Keywords: Control Technology, MEMS and nanotechnology, Manufacturing systems
Abstract: The way Extreme Ultraviolet (EUV) Light is generated is nothing short of amazing: Tin droplets about 30 micrometer in diameter are delivered into a vacuum chamber while an advanced tracking system ensures that two CO2 lasers consistently hit each droplet with sub-micrometer accuracy. First, to expand the droplet into a disk-like shape to then be blasted again by a second even more powerful laser pulse to generate EUV light with a wavelength of just 13.5nm. And this process is repeated again and again at a rate of about 50,000 times per second. None of this would be possible without the deployment of numerous control loops as well as automation and setup routines to ensure position and timing accuracies on the order of nanometers and nanoseconds. In her plenary session, Liane will share her insights into the transformative role of automation and controls used in EUV Light Sources used for producing the world’s most advanced microchips, offering a glimpse into the complexity ASML’s EUV Light Sources that are pushing the boundary of real-time precision and reliability.

Keywords: Iterative learning control, Autonomous systems, Automotive applications
Abstract: Shared control, which merges the dynamic inputs of human drivers with vehicle automation, has attracted considerable attention due to its potential to enhance both safety and driver satisfaction. However, most existing shared control strategies are based on one-size-fits-all designs, neglecting the fact that the optimal level of sharing will typically depend on the individual driver and road characteristics. In light of these limitations, along with the observation that routes are commonly repeated, we propose an iterative learning control algorithm that divides a closed-loop driving circuit into discrete segments, enabling online estimation of the driver’s performance over each segment. By adapting arbitration weights based on these driver-specific and segment-specific estimates, our method seeks to reduce lateral tracking errors, reduce path completion time, and increase secondary task performance. To validate this approach, we conducted driver-in-the-loop simulator tests with 20 participants, each driving repeatedly on the closed-loop circuit. The results demonstrate that our personalized strategy significantly improves driving performance.

Keywords: Optimization, Process control, Identification
Abstract: Feedback optimization has emerged as an effective strategy for steady-state optimization of dynamical systems. By exploiting models of the steady-state input-output sensitivity, methods of this type are often sample efficient, and their use of feedback ensures that they are robust against model error. Still, this robustness has its limitations, and the dependence on a model may hinder convergence in settings with high model error. We investigate here the effect of a particular type of model error: bias due to identifying the model from closed-loop data. Our main results are a sufficient convergence condition, and a converse divergence condition. The convergence condition requires a matrix which depends on the closed-loop sensitivity and a noise-to-signal ratio of the data generating system to be positive definite. The negative definiteness of the same matrix characterizes an extreme case where the bias due to closed-loop data results in divergence of model-based feedback optimization.

Keywords: Sensors, Adaptive control, Estimation
Abstract: In this paper we propose a framework towards achieving two intertwined objectives: (i) equipping reinforcement learning with active exploration and deliberate information gathering, such that it regulates state and parameter uncertainties resulting from modeling mismatches and noisy sensory; and (ii) overcoming the computational intractability of stochastic optimal control. We approach both objectives by using reinforcement learning to compute the stochastic optimal control law. On one hand, we avoid the curse of dimensionality prohibiting the direct solution of the stochastic dynamic programming equation. On the other hand, the resulting stochastic optimal control reinforcement learning agent admits caution and probing, that is, optimal online exploration and exploitation. Unlike fixed exploration and exploitation balance, caution and probing are employed automatically by the controller in real-time, even after the learning process is terminated. We conclude the paper with a numerical simulation, illustrating how a Linear Quadratic Regulator with the certainty equivalence assumption may lead to poor performance and filter divergence, while our proposed approach is stabilizing, of an acceptable performance, and computationally convenient.

Keywords: Adaptive control, Semidefinite programming, Aerospace applications
Abstract: This paper presents a novel data-driven method for the synthesis of linear parameter-varying (LPV) controllers aimed at adaptive disturbance rejection. The approach leverages frequency-domain input/output response data from a linear time-invariant (LTI) multiple-input multiple-output (MIMO) system, eliminating the need for a parametric model. The designed LPV controller guarantees system stability even under arbitrarily fast variations of the scheduling parameter corresponding to the estimated harmonic disturbance frequencies. Control design is carried out in the frequency domain using performance constraints at selected operating points representing stationary disturbance frequencies. Then, Integral Quadratic Constraints (IQC) are employed to analyse the closed-loop stability under scheduling parameter variations. The IQC-based algorithm also determines the admissible range of the scheduling parameter and can incorporate upper bounds on its variation rate to reflect physical system limitations. The method is experimentally validated on a hybrid microvibration damping (MIVIDA) platform for space applications. An LPV controller is designed and implemented to reject unknown time-varying harmonic disturbances. Experimental results demonstrate the effectiveness of the approach in achieving robust disturbance rejection and closed-loop stability.

