a Sampling Based Model Predictive Optimization (SBMPO) algorithm. This method is demonstrated using a simulated environment with a wide range of varying scenarios representative of real-world usages. The results show that, in some hilly cases, it is more energy efficient to take a longer path when operating skid-steer robots on mixed terrain. These results are intended to improve the accuracy of energy consumption models for robotics.

Keywords: Mechanical systems/robotics, Autonomous robots, Robotics
Abstract: The ability to track a general walking path with specific timing is crucial to the operational safety and reliability of bipedal robots for avoiding dynamic obstacles, such as pedestrians, in complex environments. This paper introduces an online, full-body motion planner that generates the desired impact-aware motion for fully-actuated bipedal robotic walking. The main novelty of the proposed planner lies in its capability of producing desired motions in real-time that respect the discrete impact dynamics and the desired impact timing. To derive the proposed planner, a full-order hybrid dynamic model of fully-actuated bipedal robotic walking is presented, including both continuous dynamics and discrete lading impacts. Next, the proposed impact-aware online motion planner is introduced. Finally, simulation results of a 3-D bipedal robot are provided to confirm the effectiveness of the proposed online impact-aware planner. The online planner is capable of generating full-body motion of one walking step within 0.6 second, which is shorter than a typical bipedal walking step.

Keywords: Nonholonomic systems, Lyapunov methods, Autonomous robots
Abstract: We present a partial-consensus controller for a network of nonholonomic mobile robots in the presence of time-varying communication delays. The control problem that we address is that of having a group of vehicles converging to a relatively common non-specified Cartesian position value, but each having an imposed desired orientation. The vehicles are assumed to communicate over a network with a connected, undirected graph topology. The proposed decentralized control law is smooth time-varying, of the delta−Persistently-Exciting class. We establish uniform global asymptotic stability for the closed-loop system.

Keywords: Autonomous robots, Robust control, Game theory
Abstract: Imperfect weather forecasts complicate robot planning. A conservative motion planning algorithm is developed to address uncertainty in autonomous boat missions. Dynamic Programming generates an optimal action for every location and game theory strategically handles uncertainty. Experimental results using Massachusetts Bay forecasts show the robot is able minimize the cost of worst-case weather.

Keywords: Robotics, Adaptive control, Uncertain systems
Abstract: This paper considers formation control of multiple wave-adaptive-modular vessels (WAM-Vs) with the aid of neighbors' information when there are parametric uncertainty and non-parametric uncertainty in the dynamics of each WAM-V. With the aid of backstepping techniques, distributed robust tracking controllers are proposed. To avoid calculation of the derivative of signals, distributed command filtered controllers are also proposed. Simulation results show the effectiveness of the proposed controllers.

Keywords: Robotics, Human-in-the-loop control
Abstract: The field of soft robotics has evinced considerable interest recently due to its importance in several practical applications. Teleoperation of a soft manipulator to do a multitude of tasks in a remote environment is one such promising application. The dexterity and conformity of a soft robot can be constructively utilized for enhanced motion planning and manipulability in cluttered environments. To that end, this paper investigates an adaptive task space bilateral teleoperation framework for soft robots with dynamic uncertainties assuming a non-redundant rigid master manipulator and a redundant soft slave manipulator under the piecewise constant curvature hypothesis. First, the dynamics of the soft robot are approximated as a rigid link manipulator with elastic joints using an existing augmented formulation in the literature. The task space adaptive bilateral teleoperation framework is then introduced based on this rigid-robot-like formulation. The null space velocity of the soft robot is also exploited to achieve sub-task objectives. Finally, the proposed control algorithms are experimentally investigated on a planar soft robot and the results are discussed pointing out the important observations.

Keywords: Robotics, Hybrid systems
Abstract: In this paper, we present a continuous norm-regulation-based limit cycle control for vertical hoppers, inspired by the phased-locked controller for an anchored spring-mass-damper system. Our approach provides continuous real-time norm regulation during the stance phase, leading to faster convergent rate and larger disturbance rejection capability as compared to the conventional impulsive or continuous stance phase control approaches. We analytically proved that the proposed controller can asymptotically stabilize the vertical hopper to a desired limit cycle with an intermittent transition between the stance and flight phases. In addition, we also proved that the convergence of the controller can be preserved even with model parameter uncertainty. Simulations and experiments were conducted to demonstrate the effectiveness of the proposed controller.

Keywords: Robotics
Abstract: This article establishes an Explicit Model Predictive Control (EMPC) scheme for controlling the adhesion of a climbing Vortex Robot (VR). The VR utilizes an Electric Ducted Fan (EDF) as the Vortex Actuator (VA), where the dynamics have been identified via an AutoRegressive-Moving-Average, with eXternal input (ARMAX) identification scheme. An explicit controller via the use of a Constraint Finite Time Optimal Control (CFTOC) approach is designed in an offline manner and implemented for the case of the VR, where the adhesion reference is provided by a static force model. The presented approach results in a lookup table realization that ensures overall system stability in all state transitions, while being able to accurately control the adhesion force for arbitrary setup orientations. The efficacy of the proposed control scheme is demonstrated through experimental results involving a moving test surface under random inclinations and robot orientations.

Keywords: Robotics, Optimization, Autonomous robots
Abstract: Mobile manipulators, constructed by mobile platforms and manipulators, have become a promising solution to future factories for introducing flexibility to manufacturing. This paper presents a method, hierarchical receding horizon control algorithm (HRHC), to assure safety and achieve higher time and space efficiency in robots surrounded by time-varying environments. HRHC contains an optimization-based motion planning module that takes account of both the mobile platform and the manipulator to utilize the kinematic redundancy, and a low-level safety controller to deal with fast changes in the environment. With this method, we verify the performance through experiments. The result shows that space efficiency is increased and the HRHC can guarantee local safety in dynamic environments.

Keywords: Robotics, Lyapunov methods, Autonomous robots
Abstract: In this paper, we investigate the task of approaching a rigid tumbling satellite (Target) with a fully-actuated manipulator-equipped spacecraft (Servicer). We consider a Servicer with an end-effector-mounted exteroceptive sensor for feedback of Target motion. This sensor, however, provides only a noisy relative pose (position and orientation) of the tumbling Target's grasping frame. For this time-varying scenario, we propose a novel method, which is a cascade interconnection of a geometric Extended Kalman Filter (EKF) observer and a geometric controller. The key idea is to estimate the unforced Target's full state-space with the proposed EKF, and then use these estimates in feed-forward and feedback terms of the control law, while exploiting the fully-actuated Servicer. This results in a cascade interconnection, for which we prove the Local Asymptotic Stability (LAS) property. Furthermore, the effectiveness of the proposed method for the approach task is demonstrated through simulation.

Keywords: Robotics, Mechanical systems/robotics
Abstract: This paper proposes a method to generate feasible trajectories for robotic systems with predefined sequences of switched contacts. The proposed trajectory generation method relies on sampling-based methods, optimal control, and reachability analysis. In particular, the proposed method is able to quickly test whether a simplified model-based planner, such as the Time-to-Velocity-Reversal planner, provides a reachable contact location based on reachability analysis of the multi-body robot system. When the contact location is reachable, we generate a feasible trajectory to change the contact mode of the robotic system smoothly. To perform reachability analysis efficiently, we devise a method to compute forward and backward reachable sets based on element-wise optimization over a finite time horizon. Then, we compute robot trajectories by employing optimal control. The main contributions of this study are the following. Firstly, we guarantee whether planned contact locations via simplified models are feasible by the robot system. Secondly, we generate optimal trajectories subject to various constraints given a feasible contact sequence. Lastly, we improve the efficiency of computing reachable sets for a class of constrained nonlinear systems by incorporating bi-directional propagation (forward and backward). To validate our methods we perform numerical simulations applied to a humanoid robot walking.

Keywords:
Abstract: The Robotarium is a remotely accessible swarm robotics lab that allows users from all over the world to upload control code, written in MATLAB, and run experiments. Since its official launch in August 2017, over 5000 remote experiments have been conducted by users from all continents (except Antarctica). The impetus behind the Robotarium project is to provide broad, democratized access to a world-class research facility, and users span the gambit from robotics researchers to middle-school students. This talk will discuss the technical challenges associated with the Robotarium as well as a lessons learned in remote-access experimentation.

Keywords:
Abstract: Automated vehicles are computers that perform several functions necessary to understand the world and make driving decisions. Developing such systems is challenging, since driving is a multi-variable, multi-objective, nonlinear and sometimes uncertain task, in which multiple agents including drivers, pedestrians, devices and environment interact in real-time.

This talk provides a technical review of lateral controls in GM's SuperCruise, the industry’s first hands-free driving technology for the highway. Several aspects of the system are discussed, including systems and components, hardware redundancy to ensure safety, hardware/software integration, and technical aspects in vehicle dynamics, sensing, fusion, path planning and controls.

Specific case studies are provided which highlight application of controls techniques to develop various functionalities that enable operation of SuperCruise.

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.

Office hours allow for individual Q&A with Program Managers.

Noon - 1:30pm - Dr. Jordan Berg

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.

Office hours allow for individual Q&A with Program Managers.

Noon - 1:30pm - Dr. Eduardo Misawa

Keywords:
Abstract: Room 57: WIC advisory board meeting Time: Jul 2, 2020 11:00 AM Mountain Time (US and Canada)

Join Zoom Meeting https://us02web.zoom.us/j/86500270870 Contact Meeting organizer for password

Meeting ID: 865 0027 0870 One tap mobile +16699006833,,86500270870#,,1#,306233# US (San Jose) +12532158782,,86500270870#,,1#,306233# US (Tacoma)

Dial by your location +1 669 900 6833 US (San Jose) +1 253 215 8782 US (Tacoma) +1 346 248 7799 US (Houston) +1 929 436 2866 US (New York) +1 301 715 8592 US (Germantown) +1 312 626 6799 US (Chicago) Meeting ID: 865 0027 0870 Password: 306233 Find your local number: https://us02web.zoom.us/u/ketFVyCt6E

Keywords:
Abstract: Room 58: Joint 2020/2021 ACC OpComm Meeting Time: Jul 2, 2020 12:00Noon Mountain Time (US and Canada)

Join Zoom Meeting https://us02web.zoom.us/j/83469084414

Meeting ID: 834 6908 4414 Contact Meeting Organizer for Password

One tap mobile +12532158782,,83469084414#,,1#,402613# US (Tacoma) +13462487799,,83469084414#,,1#,402613# US (Houston)

Dial by your location +1 253 215 8782 US (Tacoma) +1 346 248 7799 US (Houston) +1 669 900 6833 US (San Jose) +1 301 715 8592 US (Germantown) +1 312 626 6799 US (Chicago) +1 929 436 2866 US (New York) Meeting ID: 834 6908 4414 Password: 402613 Find your local number: https://us02web.zoom.us/u/khXOHFxK0

Keywords:
Abstract: Room 48: IEEE CSS MAB Time: Jul 2, 2020 12:00 PM Mountain Time (US and Canada)

Join Zoom Meeting https://us02web.zoom.us/j/87386647903

Meeting ID: 873 8664 7903 Use 2020 ACC conference password

One tap mobile +16699006833,,87386647903#,,,,0#,,000747# US (San Jose) +12532158782,,87386647903#,,,,0#,,000747# US (Tacoma)