Keywords: Optimization, Process control
Abstract: This paper develops a persistently exciting input generating Online Feedback Optimization (OFO) controller that estimates the sensitivity of a process ensuring minimal deviations from the descent direction while converging. This eliminates the need for random perturbations in feedback loop. The proposed controller is formulated as a bilevel optimization program, where a nonconvex full rank constraint is relaxed using linear constraints and penalization. The validation of the method is performed in a simulated scenario where multiple systems share a limited, costly resource for production optimization, simulating an oil and gas resource allocation problem. The method allows for less input perturbations while accurately estimating gradients, allowing faster convergence when the gradients are unknown. In the case study, the proposed method achieved the same profit compared to an OFO controller with random input perturbations, and 1.4% higher profit compared to an OFO controller without input perturbations.

Keywords: Predictive control, Robust control, Automotive applications
Abstract: The key to vehicle collision avoidance is achieving optimal avoidance performance with a reasonable computational load for real-time applications. To address these requirements, this study applies a novel approach by designing a robust tube nonlinear model predictive controller (RTNMPC) and approximating it to a neural network, thereby ensuring both optimal collision avoidance performance and real-time capability. The RTNMPC optimally controls the vehicle's steering and differential braking forces to guide it to a safe lane, minimizing the avoidance trajectory area. Tightened tire grip constraints were applied to robustly maintain vehicle maneuverability under system uncertainties and approximation errors in the neural network controller. Grip constraints were further relaxed by introducing a practical constraint tightening approach with an input saturation process based on tire grip usage. Consequently, the proposed collision avoidance system achieved both greater collision avoidance results with the lowest computational load compared to the baselines in CarSim simulations.

Keywords: Automotive applications, Optimization, Autonomous systems
Abstract: This study proposes a computationally efficient trajectory planning algorithm for emergency collision avoidance in high center of gravity (CoG) vehicles, building upon the previously developed Dual Connecting Points Optimization (DCPO) framework. While existing methods often suffer from computational inefficiency or limited avoidance performance, the DCPO approach effectively utilizes lateral acceleration up to the rollover threshold, enabling safe and responsive maneuvers for high CoG vehicles. By employing exponential lateral acceleration profiles and only two optimization variables, the algorithm achieves both rapid lane changes and real-time feasibility. This work enhances the prior DCPO formulation by redesigning the cost function to adaptively adjust trajectories based on obstacle proximity—producing smoother paths for moderate threats and sharper maneuvers for imminent ones. Simulation results confirm the proposed method’s effectiveness across diverse emergency scenarios.

Keywords: Automotive applications
Abstract: A variety of control functions are used in modern vehicles to stabilize the vehicle dynamics. These can be improved with precise information of time-varying parameters. The vehicle mass, center of gravity height and the roll moment of inertia are significant for vehicle roll dynamics as they characterize the static and dynamic behavior of the roll motion. These parameters not only affect the driving behavior but can also increase the risk of rollover. Since the roll moment of inertia is particularly significant for transient roll dynamics, the main contribution of this paper is to develop a model-based estimation algorithm to estimate the roll moment of inertia. This paper therefore presents an Unscented Kalman Filter for a simultaneous state and parameter estimation. By using a nonlinear vehicle model which represents the roll, pitch and vertical dynamics, the effects of the center of gravity height and additional masses on the inertia are taken into account. In order to improve the estimation results, an activation condition based on a linear single track model and an underlying observability analysis is presented. Based on that, a precise parameter estimation with a deviation of less than 5% to the nominal parameter is achieved.

Keywords: Automotive applications, Fault detection/accomodation, Kalman filtering
Abstract: Detecting sensor faults in automotive systems is vital to ensure the safe operation and reliability of vehicle controllers. The aim is to detect faults reliably and quickly while avoiding false positives. In particular, the road parameters in the form of banking angle and friction coefficient strongly influence the fault estimation. This work presents a framework for the reliable detection of sensor faults in the vehicle’s Inertial Measurement Unit (IMU) up to the driving dynamics limit range. The unknown road parameters are estimated in parallel and enable the use of the proposed method under varying environmental conditions. The estimation of the vehicle states, the coefficient of friction and the sensor faults is carried out using a Two-Stage Unscented Kalman Filter with a nonlinear vehicle model. The estimation of the road inclination and banking angle is carried out using a separate Extended Kalman Filter. Measurements with an Audi SQ8 e-tron under high and low friction condition are presented and show a fault detection time of 100 ms while avoiding false-positive detections for quasi steady state friction conditions. Rapid variations in the coefficient of friction lead to a slight degradation in fault estimation performance.

Keywords: Automotive applications, Predictive control, Control applications
Abstract: This paper introduces a Time Shift Governor (TSG)-guided Model Predictive Controller with Control Barrier Functions (CBFs)-based constraints for adaptive cruise control (ACC). This MPC-CBF approach is defined for obstacle-free curved road tracking, while following distance and obstacle avoidance constraints are handled using standard CBFs and relaxed Collision Cone CBFs. In order to address scenarios involving rapidly moving obstacles or rapidly changing leading vehicle's behavior, the TSG augmentation is employed which alters the target reference to enforce constraints. Simulation results demonstrate the effectiveness of the TSG-guided MPC-CBF approach.

Keywords: Mobile Robots, Autonomous systems, Robotics applications
Abstract: This paper presents ZeloS, a research platform designed and built for practical validation of automated driving methods in an early stage of research. We overview ZeloS’ hardware setup and automation architecture and focus on motion planning and control. ZeloS weighs 69 kg, measures a length of 117 cm, and is equipped with all-wheel steering, all-wheel drive, and various onboard sensors for localization. The hardware setup and the automation architecture of ZeloS are designed and built with a focus on modularity and the goal of being simple yet effective. The modular design allows the modification of individual automation modules without the need for extensive onboarding into the automation architecture. As such, this design supports ZeloS in being a versatile research platform for validating various automated driving methods. The motion planning component and control of ZeloS feature optimization-based methods that allow for explicitly considering constraints. We demonstrate the hardware and automation setup by presenting experimental data.

Keywords: Estimation, Energy Storage, Machine learning
Abstract: This paper presents a novel Graph Neural Network (GNN)-based surrogate model for predicting state evolution in reconfigurable battery packs. By leveraging graph-based representations of battery cell interconnections, the proposed approach addresses the unique challenge of estimating the imbalance in state-of-charge (SOC) and temperature of cells of a battery pack in dynamic battery configurations. Unlike conventional methods that focus on instantaneous state estimation, our GNN model predicts future SOC and temperature distributions by considering both current system state and switch configuration. The model architecture combines Graph Attention Networks with pooling operations to effectively capture cell-to-cell interactions and battery pack-level dynamics. Numerical results demonstrate that our GNN-based approach significantly outperforms baseline Feedforward Neural Network (FNN) and FNN-attention models, showing substantial improvements in prediction accuracy for both temperature and SOC while maintaining robust performance even with limited training data.

Keywords: Identification, Energy Storage, Estimation
Abstract: We address the problem of estimating model parameters of a battery without prior knowledge of the battery’s specifications, using only operational data. Many existing studies on parameter estimation for batteries use prior knowledge of the battery’s specifications, such as typical values of model parameters and the nonlinear relationship between the state-of-charge (SOC) and open-circuit voltage (OCV). However, there are cases where battery’s specifications are completely unknown, but battery modeling is still needed, such as in model-based design and condition monitoring of battery-powered products. To achieve that, we propose a parameter identification technique using the Variable Projection-based Output Error Method for the battery state-space model. Our proposed method estimates the parameters of a polynomial OCV function with respect to cumulative current, as well as the parameters of an equivalent circuit model (ECM), solely from operational data. Moreover, with the aid of Variable Projection, the number of actual parameters to be estimated is significantly reduced, enabling automatic parameter estimation. The performance of the proposed method is validated using both simulated and operational data obtained from a test drive of an electric vehicle (EV).

Keywords: Energy Storage, Reinforcement learning, Control applications
Abstract: Accurate parameter estimation in electrochemical battery models is essential for monitoring and assessing the performance of lithium-ion batteries (LiBs). This paper presents a novel approach that combines deep reinforcement learning (DRL) with an optimal experimental design (OED) framework to identify key electrochemical parameters of LiB cell models. The proposed method utilizes the twin delayed deep deterministic policy gradient (TD3) algorithm to optimize input excitation, thereby increasing the sensitivity of the system’s response to electrochemical parameters. The performance of this DRL-based approach is evaluated against a nonlinear model predictive control (NMPC) method and conventional tests. Results indicate that the DRL-based method provides superior information content, reflected in higher Fisher information (FI) values and lower parameter estimation errors compared to the NMPC design and conventional test practices. Additionally, the DRL approach offers a substantial reduction in experimental time and computational resources.