Dial by your location +1 669 900 6833 US (San Jose) +1 253 215 8782 US (Tacoma) +1 346 248 7799 US (Houston) +1 929 436 2866 US (New York) +1 301 715 8592 US (Germantown) +1 312 626 6799 US (Chicago) Meeting ID: 873 8664 7903 Password: 000747 Find your local number: https://us02web.zoom.us/u/kdkdbgghxB

Keywords:
Abstract: Room 59: IEEE CSS TAB Meeting Time: Jul 2, 2020 12:00 PM Mountain Time (US and Canada)

Join Zoom Meeting https://us02web.zoom.us/j/84056103085

Meeting ID: 840 5610 3085 Use 2020 ACC conference password

One tap mobile +16699006833,,84056103085#,,1#,426960# US (San Jose) +12532158782,,84056103085#,,1#,426960# US (Tacoma)

Dial by your location +1 669 900 6833 US (San Jose) +1 253 215 8782 US (Tacoma) +1 346 248 7799 US (Houston) +1 929 436 2866 US (New York) +1 301 715 8592 US (Germantown) +1 312 626 6799 US (Chicago) Meeting ID: 840 5610 3085 Password: 426960 Find your local number: https://us02web.zoom.us/u/kemf3DvtDX

Keywords: Learning, Optimal control
Abstract: This paper presents a safe off-policy reinforcement learning (RL) scheme to design optimal controllers for systems with uncertain dynamics. The utility function for which its optimization achieves a desired behavior is augmented with a control barrier function (CBF) candidate providing a platform for merging safety planning and optimal control design. A damping factor is included in the CBF providing a design tool to specify the relative importance of performance and safety. As one of the main contributions of this paper, it is shown that by iterative approximation of the value function, the safety properties of CBF are certified which bridges the broad capability of barrier functions into the learning-based approaches. Then, the safety of control system is proved accordingly. Stability and optimality of the control system in a safe condition are verified. Afterward, an off-policy RL algorithm is used to obtain the safe and optimal controller without requiring full knowledge about the system dynamics. The efficiency of the proposed method is demonstrated on the lane changing as an automotive control problem.

Keywords: Learning, Optimization, Game theory
Abstract: Scenarios in which multiple agents interact in a goal-directed manner can be modeled as differential games. Here each agent aims to optimize an associated cost function. Systems ranging from classical economic models to emerging applications in autonomous vehicles can fit into this framework. In scenarios such as analysis of human or animal behaviors, the cost-functions are not known. In order to design control strategies that interact with such agents, the objectives must be modeled or identified. This paper presents a methodology for identifying cost functions for interacting agents. In particular, we identify costs that lead to open-loop Nash equilibria for nonzero-sum constrained differential games. To solve this problem, we extend an inverse optimal control method, known as residual optimization, to the case of differential games. In this paper, we show that residual optimization is tractable for constrained differential games. Specifically, the residual optimization problems turn out to be fully decoupled. Numerical examples indicate that accurate costs can be learned from observing trajectories.

Keywords: Learning, Stochastic optimal control, Filtering
Abstract: This paper presents a bio-inspired central pattern generator (CPG)-type architecture for learning optimal maneuvering control of periodic locomotory gaits. The architecture is presented here with the aid of a snake robot model problem involving planar locomotion of coupled rigid body systems. The maneuver involves clockwise or counterclockwise turning from a nominally straight path. The CPG circuit is realized as a coupled oscillator feedback particle filter. The collective dynamics of the filter are used to approximate a posterior distribution that is used to construct the optimal control input for maneuvering the robot. A Q-learning algorithm is applied to learn the approximate optimal control law. The issues surrounding the parametrization of the Q-function are discussed. The theoretical results are illustrated with numerics for a 5-link snake robot system.

Keywords: Machine learning, Computational methods, Learning
Abstract: Compressed sensing refers to the recovery of high- dimensional but low-complexity objects from a small number of measurements. The recovery of sparse vectors and the recovery of low-rank matrices are the main applications of compressed sensing theory. In vector recovery, the restricted isometry property (RIP) and the robust null space property (RNSP) are the two widely used sufficient conditions for achieving compressed sensing. Until recently, RIP and RNSP were viewed as two separate sufficient conditions. However, in a recent paper [1], the present authors have shown that in fact the RIP implies the RNSP, thus establishing the fact that RNSP is a weaker sufficient condition than RIP. In matrix recovery, there are three different sufficient con- ditions for achieving low-rank matrix reconstruction, namely; Rank Restricted Isometry Property (RRIP), Rank Robust Null Space Property (RRNSP), and Robust Uniform Boundedness Property (RUBP). In this paper, using the result of [1], it is shown that actually both RRIP and RUBP imply the RRNSP, so that RRNSP is the weakest sufficient condition for matrix recovery. In contrast with the situation for vector recovery, until now there are no deterministic methods for designing a measurement operator for matrix recovery. The present results open the door towards such a possibility.

Keywords: Machine learning, Learning, Communication networks
Abstract: We propose a communication-aware Gaussian process regression algorithm that allows a network of robots to collaboratively learn about a common latent function in real time using streaming data. We quantify the improvement that inter-robot communication brings on the transient performance of the learning algorithm. Simulations are performed to validate the proposed algorithm.

Keywords: Machine learning, Learning, Computational methods
Abstract: In this paper, we study the matrix completion problem: Suppose X in R^{n_r times n_c} is unknown except for an upper bound r on its rank. By measuring a small number m ll n_r n_c of elements of X, is it possible to recover X exactly, or at least, to construct a reasonable approximation of X? At present, there are two approaches to choosing the sample set, namely probabilistic and deterministic. Probabilistic methods can guarantee exact recovery of the unknown matrix, but only with high probability. In this approach, samples are taken uniformly at random. Therefore we need to start sampling for every new matrix afresh. In the deterministic approach, sampling points can be kept fixed. At present, there are very few deterministic methods, and they mostly apply only to square matrices. In this paper, we present a deterministic method for selecting the sample set that can guarantee the exact recovery of the unknown matrix. This approach works for the recovery of rectangular as well as square matrices. We achieve this by choosing the elements to be sampled as the edge set of a Ramanujan bigraph. If samples are the edge set of a Ramanujan bigraph, then we can recover the unknown matrix from that sample set using nuclear norm minimization. A companion paper discusses the explicit construction of Ramanujan bigraphs. We provide a sufficient condition, that is if the samples taken are of the order of r^3 then we can recover the unknown entries exactly if the unknown matrix satisfies some coherence condition. We believe this the first sufficient condition available using deterministic sampling technique and nuclear norm minimization. The exact recovery of matrices using Ramnujan sampling is studied here, that is when samples does not have any measurement noise. The stable recovery of matrices, that is the deterministic matrix completion with noisy measurements is studied in another paper using the same sampling technique.

Keywords: Estimation, Energy systems, Observers for nonlinear systems
Abstract: This manuscript presents an algorithm for individual Lithium-ion (Li-ion) battery cell state of charge (SOC) estimation when multiple cells are connected in parallel, using only terminal voltage and total current measurements. For battery packs consisting of thousands of cells, it is desirable to estimate individual SOCs by only monitoring the total current in order to reduce sensing cost. Mathematically, series connected cells yield dynamics given by ordinary differential equations under classical full voltage sensing. In contrast, parallel connected cells are evidently more challenging because the dynamics are governed by a nonlinear descriptor system, including differential equations and algebraic equations arising from voltage and current balance across cells. This paper designs and analyzes an observer with linear output error injection, where the individual cell SOCs and local currents are locally observable from the total current and voltage measurements. The asymptotic convergence of differential and algebraic states is established by considering local Lipschitz continuity property of system nonlinearities. Simulation results on LiNiMnCoO_2/Graphite (NMC) cells illustrate convergence for SOCs, local currents, and terminal voltage.

Keywords: Energy systems, Simulation, Reduced order modeling
Abstract: In this paper, we utilize a Doyle-Fuller-Newman (DFN) model including capacity-loss side reactions to present a model-based design method for multi-stage charging protocols. This design method allows making a trade-off between charging time and battery ageing in a more systematic way. The results are leveraged by a highly efficient implementation of the DFN model, that has a short computation time. We show that by obtaining the Pareto front that describes the optimal trade-off between charging time and battery ageing for a single cycle, the results can be extended to the lifetime of the battery. Finally we show that the negative electrode over-potential is not always a good indicator for ageing, and that ageing will occur even when the battery operates in over-potential regions that are considered to not lead to ageing.

Keywords: Energy systems, Predictive control for nonlinear systems
Abstract: One of the factors limiting the range of electric vehicles is cell imbalance. This means that the condition of a single cell can cause the entire pack to be shut down, thus missing out on energy in the other cells. Active cell balancing can be used to overcome cell imbalance, but a balancing strategy is required that maximises the range. This paper proposes a real-time active-cell-balancing strategy based on model predictive control. To prevent the need for large prediction horizons, several supplementary balancing objectives, such as voltage, SoC and a charge-based quantity are considered. The controller performance is compared to a range benchmark, obtained using a method from previous work. Furthermore, the range increase is demonstrated on a scenario of realistic length and the influence of uncertainty in the future power demand is investigated. On average, this strategy can provide a range increase of approximately 5 percent and the controller is shown to be robust to uncertainties in the power demand, as long as a worst-case prediction is considered.

Keywords: Large-scale systems, Energy systems, Optimal control
Abstract: This paper explores tractable robust optimal control of nonlinear systems with large state spaces. Conventional applications of surrogate modeling for control replace the underlying dynamical model with a data-driven surrogate function. For large-scale systems, this approach possesses a host of shortcomings. We address these challenges by presenting a novel robust surrogate optimization framework for finite-time and receding horizon optimal control. Rather than modeling the entire state transition function, we define a surrogate model which maps the initial state and time series of control inputs to an approximate objective function value. We also define surrogate models which predict time series of relevant constraint functions. Since the bulk of the relevant information is encoded in the initial state, we apply a principal component analysis to project the state onto a reduced basis, allowing surrogate models with tractable parameterizations. To guarantee constraint satisfaction, we use phi-divergence to formulate distributionally robust chance constraints which are satisfied for worst-case realizations of the test data modeling error distribution. We validate our approach using a case study of optimal lithium-ion battery fast charging using a large-scale electrochemical battery model.

Keywords: Smart grid, Distributed control, Optimization algorithms
Abstract: In this paper, a distributed price-responsive energy management algorithm is proposed for a smart residential energy system (RES) equipped with multiple energy storage devices. First, the future system states are predicted via an iterative learning approach based on the lifted domain representation. Then, RES management is formulated as an optimization problem by taking into account the time-varying electricity rate, battery properties, and system operational constraints. Finally, we adopt the Alternating Direction Method of Multipliers (ADMM) and compute the optimal charging/discharging actions of local batteries in a distributed manner to establish a flexible, scalable, and computation-efficient power network. Numerical simulation is provided to illustrate the performance of our proposed algorithm.

Keywords: Energy systems, Automotive systems, Optimization
Abstract: The concept of Hybrid Energy Storage Systems (HESS) has recently re-emerged in the context of electrification of medium- to heavy-duty freight vehicles. While HESS offer the opportunity to capitalize on the advantages of high energy density and high power density Lithium ion batteries to reduce the overall pack weight and improve durability, they introduce new complexities in the pack energy and thermal management strategies.

This paper presents a method for the optimization of the control policy and thermal management strategy for HESS for an extended range hybrid electric vehicle based upon convex optimization. Starting from a zero-order equivalent circuit model of the battery pack and the formulation of an energy-optimal control problem, the equations are convexified to optimize the instantaneous power split over a prescribed duty cycle, and the method is benchmarked against dynamic programming (DP). To manage the pack temperature, a radiator model is convexified and incorporated into the overall optimization framework.

Keywords: Automotive control, Automotive systems, Linear parameter-varying systems
Abstract: Tire pressure has a high impact on the tire-road contact 0because it influences the characteristics of the tire forces. During the maneuvering of the vehicle the pressures of the tires may decrease over time, which results in performance degradation or the loss of controllability. This paper proposes a novel integration of tire pressure estimation and path-following control design based on machine learning and Linear Parameter-Varying (LPV) methods. In the estimation process the vehicle dynamic signals, which are available from the conventional on-board sensors, are fused. The values of the estimated tire pressures are incorporated in the LPV control as scheduling variables. The results of the control system are the steering and the differential drive interventions on the vehicle. The effectiveness of the method is illustrated through comprehensive simulation scenarios through the CarMaker simulation enviroment.

Keywords: Automotive control, Automotive systems, Linear parameter-varying systems
Abstract: The paper presents a new big data based control design for autonomous vehicles. The main contribution of this work is the longitudinal velocity optimization process, which is based on the approximation of the reachability sets of a passenger vehicle by using a machine-learning approach. The data, which is used for the approximation, is provided by the high-fidelity car simulation software, CarSim. The approximation is performed by applying a well-known decision tree algorithm, C4.5. The reachability sets are computed for different longitudinal velocities. Moreover, a LPV technique based lateral control design is proposed, which is used to guarantee the trajectory tracking of the vehicle. To enhance the capability of the LPV controller, the control scheme is extended with the longitudinal velocity optimization process. Thus, the stable and safe motion of the vehicle is guaranteed.

Keywords: Automotive control, Energy systems, Mechatronics
Abstract: Regenerative suspension systems, unlike traditional passive, semi-active or active setups, are able to convert the traditionally wasted kinetic energy into electricity. This paper discusses flexible multi-objective control design strategies based on LMI formulations to suitably trade-off between the usual road handling and ride comfort performance and the amount of energy to be harvested. An electromechanical regenerative vehicle suspension system is considered where the viscous damper at each wheel is replaced by a linear electrical motor which is actively governed. It is shown by simulations that multivariable centralized control laws designed on the basis of a full-car model of the suspension system are able to achieve larger amount of harvested energy under identical ride comfort prescriptions with respect to scalar control strategies, designed on the basis of a single quarter-car model and implemented independently on each wheel in a decentralized way.

Keywords: Automotive control, Optimization, Modeling
Abstract: Traditional shock absorbers dissipate large amount of vibration energy into heat waste via viscous oil dampers. To harvest such energy and improve the vehicle suspension performance, a novel energy-harvesting shock absorber that uses a mechanical motion rectifier (MMR) is introduced with a rule-based controller to improve vehicle ride comfort as well as harvested energy under the random road excitations with different roughness classes. The ride comfort performance of the controlled MMR shock absorber with rule-based strategy is compared with the passive shock absorber, the controlled traditional shock absorber with skyhook strategy, and the controlled MMR shock absorber with SH-PDD (skyhook-power driven damper) strategy. The rule-based MMR shock absorber shows the best ride comfort performance.

Keywords: Automotive control, Optimization algorithms, H-infinity control
Abstract: Most of recent researches on the automotive field focus on autonomous vehicles. These vehicles are equipped by conventional chassis systems. The goal is to control the vehicle’s traction, brakes, and front steering. This paper discusses the importance of advanced chassis systems, as driving/braking torque vectoring, for both autonomous and non-autonomous vehicles, especially in a race mode. Reliable co-simulation results shows that expending the vehicle’s potential leads to high performances and safety with respect to severe situations when optimal control allocation is ensured. Therefore, future passenger cars shall not only be equipped by additional sensors, but also by advanced systems along with adequate control algorithms.

Keywords: Automotive control, Robust control, Estimation
Abstract: This paper considers cooperative adaptive cruise control (CACC) for automated vehicles on banked and curved roads. In such context, the existing longitudinal control alone cannot guarantee robustly stable CACC, because the vehicle longitudinal dynamics are coupled with the lateral dynamics. This gives rise to the necessity of incorporating the lane keeping (LK) control with CACC for the following vehicle to guarantee lateral stability and adaptive cruise. A robustly independent design strategy is proposed for determining a linear parameter varying (LPV) LK controller and a constant gain CACC controller. The influence of sensor bias and noise is also considered and a linear time-varying generalized augmented state observer (LTV-GASO) is developed to obtain optimal state estimation of the CACC and LK systems. Designs of the observers and controllers for CACC and LK systems follow a similar scheme with an easily solved linear optimization problem formulation. This facilitates the overall vehicle control system design and implementation. Simulation of a platoon in actual traffic environment illustrates efficacy of the proposed designs in achieving robustly stable vehicle following and lateral stability.

Keywords: Automotive control, Automotive systems, Optimal control
Abstract: Driving style is known to have a big impact on fuel consumption, and ecodriving is usually understood as an approach to determine the optimal speed profile leading to the highest fuel efficiency for a given route and conditions. Ecodriving has been mainly studied for highway traffic and without taking into account the influence of other road users, dynamic programming (DP) being the standard method for computation. However, under real conditions the presence of other road users forces to deviations from the ideal speed profile. Against this background, there have been suggestions to split the problem into two optimization layers. In this paper, we follow the same idea and examine different approaches for the lower layer to cope with the actual safety conditions but recovering as much as possible of the ideal efficiency, using the ideal solution as a reference for the lower layer . It is shown that a significant part of the fuel savings from the optimal solution can be recovered under rather general conditions.

Keywords: Automotive control, Automotive systems, Pattern recognition and classification
Abstract: Variability of traffic conditions is a well known fact. Traditionally, vehicles have been developed to cope with many very different conditions, both in terms of environment and traffic, albeit at the price of, among other, reduced efficiency under many conditions. The progressive increase of on board computational power, sensors and other available information offers new possibilities, in particular, besides achieving their main function, ADAS can be tailored to achieve additional benefits, e.g. reduce fuel consumption or improve driving comfort. However, we may expect that the levels of improvement strongly depend on the specific traffic conditions. To check this hypothesis, in this paper we use measured data to build up clusters of traffic situations, and then analyze by simulation the achievable improvements of fuel consumption vs. comfort at the example of a specific ADAS, both in the case of averages and of single vehicles from the clusters. As the examples confirm, the trade-off is indeed context dependent, and control tuning should be adapted accordingly.

Keywords: Multivehicle systems, Decentralized control
Abstract: This paper considers a platoon of connected and automated vehicles implementing the pulse-and-glide (PnG) strategy. PnG is an eco-driving technique that periodically turns on and off the engine to achieve better fuel economy. Although the literature reported significant fuel savings with PnG, such benefits may degrade considerably due to the inability to complete the target PnG cycles when the PnG vehicles are not driving alone. Moreover, the variety in powertrains and preferences between ride comfort and fuel economy motivate different accelerations in PnG. Therefore, platooning of PnG vehicles is a challenging task. To overcome this challenge, this paper presents a decentralized PnG synchronization method to coordinate the PnG vehicles in platoons, so that the vehicles can realize their target PnG cycles. In particular, each vehicle maintains a virtual oscillator that is synchronized using vehicle-to-vehicle communication with the other virtual oscillators in the platoon via the Kuramoto model. With the virtual oscillator, the desired vehicle states are obtained via the phase-angle-parameterized target phase portrait of PnG. A state feedback controller is then used to track the desired vehicle states. Thus the PnG vehicles are synchronized after the synchronization of the virtual oscillators. A range keeping mechanism integrated with this feedback controller is also designed. Numerical simulations show that this synchronization method helps the PnG vehicles in a platoon realize their target PnG cycles even if their acceleration limits are different. This synchronization is achieved while maintaining the desired range with a small amount of range oscillations, leading to a more compact platoon.

Keywords: Automotive control, Feedback linearization, Optimal control
Abstract: Recent advancements in vehicle connectivity and advanced driver-assistance systems allow for more efficient driving in automated driving applications on the road. The practice of truck platooning utilizes following distances as small as a few meters of each vehicle in a string to benefit from slipstream effects and reduce aerodynamic drag. By this, fuel economy is then improved in the vehicles. In this work, the impact of a connected and anticipative cruise controller in a truck platooning application is explored. We utilize two separate optimal controllers: 1) a time-invariant kinematic model with a first-order lag on the acceleration of the vehicle - intended strictly as a connected gap-tracking controller, and 2) a time-varying dynamic model which considers linearized aerodynamic drag terms - intended as an engine demand optimizer. We consider penalty terms in the cost functional to promote string compactness and reduce accelerations incurred, and propose a set of linear constraints which restrict truck capabilities to those of a realistic engine. We compare our optimal controllers to an Intelligent Driver Model baseline modeled after human response, and we find that our time-invariant connected truck platoon performs with 14% better fuel economy, whereas our time-varying connected truck platoon performs with 20% better fuel economy. By comparing our time-invariant and time-varying controllers, we conclude that demanded work on the engine due to maintaining a strict gap between trucks is detrimental to fuel economy.

Keywords: Automotive control, Control applications, Automotive systems
Abstract: In this work, we integrate two once separate concepts for longitudinal control of heavy duty vehicles: responding to elevation changes to improve fuel economy using preview and reacting to the motion of preceding vehicles using feedback. The two concepts are unified to provide a safe yet fuel efficient connected and automated technology for heavy duty vehicles. First, we establish an integrated control framework of the two concepts based on barrier function theory and then we discuss the detailed control design of each concept. Finally, we demonstrate the benefits of the proposed design against a naive switching controller by experimentally evaluating the performance of a connected automated truck.

Keywords: Automotive control, Predictive control for nonlinear systems, Optimization
Abstract: The growing vehicle connectivity and autonomy in the ground transportation system are not only able to improve traffic safety but also fuel efficiency. This paper proposes a receding-horizon optimization-based nonlinear model predictive control (NMPC) algorithm to achieve fuel-saving speed planning for connected vehicles. The NMPC method solves for the fuel-optimal speed profile of connected vehicles considering a short speed preview of the preceding vehicle. By utilizing such previewing information through vehicle connectivity, the fuel consumption of the connected vehicles is reduced by avoiding unnecessary braking and acceleration, particularly in transient operating conditions. In order to analyze the effectiveness of NMPC design, dynamic programming (DP) method is adopted as a benchmark algorithm where the full speed preview of the preceding vehicle is known. The performances of NMPC and DP designs in driving behavior and fuel economy are quantitatively explored and compared under several standard driving cycles. Results show a promising improvement of the performance by adopting the proposed design and reveal the potential fuel benefits brought by vehicle connectivity and autonomy.