Keywords: Energy Storage, Estimation, Modeling
Abstract: This paper examines the lithium-sulfur (Li-S) battery state estimation problem. Li-S batteries promise to deliver higher energy storage capacities per unit mass compared to today’s lithium-ion batteries. This is particularly attractive for applications such as aviation electrification. Unfortunately, state estimation for Li-S batteries is quite challenging due to the high order and nonlinearity of the underlying physics. Previous research builds Li-S battery models, parameterizes them from experimental data, and utilizes them for state estimation. Model complexity is visible as a challenge throughout this literature, including the fact that many existing Li-S battery models take the form of differential algebraic equations (DAEs). We address this complexity challenge by making three contributions to the literature. First, we employ model causality inversion to translate an existing Li-S battery model in the literature from DAEs to explicit ordinary differential equations (ODEs). Second, we show that the resulting 7th order ODE model contains two redundant state variables, thereby reducing it to a 5th order model. Finally, we exploit the structure of the model to construct a simple switching state estimation law. Simulation studies show that this enables accurate state estimation using a much simpler approach compared to the literature.

Keywords: Energy Storage, Machine learning, Estimation
Abstract: This paper demonstrates a fast, low-cost, highly applicable, and robust battery health diagnostic method applicable at any SOC level using a 2-minute step current response. This response is first used to obtain cell physical parameters from ECM modeling and then serves as input for a voting regressor, an ensemble learning-based machine learning algorithm, to estimate four health metrics: cell capacity, negative electrode capacity, positive electrode capacity, and lithium inventory. Diagnostic accuracy is validated using experimental data from 36 pouch cells across three scenarios: charge-only, discharge-only, and SOC-agnostic, subject to different data availability. Results show that the MAPE for cell capacity, negative electrode capacity, positive electrode capacity, and lithium inventory estimation range from 1.51% to 1.89%, 4.48% to 5.00%, 1.52% to 1.64%, and 2.01% to 2.17%, respectively. Also, similar diagnostic accuracy can be achieved across scenarios even when cell SOC is unknown, demonstrating the effectiveness, robustness, and applicability of the proposed battery health diagnostic framework enabling real-time aging-aware control, enhancing battery safety, lifespan, and performance.

Keywords: Control applications, Energy Storage, Optimization
Abstract: Fast charging of lithium-ion batteries is essential for the widespread adoption of electric vehicles (EVs), but it introduces safety risks such as over-temperature and lithium plating (LP). This paper presents a novel fast charging algorithm based on Control Barrier Functions (CBFs) to mitigate these risks while ensuring computational efficiency. The proposed method leverages an extended equivalent circuit model (eECM) to predict battery behavior and enforce constraints on anode potential, temperature, state of charge (SoC), and terminal voltage. In contrast with Model Predictive Control (MPC)-based approaches, which require iterative optimization solvers, our CBF formulation provides an analytical solution, significantly reducing computational complexity. Numerical simulations demonstrate that the CBF-based approach achieves safety and performance levels comparable to MPC while requiring substantially less computation time.

Keywords: Time delays, Complex networks, Communication networks
Abstract: This paper studies a consensus problem in multidimensional networks having the same agent-to-agent interaction pattern under both intra- and cross-layer time delays. Several conditions for the agents to asymptotically reach a consensus are derived, which involve the overall network's structure, the local interacting pattern, and the assumptions specified on the time delays. The validity of these conditions is proved by direct eigenvalue evaluation and supported by numerical simulations.

Keywords: Cooperative control, Time delays, LMIs
Abstract: This paper investigates the convergence conditions of leader-follower matrix-weighted consensus networks in the presence of constant time delays. Several delayed consensus algorithms for networks of single-integrator agents, relying solely on relative state information, are analyzed. Convergence to a matrix-weighted leader-follower consensus is established using eigenvalue-based criteria and the Lyapunov-Krasovskii method. It is shown that, in most cases, when the leaders have non-identical initial states, the followers asymptotically converge to distinct clusters, which may lie outside the convex hull of the leaders’ initial states. The theoretical results are supported with numerical simulations.