Keywords: Smart grid, Nonlinear systems identification, Control of networks
Abstract: The rapid penetration of uncertain and unpredictable renewable energy resources into the power grid has made generation planning and real-time power balancing a challenge, thus rendering demand side advanced control a necessity. In this work, we introduce field validation of a distributed optimization framework for load flexibility control in the grid where aggregated power of multiple loads follow a command ancillary service command power. The proposed scheme is a model predictive control (MPC) approach where the pertaining large scale optimization problem is solved in a distributed fashion based on Newton iteration to provide fast convergence, and is scalable to large numbers of loads at the aggregation level [1]. The target load is a campus comprising commercial buildings. We introduce a simplified model of a complex commercial building HVAC system to achieve observability given limited available measurements. Moreover, the simplified model makes the modeling and estimation scalable in the sense that it can be applied to different buildings with minimal adjustments. We demonstrate the performance of estimator and MPC controller via closed-loop control of a commercial building on the Navy Yard campus in Philadelphia to track a commanded power while comfort measures are kept within acceptable range. It is shown that the estimator tracks output measurements well while parameters stay relatively constant and the optimizer converges in real-time. The paper is an introductory step for optimal building control with the purpose of providing grid ancillary services. It introduces a scalable approach from the perspective of modeling, estimation and distributed optimization implementing MPC on an aggregation of multiple commercial buildings. While the candidate distributed energy resource (DER) is a commercial building, the novel MPC approach is not restricted to buildings; rather, it is designed to be applicable to general large scale heterogeneous loads, as outlined in [1]

[1] R. Ghaemi, M. Abbaszadeh, and P. Bonanni, “Optimal flexibility control of large-scale distributed heterogenous loads in the power grid,” in IEEE Transactions on Control of Network Systems. IEEE, 2019, pp. 1256 – 1268.

Keywords: Building and facility automation, Control applications
Abstract: We propose a reinforcement learning-based (RL) controller for energy efficient climate control of commercial buildings. Model-based control techniques like model predictive control (MPC) for this problem are challenging to implement as they need simple yet accurate models, which are hard to obtain due to the complexity in hygrothermal dynamics of a building and its HVAC system. RL is an attractive alternative to MPC since once the policy is learned, computing the control in real time involves solving a simple low dimensional optimization problem that does not involve a model of building physics. However, training an RL controller is computationally expensive, and there are many design choices that affect performance.

We compare in simulations the proposed RL controller, an MPC controller, and a baseline rule-based controller that is widely used in practice. Both the RL and MPC controllers are able to maintain temperature and humidity constraints, and they both reduce energy use significantly compared to the baseline, though the savings by RL is smaller than that by MPC.

Keywords: Control applications, Building and facility automation, Energy systems
Abstract: The thermal inertia of buildings, along with the flexibility associated with thermostatically controlled loads (TCL) allows heating, ventilation and cooling (HVAC) systems to be used for grid demand response (DR). In this work, we consider a hydronic HVAC system that serves multiple units in a residential building to meet their space heating requirements. We aim to determine the optimal power flow to each unit that minimizes the power costs incurred by the building's occupants while keeping in consideration their thermal comfort. The DR program is assumed to allow the building temperatures to deviate from the set-points up to a maximum limit. Despite the complex, non-linear structure of the problem, we show how the optimal solutions can be obtained efficiently using quadratic programming. Since HVAC systems can run on either electricity or natural gas, we study the efficacy of the DR regime for both hourly electricity prices and flat gas prices over the course of 24 hours. We also study the optimal thermal power and the evolution of unit temperatures for various energy pricing schemes.

Keywords: Building and facility automation, Supervisory control, Stochastic optimal control
Abstract: Building thermal energy storage (TES) can provide value to building owners while helping the electric grid. In this paper, a transactive approach to controlling thermal energy storage is developed for multiple buildings considering electric grid incentives. A two-stage building control framework is proposed to plan day-ahead electricity procurement and real-time TES operation. Day-ahead planning is decided by a two-stage stochastic optimization framework to account for uncertainty in the occupant behavior and weather of the following day. In the real-time operation, a transactive market mechanism is utilized for load flexibility created by TES operation. Real-time operations are based on solving a model predictive control (MPC) problem at the aggregator level to dispatch thermal storage via transactive markets. Simulation case studies were conducted to evaluate the proposed framework by comparing the performance of the stochastic planning and control with the deterministic approach. This paper demonstrates the effectiveness of the developed framework in operating a portfolio of thermal storage resources in consideration of uncertainty.

Keywords: Building and facility automation, Modeling, Neural networks
Abstract: Smart buildings and building automation are key components for achieving greater energy efficiency. In order to implement predictive control for the HVAC system and natural ventilation, the model needs to have the capability to predict building thermal responses under various environmental and operational conditions. This task can be accomplished by using a deep neural network, which would capture the effects of complicated physical processes, such as natural ventilation. However, a deep neural network comes with a high demand for training data. In real-world applications, the target building may not have the necessary amount of operational data available. This study demonstrates how transfer learning could solve this dilemma. By freezing partial weights of a deep neural network model that is pretrained on multi-year data from a base case building, it can be quickly deployed to different buildings in other climates. With much fewer trainable parameters, the model can then be well-trained on only 15 days of data from the new target building. The base case and target case can have entirely different floor areas, building materials, and window sizes. This transfer learning model performs significantly better than a comparable model that is only trained on source data or target data, achieving high predicting accuracy on both indoor air temperature and relative humidity in the time horizon from 10 minutes to two hours. This methodology can be applied to the design of the control system in a new building. It has a high sample efficiency and shortens the minimum data collection period for model training.

Keywords: Reduced order modeling, Control applications, Estimation
Abstract: This paper focuses on an application of dynamic mode decomposition (DMD) identification methods and robust estimation theory to thermo-fluid systems modeled by the Boussinesq equations. First, we use Dynamic Mode Decomposition with control (DMDc) to construct a reduced order linear model for the Boussinesq equations. Due to inherent model uncertainties in real applications, we propose robust estimators that minimize an H-infinity norm from disturbance to estimation error. The disturbances we consider here stem from uncertainty in boundary conditions and unknown inputs acting on walls. Numerical simulations on a challenging turbulent flow, of the 2D Boussinesq equations, is used to demonstrate the potential of our approach.

Keywords: Kalman filtering, Distributed control, Control applications
Abstract: As the wind energy industry continues to push for increased power production and lower cost of energy, the focus of research has expanded from individual turbines to entire wind farms. Among a host of interesting problems to be solved when considering the wind farm as a whole, we consider the challenge of scalar field estimation, based on information already collected at the individual turbine level. We aim to estimate the large-scale, low-frequency characteristics of the wind field, such as the mean wind direction and the overall decrease in wind speed across the farm, and employ a Kalman filter that models the wind field using a polynomial function. We compare the proposed method's performance to both a simple averaging technique and filtering of individual turbine measurements. The method presented is not limited to wind turbines and is applicable in other situations where multiple remote agents are used to estimate a scalar field.

Keywords: Smart grid, Linear parameter-varying systems, Control applications
Abstract: Flexible loads have great potential to improve the electric grid's flexibility and stability. To effectively control large ensembles of heterogeneous loads, reliable models thereof are required. This paper presents a data-driven modeling and control approach to manage flexible loads for providing grid services. We leverage a linear parameter-varying autoregressive moving average (LPV-ARMA) model to describe the aggregate load response, where the parameters in the model are used to capture external environmental impacts (e.g., weather). A gain-scheduling feedback controller is then developed to adapt to environmental variations. This data-driven approach can be easily applied to different types of loads in various environmental conditions. In addition to the ensemble controller, distributed load controllers are designed to deliver grid services, while maintaining the quality of service of inherent load tasks. We demonstrate the work on the IEEE 37-node distribution system for real-time power regulation services through control of thermostatically controlled loads.

Keywords: Agents-based systems, Energy systems, Smart grid
Abstract: Traditionally, electrical power was generated in big power plants. The cost of producing energy was related to the cost of fuel, e.g., carbon or gas, and by the cost of maintaining the power plants. With the advent of distributed energy resources, power can be produced directly at the edge of the electrical network by a new type of agents: the prosumers. Prosumers are entities that both consume and generate power, e.g., by means of photovoltaic panels. The cost of the power produced by prosumers is no longer related to fuel consumption since energy coming from distributed generators is essentially free. Rather, the cost is related to the remuneration that is due to the prosumers for the services they provide. The proposed control strategy minimizes the active power generation cost in the aforementioned scenario. The control scheme requires that the prosumers measure their voltage and then adjust the amount of injected power, according to a continuous time feedback control law that is indeed a projected gradient descent strategy. Simulations are provided in order to illustrate the algorithm behavior.

Keywords: Smart grid, Estimation, Iterative learning control
Abstract: This paper outlines a combination of two data-driven approaches leveraging sum-of-squares (SoS) optimization to: i) learn the power-voltage (p-v) characteristic of photovoltaic (PV) arrays, and ii) rapidly regulate operation of the companion PV inverter to a desired power setpoint. Estimation of available headroom in PV systems is critical to the task of providing ancillary services, and the proposed method puts forth a computationally tractable solution with minimal data needs for the same. In addition to providing this key contribution to application, from an algorithmic vantage point, we present an interior-point method to solve a linear regression reformulation of the original polynomial fitting problem with SoS constraints. We validate the proposed algorithms through time-domain numerical simulations (incorporating the PV source and a 15-th order inverter model) for a variety of large-signal disturbances (step changes in real-power demand, rapid changes in irradiance) and demonstrate that the method provides an effective strategy to seamlessly and concomitantly discover the p-v curve and regulate operation to a desired setpoint.

Keywords: Smart grid, Switched systems, Power systems
Abstract: We consider the problem of guaranteeing transient stability of a network of interconnected angle droop controlled microgrids, where voltage phase angle measurements from phasor measurement units (PMUs) may be lost, leading to poor performance and instability. In this paper, we propose a novel mixed voltage angle and frequency droop control (MAFD) framework to improve the reliability of such angle droop controlled microgrid interconnections. In this framework, when the phase angle measurement is lost at a microgrid, conventional frequency droop control is temporarily used for primary control in place of angle droop control. We model the network of interconnected microgrids with the MAFD architecture as a nonlinear switched system. We then propose a dissipativity-based distributed secondary control design to guarantee transient stability of this network under arbitrary switching between angle droop and frequency droop controllers. We demonstrate the performance of this control framework by simulation on a test 123-feeder distribution network.

Keywords: Smart grid, Energy systems, Optimization
Abstract: This paper proposes a mathematical framework to optimally operate a plug-in electric vehicle (PEV) charging station, using differentiated charging services. The mathematical framework specifically exploits human behavioral modeling to alleviate ``overstay'' -- when a PEV remains plugged-in after charging service is complete. Discrete Choice Modeling is utilized to capture human decision-making behavior among multiple charging service options that differ in both price and quality-of-service. We reformulate an associated non-convex problem to a multi-convex problem via the Young-Fenchel transform. We then apply Block Coordinate Descent algorithm to efficiently solve the multi-convex problem. Simulation results show a strong potential of the proposed method in realizing benefits in three ways: (i) net profits gains, (ii) overstay reduction, and (iii) increased quality-of-service.

Keywords: Predictive control for linear systems, Output regulation, Learning
Abstract: This paper presents recursive-least-squares-based model predictive control (RLSMPC). RLSMPC uses only output feedback, and thus does not require full-state measurements. Online learning is performed through concurrent system identification, and thus no a priori model is needed. RLSMPC employs separate RLS algorithms for identification, offset determination, and control. Variable-rate forgetting is used to facilitate system identification and offset estimation.

Keywords: Predictive control for linear systems, Delay systems, Distributed parameter systems
Abstract: This paper studies linear model predictive control of real matrix-valued single delay systems. The delay system is written as an abstract infinite-dimensional control system which is then mapped into an infinite-dimensional discrete-time control system using Cayley-Tustin discretization. A constrained model predictive control (MPC) problem is formulated for the discrete-time system where a terminal penalty function is utilized to cast the infinite-horizon optimization problem into a finite-horizon one. The proposed MPC design is demonstrated on an example of constrained stabilization of a two by two system. We will demonstrate that the proposed discrete-time MPC law not only stabilizes the discrete-time system but can be utilized in stabilizing the original continuous-time system as well, which is due to several favorable properties of the Cayley-Tustin discretization.

Keywords: Predictive control for linear systems, Adaptive control, Uncertain systems
Abstract: This paper addresses the problem of controlling discrete-time linear time-invariant (LTI) systems with parametric uncertainties in the presence of hard state and input constraints. A suitably designed indirect adaptive controller is combined with a model predictive control (MPC) algorithm. An estimated model, corresponding to the uncertain plant, is used for predictions of the future states. The parameters of the estimated model are updated using a gradient descent based adaptive update law. The errors arising due to model mismatch between the estimated system model and the actual uncertain plant are accounted for using a constraint tightening method in the MPC algorithm. The proposed adaptive MPC strategy is proved to be recursively feasible.

Keywords: Predictive control for nonlinear systems, Machine learning, Modeling
Abstract: This work proposes two methods for incorporating structural process knowledge in recurrent neural network (RNN) modeling for a general class of nonlinear dynamic process systems. Specifically, based on the structural a pri- ori knowledge of the relationship between the process state variables, the first approach is to design a partially-connected structure for RNN models. Additionally, a weight-constrained RNN optimization problem equipped with a constraint on weight parameters and a regularization term in loss function is proposed to develop an RNN model that satisfies the assumption on input-output relationship. The proposed partially-connected RNN modeling method is applied to a chemical process example to demonstrate its better approximation performance compared with the fully-connected RNN model that does not incorporate any structural process knowledge.

Keywords: Predictive control for nonlinear systems, Mechatronics
Abstract: This paper proposes a sampling-based model predictive control scheme with a single degree of freedom in control. A variable horizon and stabilizing terminal conditions ensure recursive feasibility, asymptotic stability, and improvement of the closed-loop performance. The initial derivation leads to a computationally demanding mixed-integer nonlinear programming problem. To address this issue, the paper presents an algorithm that seeks the optimal horizon while rolling out a finite number of control input candidates. The proposed derivative-free optimization even finds the control invariant terminal set online at reasonable computational overhead. The approach is straightforward to implement and aims for fast, albeit small-scale systems. A comparative analysis with a fixed-horizon scheme and standard model predictive control on a nonlinear benchmark system shows high closed-loop performance with low computational effort.

Keywords: Predictive control for nonlinear systems, Predictive control for linear systems, Nonlinear systems identification
Abstract: It is now well known that dynamics of nonlinear systems can be lifted to higher or infinite dimensional spaces and represented as linear systems. We call such linear system representations and approximations, ’lifting linear’ representations. Once we have such a linear system representation, we can apply linear control theorems to the previous nonlinear systems.The behavior of the lifting linear system are closer to the original systems than those of Taylor series approximations. In this paper, we compare performances of MPC (Model Predictive Control) for nonlinear systems using several different approximated linear models. When the systems are represented as linear systems, MPC can be solved by convex quadratic optimization methods if the constraints are linear. It will be shown that computational times become much shorter and the optimality of the solutions are improved using particular lifting linearizations.

Keywords: Robotics, Control applications, Stability of nonlinear systems
Abstract: This paper examines the effect of control allocation on the passive input-output properties of a redundantly-actuated parallel robotic manipulator with a point-mass payload. In particular, it is shown that the mapping matrix used for control allocation is to satisfy a forward velocity kinematic constraint in order to preserve the manipulator's passive input-output mapping from modified control torques in task space to the velocity tracking error of the payload. A method to generate the control allocation matrix using load-sharing parameters is proposed, and is shown to have a physically intuitive relationship to the control effort of the individual actuators. A numerical example of a cable-driven parallel robot is presented, which illustrates the intuitive nature of the proposed control allocation method compared to a pseudoinverse method in the literature.

Keywords: Robotics, Emerging control applications, Intelligent systems
Abstract: Disassembly currently is a labor-intensive process with limited automation. The main reason lies in the fact that disassembly usually has to address model variations from different brands, physical uncertainties resulting from component defects or damage during usage, and incomplete product information. To overcome these challenges and to automate the disassembly process through human-robot collaboration, this paper develops a disassembly sequence planner which distributes the disassembly task between human and robot in a human-robot collaborative setting. This sequence planner targets to address potential issues including distinctive products, variant orientations, and safety constraints of human operators. The proposed disassembly sequence planner identifies the locations and orientations of the to-be-disassembled items, determines the starting point, and generates the optimal disassembly sequence while complying with the disassembly rules and considering the safe constraints for human operators. This algorithm is validated by numerical and experimental tests: the robot can successfully locate and disassemble the pieces following the obtained optimal sequence, and complete the task via collaboration with the human operator without violating the constraints.

Keywords: Robotics, Lyapunov methods, Mechanical systems/robotics
Abstract: With the goal of moving towards implementation of increasingly dynamic behaviors on underactuated systems, this paper presents an optimization-based approach for solving full-body dynamics based controllers on underactuated bipedal robots. The primary focus of this paper is on the development of an alternative approach to the implementation of controllers utilizing control Lyapunov function based quadratic programs. This approach utilizes many of the desirable aspects from successful inverse dynamics based controllers in the literature, while also incorporating a variant of control Lyapunov functions that renders better convergence in the context of tracking outputs. The principal benefits of this formulation include a greater ability to add costs which regulate the resulting behavior of the robot, in addition, the model error-prone inertia matrix is used only once, in a non-inverted form. The result is a successful demonstration of the controller for walking in simulation, and applied on hardware in real-time for dynamic crouching.

Keywords: Robotics, Manufacturing systems, Mechanical systems/robotics
Abstract: In this paper we describe a novel path following scheme for robot end-effectors that is particularly suitable for robotic surface finishing operations where constant velocity of travel on the surface is desirable. The scheme is applicable to general situations where the path is typically given in terms of measured data from a sensor, and also to paths that are specified in terms of analytical curves (circular or ellipsoidal). Considering the given data points as control points, we utilize cubic spline interpolation to generate a closed-form geometric description for the path. Since velocity control is quite common in many industrial robots and most surface finishing tasks require travel with constant velocity along the path, we consider a kinematic model for the end-effector with control inputs as rate of change of orientation and translational velocity. By utilizing a path variable and the tangent vector along the path, we describe the complete path as the path that is taken from the initial robot end-effector point to the desired path and subsequent travel on the desired path. To evaluate the performance of the scheme, we have conducted a number of real-time experiments on an industrial robot for circular paths and for paths generated for gear deburring and chamfering, and results from those experiments will be discussed.

Keywords: Robotics, Mechanical systems/robotics
Abstract: Snake robots have the potential to maneuver through tightly packed and complex environments. One challenge in enabling them to do so is the complexity in determining how to coordinate their many degrees-of-freedom to create purposeful motion. This is especially true in the types of terrains considered in this work: environments full of unmodeled features that even the best of maps would not capture, motivating us to develop closed-loop controls to react to those features. To accomplish this, this work uses proprioceptive sensing, mainly the force information measured by the snake robot's joints, to react to unmodeled terrain. We introduce a biologically-inspired strategy called directional compliance which modulates the effective stiffness of the robot so that it conforms to the terrain in some directions and resists in others. We present a dynamical system that switches between modes of locomotion to handle situations in which the robot gets wedged or stuck. This approach enables the snake robot to reliably traverse a planar peg array and an outdoor three-dimensional pile of rocks.

Keywords: Robotics, Network analysis and control, Communication networks
Abstract: We consider a team of mobile robots with limited communication range tasked to coordinate in large environments. We require that the robots maximize the coverage of the environment while maintaining an r-robust communication network. r-robustness guarantees the team's ability to achieve resilient consensus: i.e., achieve consensus in the presence of non-cooperative agents. Existing works for static networks showed that the individual robot communication ranges need to be large to satisfy the r-robustness condition. This paper relaxes this requirement on the communication range by leveraging on the robot motion. Specifically, we design modular dynamic formations for a subset of robots where the robots move along simple closed curves. These modules can be composed together to form larger formations. We derive conditions based on periodic robot connections for individual and interconnected modules. We present module designs that satisfy the sufficient conditions in a lattice space. Simulations are provided to support our theoretical results.

Keywords: Numerical algorithms, Computational methods, Large-scale systems
Abstract: Traditional model reduction derives reduced models from large-scale systems in a one-time computationally expensive offline (training) phase and then evaluates reduced models in an online phase to rapidly predict system outputs; however, this offline/online splitting means that reduced models can be expected to faithfully predict outputs only for system behavior that has been incorporated into the reduced models during the offline phase. This work considers model reduction with the online adaptive empirical interpolation method (AADEIM) that adapts reduced models in the online phase to system behavior that was not anticipated in the offline phase by deriving updates from a few samples of the states of the large-scale systems. The contribution of this work is an analysis of the AADEIM sampling strategy for deciding which parts of the large-scale states to sample to learn reduced-model updates. The analysis shows that the AADEIM sampling strategy is optimal up to a factor 2. Numerical results demonstrate the theoretical results by comparing the quasi-optimal AADEIM sampling strategy to other sampling strategies on various examples.

Keywords: Human-in-the-loop control, Adaptive control
Abstract: Communication and cooperation among team members can be enhanced significantly with physical interaction. Successful collaboration requires the integration of the individual partners' intentions into a shared action plan, which may involve a continuous negotiation of intentions and roles. This paper presents an adaptive haptic shared control framework wherein a human driver and an automation system are physically connected through a motorized steering wheel. By virtue of haptic feedback, the driver and automation system can monitor each other actions, and can still intuitively express their control intentions. The focus of this project is to develop a systematic model for an automation system that can vary its impedance such that its interaction with a human partner occurs as smoothly as that same interaction would between two humans. To this end, we defined a cost function that not only ensures the safety of the collaborative task but also takes account for the assistive behavior of the automation system. We employed a predictive controller based on modified non-negative least square to modulate the automation system impedance such that the cost function is optimized. The results demonstrate the significance of the proposed approach for negotiating the control authority, specifically when human and automation are in a non-cooperative mode. Furthermore, the performance of the adaptive haptic shared control is compared with the traditional haptic shared control paradigm. Finally, future experimental plan, its challenges, and our solution for those challenges are discussed.

Keywords: Neural networks, Adaptive control, Uncertain systems
Abstract: Employing the standard observer architecture to embed a neural network to estimate the states as well as uncertainty of a dynamic system, the modified state observer(MSO) is a technique that has found some successful applications in the engineering community, such as orbit uncertainty estimation problem, atmospheric reentry uncertainty estimation problem, and control design problem of nonlinear electrohydraulic system with parameter uncertainty. For implementation, however, it is desirable to have a discrete version that can be built into a microcontroller. In this paper, we formulate the discrete time version of the MSO, called the discrete time modified state observer (DMSO). Necessary mechanisms are developed using the Lyapunov theory. Finally, to prove the validity of the discrete time modified state observer, simulation studies are performed using a two wheeled inverted pendulum robot, a benchmark unstable nonlinear system.

Keywords: Neural networks, Machine learning, Learning
Abstract: We propose a novel mathematical semantic embedding for problem retrieval in adaptive tutoring. The goal is to retrieve problems with similar mathematical concepts. There are two challenges: First, problems conducive to tutoring are never exactly the same in terms of underlying concepts: those problems often mix concepts in innovative ways. Second, it is difficult for human to determine a consistent similarity score across a large enough training set. To address these two challenges, we develop a hierarchical problem embedding algorithm, Prob2Vec, which consists of abstraction and embedding steps. Prob2Vec is able to distinguish very fine-grained differences among problems, an ability humans need time and effort to acquire. In addition, the associated concept labeling is a multi-label problem with imbalanced training data set suffering from dimensionality explosion. Robust concept labeling is achieved with a novel negative pre-training algorithm that dramatically reduces false negative and positive ratios for classification. Experimental results show that Prob2Vec achieves 96.88% accuracy on a problem similarity test, in contrast to 75% from directly applying state-of-the-art sentence embedding methods.

Keywords: Adaptive systems, Uncertain systems, LMIs
Abstract: This paper focuses on a model reference adaptive control architecture for a class of uncertain systems with unmeasurable coupled dynamics. In particular, we propose a decoupling approach that yields asymptotic system error convergence; that is, asymptotic convergence between the trajectories of an uncertain system and a given reference model. The proposed approach reveals a tight sufficient condition for closed-loop system stability and it does not utilize any measurements from the coupled dynamics. We also present a numerical example for illustrating the presented approach.

Keywords: Autonomous systems, Evolutionary computing, Agents-based systems
Abstract: An unmanned surface vehicle takes on the role of an autonomous science agent, generating missions based on data. The vehicle explores over long durations and large regions, continually selecting and visiting target survey locations. Select system components are implemented and evaluated for automatic target generation, onboard mapping and path planning that optimizes energy use and data collection reward.

Keywords: Autonomous systems, Game theory, Learning
Abstract: This paper addresses the problem of tracking an actively evading target by employing a team of coordinating unmanned aerial vehicles while also learning the level of intelligence for appropriate countermeasures. Initially, under infinite cognitive resources, we formulate a game between the evader and the pursuing team, with an evader being the maximizing player and the pursuing team being the minimizing one. We derive optimal pursuing and evading policies while taking into account the physical constraints imposed by Dubins vehicles. Subsequently, we relax the infinite rationality assumption, via the use of level-k thinking. Such rationality policies are computed by using a reinforcement learning-based architecture and are proven to converge to the Nash policies as the thinking levels increase. Finally, simulation results verify the efficacy of the approach.

Keywords: Autonomous systems, Human-in-the-loop control, Markov processes
Abstract: Understanding human driver behaviors is crucial for an advanced driver assist system in the path control of a vehicle. In this study, an uncertain parameter in a human driver internal model is estimated using a Hidden Markov Model-based learning process. Based on the identified parameter, a Linear Quadratic Gaussian controller is designed for the vehicle to follow an online planned, optimal path, while reducing the row alignment error caused by the deviation of the driver internal model from the actual vehicle model, as well as sensor/actuator noise. Simulation results are used to show the effectiveness of the proposed controller.

Keywords: Autonomous systems, Machine learning
Abstract: We present an integrated approach for perception and control for an autonomous vehicle and demonstrate this approach in a high-fidelity urban driving simulator. Our approach first builds a model for the environment, then trains a policy exploiting the learned model to identify the action to take at each time-step. To build a model for the environment, we leverage several deep learning algorithms. To that end, first we train a variational autoencoder to encode the input image into an abstract latent representation. We then utilize a recurrent neural network to predict the latent representation of the next frame and handle temporal information. Finally, we utilize an evolutionary-based reinforcement learning algorithm to train a controller based on these latent representations to identify the action to take. We evaluate our approach in CARLA, a high-fidelity urban driving simulator, and conduct an extensive generalization study. Our results demonstrate that our approach outperforms several previously reported approaches in terms of the percentage of successfully completed episodes for a lane keeping task.

Keywords: Emerging control applications, Human-in-the-loop control, Autonomous systems
Abstract: Currently, studying the vehicle-human interactive behavior in the emergency needs a large amount of datasets in the actual emergent situations that are almost unavailable. Existing public data sources on autonomous vehicles (AVs) mainly focus either on the normal driving scenarios or on emergency situations without human involvement. To fill this gap and facilitate related research, this paper provides a new yet convenient way to extract the interactive behavior data (i.e., the trajectories of vehicles and humans) from actual accident videos that were captured by both the surveillance cameras and driving recorders. The main challenge for data extraction from real-time accident video lies in the fact that the recording cameras are un-calibrated and the angles of surveillance are unknown. The approach proposed in this paper employs image processing to obtain a new perspective which is different from the original video's perspective. Meanwhile, we manually detect and mark object feature points in each image frame. In order to acquire a gradient of reference ratios, a geometric model is implemented in the analysis of reference pixel value, and the feature points are then scaled to the object trajectory based on the gradient of ratios. The generated trajectories not only restore the object movements completely but also reflect changes in vehicle velocity and rotation based on the feature points distributions.

Keywords: Time-varying systems, Autonomous systems, Agents-based systems
Abstract: This paper formally expands the application domain of robot motion planning methods that are based on navigation functions to the case of moving obstacles. It generalizes the navigation function methodology from static sphere world environments, to dynamic ones. Specifically, it allows the obstacles' locations to be time-varying, albeit unknown, and accommodates the case where the navigation goal is not a single isolated point, but rather a spherical manifold. For such cases, the paper presents analytical bounds on the tuning parameters that guarantee the navigation function properties of the time-varying potential function, uniformly in time. Thus using the same choice of tuning parameters, the agent is ensured that at every instance in time, the artificial potential field that directs it to its destination is free of local minima. The parameter bounds naturally depend on the geometry of the agent workspace, and include conditions on how close the obstacles can approach each other, the fixed workspace boundary, and the destination sphere. The bounds presented here are conservative; their analytic determination serves mainly the purpose of theoretically guaranteeing completeness properties for the methodology in the time-varying obstacle case.

Keywords: Network analysis and control, Agents-based systems, Distributed control
Abstract: This paper presents a family of discrete-time distributed algorithms that enable nodes in an undirected, connected network to solve, in a fully decentralized fashion, a system of modular congruences whose residues and pairwise coprime moduli are locally known to the nodes. We show that each algorithm in the family is able to determine, in finite time, the congruence class of solutions whose existence and uniqueness is guaranteed by the Chinese remainder theorem. We also describe and analyze three specific algorithms from the family called Synchronous Updating (SU), Pairwise Equalizing (PE), and Groupwise Equalizing (GE), relating the convergence rate of SU to the network diameter and those of PE and GE to their asynchronous update patterns. Finally, we provide simulation results that illustrate their effectiveness.

Keywords: Network analysis and control, Control of networks, Cooperative control
Abstract: In this paper, we consider a network of agents with Laplacian dynamics, and study the problem of improving network robustness by adding maximum number of edges within the network while preserving a lower bound on its strong structural controllability (SSC). Edge augmentation increases network's robustness to noise and structural changes, however, it could also deteriorate network controllability. By exploiting relationship between network controllability and distances between nodes in graphs, we formulate an edge augmentation problem with a constraint to preserve distances between certain node pairs, which in turn guarantees that a lower bound on SSC is maintained even after adding edges. In this direction, first we choose a node pair and maximally add edges while maintaining the distance between selected nodes. We show that an optimal solution belongs to a certain class of graphs called clique chains. Then, we present and analyze two algorithms to add edges while preserving distances between a certain collection of nodes. Finally, we evaluate our results on various networks.

Keywords: Network analysis and control, Modeling, Identification
Abstract: Network representation provides a natural framework for the study of real world complex systems. In many cases, however, a faithful network representation that captures the interaction between individual components is not readily available. With enough data such interactions can be uncovered, but most real-world phenomena manifest in the form of a single cascade of infections or behaviors (more generally referred to as information) that percolates through the network. This problem can be represented as reconstructing a network where node states are visible as they reach a certain threshold. In this paper, we model an information cascade as the step response of a linear time invariant (LTI) directed network of nodes having first order dynamics. This simple representation allows us to solve for individual nodal parameters and the associated combinations of edges using data from a single perturbation. As expected, we obtain more than one valid network solutions that are able to recreate the response. We therefore evaluate the dependence of the number of valid network solutions on the amount of prior knowledge about node dynamics or connectivity. This is particularly relevant in situations where the experimenter may be able to generalize nodal dynamics or local topology within a network. Our results indicate that the number of solutions can be greatly reduced provided some knowledge of nodal parameters and network topology is available.

Keywords: Network analysis and control, Networked control systems
Abstract: In this paper we introduce a discrete time competing virus model and the assumptions necessary for the model to be well posed. We analyze the system exploring its different equilibria. We provide necessary and sufficient conditions for the estimation of the model parameters from time series data and introduce an online estimation algorithm. We employ a dataset of two competing subsidy programs from the US Department of Agriculture to validate the model by employing the identification techniques. To the best of our knowledge, this work is the first to study competing virus models in discrete-time, online identification of spread parameters from time series data, and validation of said models using real data. These new contributions are important for applications since real data is naturally sampled.

Keywords: Network analysis and control, Neural networks, Stability of nonlinear systems
Abstract: This paper analyzes the free recall dynamics of a working memory model. Free recalling is the reactivation of a stored pattern in the memory in the absence of the pattern. Our free recall model is based on an abstract model of a modular neural network composed on N modules, hypercolums, each of which is a bundle of minicolums. This paper considers a network of N modules, each consisting of two minicolumns, over a complete graph topology. We analyze the free recall dynamics assuming a constant, and homogeneous coupling between the network modules. We obtain a sufficient condition for synchronization of network’s minicolumns whose activities are positively correlated. Furthermore, for the synchronized network, the bifurcation analysis of one module is presented. This analysis gives a necessary condition for having a stable limit cycle as the attractor of each module. The latter implies recalling a stored pattern. Numerical results are provided to verify the theoretical analysis.

Keywords: Networked control systems, Communication networks, Linear systems
Abstract: The networking of Cyber-Physical Systems (CPS) is a worldwide trend. With the growing accessibility of data, vulnerability to cyber attacks increases. Cyber attacks on Industrial Control Systems (ICS) can have disastrous physical consequences, which motivate the study on cyber security of CPS. If the cyber attack remains undetected by the monitoring system, it is called stealthy. Stealthiness of the attack produces the risk that an adversary can take control over a CPS via a communication network and remain unnoticed. We aim to call attention to this problem by proposing a new attack scenario called local covert attack. Different from the previously known covert attack introduced in [1], the local covert attack proposed here needs less disruption resources and does not require the access to all control input signals and sensor output signals. We show that the local covert attack can be made completely stealthy by applying the decoupling technique or by combining the basic idea of covert attack and zero-dynamics attack. Simulation results are provided to demonstrate the stealthiness of the proposed local covert attacks.

Keywords: Estimation, Lyapunov methods, LMIs
Abstract: This paper investigates the scenario where a mobile sensor must observe (i.e., estimate the state of) another system, called the target. The estimation is affected by a distance-dependent noise, and for this reason we propose to control the sensor so that the effect of the noise is minimized. We propose a novel approach in which the controller and the observer are designed in tandem, with the common objective of obtaining a better estimation. We give sufficient design conditions for a general class of nonlinear systems satisfying a Lipschitz-like condition on the nonlinearity. A numerical example illustrates the obtained results.

Keywords: Estimation, Neural networks, Kalman filtering
Abstract: In this paper, we consider the task of estimating the state of dynamic object by applying an unscented filter to pose estimates generated by a neural network. To incorporate the rotational state of the system into the filter, we use a parameterization of the tangent space of the group of rotation matrices SO(3). We then characterize the noise in the pose estimation neural network by considering simple motions of the object, as well as using a Monte Carlo approach. Finally, using synthetically generated images, we show in simulation how the unscented filter can improve the accuracy of the pose estimates from the neural network.

Keywords: Estimation, Neural networks, Sensor networks
Abstract: This paper proposes a novel adaptive neural network (NN) based distributed state estimation scheme for a heterogeneous sensor network (HSN), to estimate the state vector of an unknown nonlinear process/target by using sensed output when the target input remains unknown. The active nodes in the HSN can sense the target output based on the detection range. By using a connected graph, the active nodes will communicate their estimated state vector from their adaptive NN observer to other passive nodes in the neighborhood that cannot sense the target, so that they can estimate the target state vector. Next, a subset of nodes in the HSN, referred to as the mobile nodes, track the moving target by using their estimated state information and a state feedback controller. For the communication topology considered, it is shown that the distributed state estimation, the NN observer weight estimation, and the tracking errors are uniformly ultimately bounded. Simulation results verify the theoretical claims.

Keywords: Estimation, Sensor fusion, LMIs
Abstract: Among algorithms used for sensor fusion for attitude estimation in unmanned aerial vehicles, the Extended Kalman Filter(EKF) is the most commonly used for estimation. In this paper, we propose a new version of H2 estimation called extended H2 estimation that can overcome the limitations of the extended Kalman Filter, specifically with respect to computational speed, memory usage, and root mean squared error. We formulate a new attitude estimation algorithm, where the filter gain is designed offline about a nominal operating point, but the filter dynamics is implemented using the nonlinear system dynamics. We refer to this implementation of the H2 optimal estimator as the extended H2 estimator. The solution presented is tested on two cases, corresponding to slow and rapid motions, and compared against the EKF in the performance metrics mentioned above.

Keywords: Estimation, Simulation
Abstract: Maximum likelihood estimations (MLEs) and corresponding confidence intervals are commonly used in statistical inference. In practice, people usually construct approximate confidence intervals with the Fisher information at given sample data based on the asymptotic normal distribution of MLE. Two common Fisher information numbers (FINs, for scalar parameters) are the observed FIN (the second derivative of negative log-likelihood function) and the expected FIN (the expectation of the observed FIN). In this article, we prove that under certain conditions and with MSE criterion, approximate confidence intervals with the expected FIN are more accurate than those with the observed FIN. The fact is illustrated in a numerical study related to a standard signal-plus-noise problem.

Keywords: Estimation, Time-varying systems, Observers for nonlinear systems
Abstract: The aim of this paper is to cope with estimation issues of discrete-time nonlinear time-varying systems with input and output. Inspired by [12], a new design technique of fixed-time observers is proposed. It relies on the use of past values of the output and the theory of the monotone systems to construct dead bit observer or fixed-time interval estimator depending on the absence or the presence of uncertainties. Finally, simulations are conducted to verify the effectiveness of the proposed schemes.

Keywords: Control applications, Mechatronics, Robust control
Abstract: In this paper, a robotic manipulator is developed to allow the appropriate exertion of the coupling force needed for proper implementation of Smart-touch inspection. A force feedback control system works in tandem with a load cell sensor and DC motor, which allows the manipulator to guide two fingers to grasp a bolted connection for Smart-touch inspection. The envisioned device can be mounted onto a remotely operated vehicle (ROV) and eventually a fully autonomous vehicle. The load-cell on one of the fingers senses the touching force, which provides force feedback information for the DC motor on the joint of the robotic manipulator. To reject the system's uncertainties and nonlinearities, such as dead-zone, backlash and sensing noise, a reduced-order active disturbance rejection control (R-ADRC) is developed for the robotic manipulator to maintain its grasping force at the inspection location. Experimental results have shown that the R-ADRC can actively reject the system's uncertainties and work better than proportional-integral-derivative (PID) control in terms of less steady-state error and overshoot.

Keywords: Control applications, Power systems, Robust control
Abstract: In this paper, a robust controller is developed for an automatic voltage regulator (AVR) in power systems. In reality, the voltage at each substation fluctuates constantly due to disturbances. These voltage disturbances are mainly caused by load changes and loss of equipment. They degrade voltage/power quality, and if not corrected in time could lead to voltage collapse and black out. The AVR at generating stations is a crucial system to maintain the magnitude of terminal voltage at nominal value. But AVR itself is not robust enough against the disturbances. Therefore an additional controller is needed to improve the dynamic performance of the AVR and maintain the nominal voltage at each bus. An error driven active disturbance rejection controller (EDADRC) is originally designed and applied to enhance the performance of the AVR. In EDADRC, a feedback controller is constructed based on the difference between nominal and actual voltages. An observer is used to estimate the external disturbance, which is then compensated by the controller. The stability of EDADRC is theoretically proved. Both PID and EDADRC are implemented on a real time Simscape Electrical platform with actual power equipment. Simulation results demonstrate the superiority of the EDADRC to PID controller in disturbance rejection with comparable control efforts. They also verify the effectiveness of EDADRC in voltage regulation.

Keywords: Control applications, Robust control, Robotics
Abstract: In this article, the problem of designing Active Disturbance Rejection Control (ADRC) for a class of second-order mechanical systems, expressed with Euler-Lagrange equations, is studied. A specific and practically motivated case is considered here, namely trajectory tracking task without the use of signal time-derivatives in the tracking controller. A general solution is proposed showing how to synthesize and tune the observer and the controller parts of the ADRC scheme. A special Extended State Observer is used in the design and here it takes the form of a Generalized Proportional Integral Observer (GPIO), which uses a Taylor series approximation of the total disturbance. A set of experimental results, obtained using a two degree-of-freedom robotic manipulator, shows the effectiveness of the proposed governing scheme in terms of trajectory realization and disturbance rejection without the use of signal time-derivatives in the controller.

Keywords: Randomized algorithms, Robust control
Abstract: We present two data-driven methods for estimating reachable sets with probabilistic guarantees. Both methods make use of a probabilistic formulation allowing for a formal definition of a data-driven reachable set approximation that is correct in a probabilistic sense. The first method recasts the reachability problem as a binary classification problem, using a Gaussian process classifier to represent the reachable set. The quantified uncertainty of the Gaussian process model allows for an adaptive approach to the selection of new sample points. The second method uses a Monte Carlo sampling approach to compute an interval-based approximation of the reachable set. This method comes with a guarantee of probabilistic correctness, and an explicit bound on the number of sample points needed to achieve a desired accuracy and confidence. Each method is illustrated with a numerical example.

Keywords: Transportation networks, Multivehicle systems, Robust control
Abstract: A station-based vehicle sharing system consists of a fleet of vehicles (usually bikes or cars) that can be rented at one station and returned at another station. We study how to achieve guaranteed service availability in such systems. Specifically, we are interested in determining a) the fleet size and b) a vehicle rebalancing policy that guarantees that a) every customer will find an available vehicle at the origin station and b) the customer will find a free parking spot at the destination station. We model the evolution of the number of vehicles at each station as a stochastic process. The proposed rebalancing strategy iteratively solves a chance-constrained optimization problem to find a rebalancing schedule that ensures that no service failures will occur in future with a given level of confidence. We show that such a chance-constrained optimization problem can be converted into a linear program and efficiently solved. As a case study, we apply the proposed method to control a simulated bike-sharing system in Boston using real-world historical demand. Our results demonstrate that our method can indeed ensure the desired level of service availability even when the demand does not fully conform to the assumptions of the underlying stochastic model. Moreover, compared with a state-of-the art rebalancing method, the proposed method is able to achieve nearly full service availability while making less than half of the rebalancing trips.

Keywords: Uncertain systems, Robust control, H-infinity control
Abstract: This paper considers the construction of worst-case perturbations for uncertain systems. The uncertain system is modeled as an interconnection of a linear time-invariant (LTI) system and a norm-bounded LTI uncertainty. The worst-case gain (measured in the H-infinity norm) for the uncertain system can be assessed via skewed-mu analysis. The standard approach is to compute upper and lower bounds for the worst-case gain on a frequency grid. A worst-case LTI perturbation is then constructed to maximize the gain at a single frequency. This perturbation can be used within a high fidelity nonlinear simulation to further explore system robustness. A drawback of this existing approach is that the worst-case perturbation constructed at a single frequency may not necessarily induce poor time-domain performance. It is beneficial to construct a perturbation that maximizes the gain at multiple frequency points, e.g. where the system is most sensitive or where disturbances have large frequency content. This paper provides an algorithm to construct a single perturbation which causes the uncertain system to achieve its largest possible gain at multiple chosen frequency points. This is achieved by interpolating through worst-case samples at the individual frequencies using the boundary Nevanlinna-Pick interpolation. Simple numerical examples are provided to demonstrate the proposed approach.

Keywords: Distributed parameter systems, Delay systems, Estimation
Abstract: We provide a method for gradient-based multivariable extremum seeking that allows different delays in each input channel. We allow output delays, and for the delays to be arbitrarily long. We use averaging and a perturbation based estimate, but our sequential predictor based approach provides a useful alternative to the distributed terms that were used in earlier delay compensation approaches to extremum seeking. We illustrate our approach in a source seeking example.

Keywords: Distributed parameter systems, Distributed control
Abstract: This paper is motivated by economic aspects of fixed initial and operating costs for control of spatially distributed systems. In particular, the paper investigates the possibility of a large number of inexpensive actuating and sensing devices, as an alternative to (a reduced number of) expensive high capacity devices. While such an alternative reduces the fixed initial costs associated with actuators and sensors, it may also lead to increased operating costs resulting from communication requirements between the now-networked actuator-sensor-control units. To simplify the controller architecture, a proportional controller is assumed that amounts to a static output feedback controller. In a network of n actuator-sensor pairs, an all-to-all communication topology results in a fully populated static output feedback matrix with as much as n(n-1) communication links. In addition to a traditional performance index used to obtain the static output feedback gain matrix, this paper proposes a mixed index wherein both the traditional performance index and the number of communication links (representing operating costs associated with information exchange links), are taken into account. As an example, the proposed scheme is applied to a parabolic partial differential equation having four actuator-sensor pairs. The resulting optimization produces a sparse static gain matrix with a communication topology that has half the graph edges of the fully connected case and with essentially the same performance.

Keywords: Distributed parameter systems, Lyapunov methods, Output regulation
Abstract: This paper presents an observer and an output feedback control with respect to a reference solution for the one-phase Stefan problem under input hysteresis. The one-phase Stefan problem describes evolution of the temperature and melting-solidification front in liquid-solid material. The setting models an industrial casting process, and experiments have revealed the existence of hysteresis due to boiling of the cooling water at the surface of the casting process. Therefore, one-phase Stefan problem with water cooling hysteresis under Neumann boundary actuation is considered. Full state feedback control law for this problem was designed and proved to provide asymptotic convergence of both temperature distribution and the solidification front. However, for the casting process, only boundary sensing is available. To address the latter problem, the present paper proposes an observer that estimates the temperature profile based on the available surface temperature measurement, taking into account boundary input hysteresis. The stability of the observer is proved with Lypanov method. Finally, an output feedback control law is proposed and proved to ensure asymptotic convergence of the temperature and the solidification front errors to zero. A numerical example presents the application of the method proposed.

Keywords: Computational methods, Distributed parameter systems, Control software
Abstract: In this paper, we present PIETOOLS, a MATLAB toolbox for the construction and handling of Partial Integral (PI) operators. The toolbox introduces a new class of MATLAB object, opvar, for which standard MATLAB matrix operation syntax (e.g. +, *, ' etc.) is defined. PI operators are a generalization of bounded linear operators on infinite-dimensional spaces that form a *-subalgebra with two binary operations (addition and composition) on the space Rtimes L_2. These operators frequently appear in analysis and control of infinite-dimensional systems such as Partial Differential Equations (PDE) and Time-delay systems (TDS). Furthermore, PIETOOLS can: declare opvar decision variables, add operator positivity constraints, declare an objective function, and solve the resulting optimization problem using a syntax similar to the sdpvar class in YALMIP. Use of the resulting Linear Operator Inequalities (LOI) are demonstrated on several examples, including stability analysis of a PDE, bounding operator norms, and verifying integral inequalities. The result is that PIETOOLS, packaged with SOSTOOLS and MULTIPOLY, offers a scalable, user-friendly and computationally efficient toolbox for parsing, performing algebraic operations, setting up and solving convex optimization problems on PI operators.

Keywords: Distributed parameter systems
Abstract: The work provides a general model of communication attacks on a networked infinite dimensional system. The system employs a network of inexpensive control units consisting of actuators, sensors and control processors. In an effort to replace a reduced number of expensive high-end actuating and sensing devices implementing an observer-based feedback, the alternate is to use multiple inexpensive actuators/sensors with static output feedback. In order to emulate the performance of the high-end devices, the controllers for the multiple actuator/sensors implement controllers which render the system networked. In doing so, they become prone to communication attacks either as accidental or deliberate actions on the connectivity of the control nodes. A single attack function is proposed which models all types of communication attacks and an adaptive detection scheme is proposed in order to (i) detect the presence of an attack, (ii) diagnose the attack and (iii) accommodate the attack via an appropriate control reconfiguration. The reconfiguration employs the adaptive estimates of the controller gains and restructure the controller adaptively in order to minimize the detrimental effects of the attack on closed-loop performance. Numerical studies on a 1D diffusion PDE employing networked actuator/sensor pairs are included in order to further convey the special architecture of detection and accommodation of networked systems under communication attacks.

Keywords: Distributed parameter systems, Learning, Autonomous robots
Abstract: In this paper we propose an online active sensor planning strategy for model-based acoustic source identification (SI) in non-convex domains utilizing the three-dimensional Helmholtz partial differential equation (PDE). After discretizing the PDE using the finite element method, we formulate the SI problem as a PDE-constrained optimization problem. To make the solution computationally tractable, we employ proper orthogonal decomposition to reduce the dimension of the pressure field. Given a set of initial measurements, we solve the SI and sensor planning problems in a feedback loop. Specifically, given a set of measurements, we first solve the SI problem to get an estimate of the source field. We then fit a set of nonlinear basis functions to the solution in order to reduce the number of unknowns required to describe the source field. We finally utilize the Fisher information matrix (FIM) along with the current source parameter estimates to select the next best measurement location. Specifically, we choose a sequence of waypoints that sequentially maximize the minimum eigenvalue of the FIM with respect to the unknown source parameters. We present numerical and experimental results that showcase our proposed method. This work presents the first active sensor planning method for PDE-based acoustic SI that is investigated in practice.

Keywords: Distributed parameter systems, Nonlinear output feedback, Process Control
Abstract: We focus on Lyapunov-based output feedback control for a class of distributed parameter systems with spatiotemporal dynamics described by input-affine semilinear dissipative partial differential equations (DPDEs). The control problem is addressed via adaptive model order reduction. Galerkin projection is applied to discretize the DPDE and derive low-dimensional reduced order models (ROMs). The empirical basis functions needed for this discretization are updated using adaptive proper orthogonal decomposition (APOD) which needs measurements of the complete profile of the system state (called snapshots) at revision times. The main objective of this paper is to minimize the demand for snapshots from the spatially distributed sensors by the control structure while maintaining closed-loop stability and performance. A control Lyapunov function is defined and its value is monitored as the system evolves. Only when the value violates a closed-loop stability threshold, snapshots are requested for a brief period by APOD after which the ROM is updated and the controller is reconfigured. The proposed approach is applied to stabilize the Kuramoto-Sivashinsky equation.

Keywords: Nonlinear output feedback, Estimation, Electrical machine control
Abstract: We show that for PWM-operated devices, it is possible to benefit from signal injection without an external probing signal, by suitably using the excitation provided by the PWM itself. As in the usual signal injection framework conceptualized in [1], an extra “virtual measurement” can be made available for use in a control law, but without the practical drawbacks caused by an external signal.

Keywords: Modeling, Nonlinear output feedback, Stability of nonlinear systems
Abstract: Self-excited systems arise in many applications,such as biochemical systems, mechanical systems with fluid-structure interaction, and fuel-driven systems with combustion dynamics. This paper presents a Lur’e model that exhibits biased oscillations under constant inputs. The model involves arbitrary asymptotically stable linear dynamics, time delay, a washout filter, and a saturation nonlinearity. For all sufficiently large scalings of the loop transfer function, these components cause divergence under small signal levels and decay under large signal amplitudes, thus producing an oscillatory response.A bias-generation mechanism is used to specify the mean of the oscillation. The main contribution of the paper is analysis of a discrete-time version of this model.

Keywords: Nonlinear output feedback, Output regulation, Delay systems
Abstract: A prescribed-time output-feedback stabilizing controller is designed for a class of uncertain nonlinear strict-feedback-like systems with unknown time delays. The control design is based on a novel adaptation of the dual dynamic high-gain scaling based observer and controller design methodology wherein a Lyapunov-Krasovskii functional is first used to show existence of closed-loop solutions over the prescribed time interval. Thereafter, dynamics of an adaptation state variable and high-gain scaling parameter are used to show that unknown parameters are dominated within a sub-interval of the prescribed time after which a formulation of a Lyapunov function dynamics as a scalar delay differential equation is introduced to establish exponential convergence within the remaining sub-interval of the prescribed time.

Keywords: Stability of nonlinear systems, Nonlinear output feedback, Lyapunov methods
Abstract: Global asymptotic tracking by output feedback is studied for a class of nonminimum-phase nonlinear systems in output feedback form. It is proved that the tracking problem is solvable by an n-dimensional output feedback controller under the two conditions: 1) the nonminimum-phase nonlinear system can be rendered minimum-phase by a virtual output; 2) the internal dynamics of the nonlinear system driven by a desired signal and its derivatives has a bounded solution trajectory. With the help of a new coordinate transformation, a constructive method is presented for the design of a dynamic output tracking controller. An example is given to validate the proposed output feedback tracking control scheme.

Keywords: Nonlinear output feedback, Optimal control, Lyapunov methods
Abstract: We consider the optimal control design of a class of affine nonlinear systems whose right-hand sides are analytic functions. We build upon ideas from Carleman linearization, which is a nonlinear procedure to transform (lift) a finite-dimensional nonlinear system into an infinite-dimensional linear system with no loss, and lift a Hamilton-Jacobi-Bellman (HJB) equation into an infinite-dimensional quadratic form that resembles the familiar algebraic Riccati equation. Then, we propose an efficient method to calculate the solution of the resulting infinite-dimensional equation using one algebraic Riccati equation (of the same dimension as the original nonlinear system) and a series of linear matrix equations in an iterative manner. One can obtain arbitrarily near-optimal solutions using finite truncations. It is shown that the resulting approximate solutions are symmetric. Using these approximate solutions to the HJB equation, we construct approximate optimal control laws. Our simulation results assert that our method enjoys high accuracy in comparison to the actual optimal feedback control laws and the accuracy increases as higher-order truncations are used.

Keywords: Cooperative control, Optimization algorithms, Agents-based systems
Abstract: We address the problem of multiple local optima arising in cooperative multi-agent optimization problems with non-convex objective functions. We propose a systematic approach to escape these local optima using boosting functions. These functions temporarily transform a gradient at a local optimum into a "boosted" non-zero gradient. Extending a prior centralized optimization approach, we develop a distributed framework for the use of boosted gradients and show that convergence of this distributed process can be attained by employing an optimal variable step size scheme for gradient-based algorithms. Numerical examples are included to show how the performance of a class of multi-agent optimization systems can be improved.

Keywords: Iterative learning control, Cooperative control
Abstract: Collaborative tracking control and formation control are common approaches in which multiple agents work together to perform a global objective. They are increasingly used in a diverse range of applications, however few controllers simultaneously address both tasks. To improve performance of repeated tasks, iterative learning control (ILC) has been independently applied to both methodologies. However, focus has been on centralized structures, and existing solutions typically have limited convergence rates and robustness properties.

This paper addresses these limitations by developing a powerful decentralised ILC framework that unites both collaborative tracking and formation control objectives. It enables broad classes of ILC algorithm to be derived with well-defined convergence rates, optimal tracking solutions, and transparent robustness properties. The framework is illustrated through derivation of three new ILC updates: inverse, gradient and norm optimal ILC. Convergence analysis for the proposed framework is also given.

Keywords: Cooperative control, Robotics, Adaptive control
Abstract: In this paper we present a novel adaptive cooperative manipulation controller for multiple mobile robots with rolling contacts. Our approach exploits rolling effects of passive end-effectors and does not require force/torque sensing. Moreover, the proposed scheme is robust to uncertain dynamics of the object and agents including object center of mass, inertia, weight, and Coriolis terms. In addition, we present a novel closed-form internal force controller that guarantees no slip throughout the manipulation task. The adaptive controller design ensures boundedness of the estimated model parameters in predefined sets. Numerical simulations validate the effectiveness of the proposed approach.

Keywords: Cooperative control, Stability of nonlinear systems
Abstract: In this paper, we investigate the almost output consensus problem for nonlinear multi-agent systems under the influence of external disturbances. The concept of almost output consensus of nonlinear multi-agent systems is defined. Conditions on the nonlinear systems are established under which distributed consensus protocols are designed in a recursive manner. These protocols are shown to achieve almost output consensus, that is, output consensus of the system is achieved in the absence of disturbances and the L₂-gain from the disturbances to the output consensus error of agents with the same initial condition can be made arbitrarily small. A numerical example is shown to verify the theoretical results.

Keywords: Cooperative control
Abstract: The concept of reach-avoid (RA) games via coverage control is generalized to players with nonlinear dynamics and in arbitrary dimensions. Pursuer coordination on defense surfaces is formally shown sufficient as a cooperative strategy for RA games in any dimensions. Nonlinear control synthesis strategies with convergence guarantees are provided to enforce coverage on said surfaces. The effectiveness of two coverage control formulations is verified through simulation.