Keywords: Linear systems, Complex networks, Distributed control
Abstract: This paper studies the resilient consensus of a multi-agent system in the presence of an external source and misbehaving agents. A dynamic model is considered that integrates an external signal from the source to the weighted-mean-subsequence-reduced (W-MSR) based algorithm. Graph-theoretic conditions for resilient consensus are derived based on a new notion of tier-k conformity, ensuring that the consensus value remains independent of the misbehaving agents and is determined by the external signal. For the special case where all agents receive the external signal, exponential convergence to the consensus value is proven. Numerical examples are included to illustrate and validate the theoretical findings.

Keywords: Distributed control, Autonomous systems, Cooperative control
Abstract: In this study, we propose distributed adaptive control laws for bearing-based formation control systems, where the control gain is dynamically adjusted based on the bearing formation error. Unlike conventional control approaches that do not explicitly design the control gain or use a constant gain, our proposed adaptive control laws enable dynamic gain adaptation based on real-time bearing formation errors. Therefore, we can expect a rapid reduction of formation errors while minimizing unnecessary control efforts in a steady state. Also, we conduct a rigorous Lyapunov analysis to prove the stability of the proposed adaptive control laws. Numerical simulation examples will be provided to analyze the performance of the adaptive control laws, demonstrating fast convergence and formation stabilization.

Keywords: Nonlinear systems, Distributed control, Complex networks
Abstract: This paper investigates the consensus problem of a multi-agent system whose states evolve on the 2-Sphere. To design the consensus protocol, we associate to each agent a virtual system whose virtual state communicates with its neighbors. Then, we use the virtual state as a reference for the actual system. The communication between the virtual agents is modeled by an undirected graph. The stability analysis shows that the agents reach consensus with an exponential convergence rate. We provide simulation results that validate the effectiveness of the proposed consensus algorithm.

Keywords: Predictive control, Nonlinear systems, Biosystems
Abstract: This paper presents a scenario-based model predictive control (MPC) scheme designed to control an evolving pandemic via non-pharmaceutical intervention (NPIs). The proposed approach combines predictions of possible pandemic evolution to decide on a level of severity of NPIs to be implemented over multiple weeks to maintain hospital pressure below a prescribed threshold, while minimizing their impact on society. Specifically, we first introduce a compartmental model which divides the population into Susceptible, Infected, Detected, Threatened, Healed, and Expired (SIDTHE) subpopulations and describe its positive invariant set. This model is expressive enough to explicitly capture the fraction of hospitalized individuals while preserving parameter identifiability w.r.t. publicly available datasets. Second, we devise a scenario-based MPC scheme with recourse actions that captures potential uncertainty of the model parameters. e.g., due to population behavior or seasonality. Our results show that the scenario-based nature of the proposed controller manages to adequately respond to all scenarios, keeping the hospital pressure at bay also in very challenging situations when conventional MPC methods fail.

Keywords: Complex networks, Reinforcement learning
Abstract: In this paper, we propose a new approach to reduce infection spread among individuals in a networked susceptible-infected-susceptible (SIS) model. Our approach assumes that each agent can adjust the strength of its connections with neighbors. An agent is randomly selected to learn about its local network and implement strategies to minimize infection spread over 100,000 episodes. We leverage deep deterministic policy gradient (DDPG), a deep reinforcement learning algorithm. We implement and evaluate our approach on graphs of various sizes and configurations. Additionally, we examine cooperative multi-agent scenarios in which agents collaborate to control the spread of infections.

Keywords: Modeling, Identification, Optimization
Abstract: Bariatric surgery, particularly Roux-en-Y Gastric Bypass (RYGB), significantly alters postprandial physiology, including enhanced gastric emptying and increased glucose absorption. These changes can contribute to the development of post-bariatric hypoglycemia (PBH), and its associated a clinically relevant complication. As demonstrated in type 1 diabetes, the development of mathematical models and their use in simulation environments can greatly accelerate the development and testing of novel clinical approaches. This work aims to identify the most appropriate meal absorption model structure from existing models in the literature by incorporating parameters that reflect physiological alterations following RYGB surgery. Glucose rate of appearance (Ra) mean data from 12 subjects both pre- and post-surgery is fitted. The best-fitting model was selected using the Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC). The selected model was able to accurately capture the exaggerated postprandial glycemic excursions observed in the data of RYGB subjects (RMSE= 0.382 g/kg/min). Such a model can be further integrated into larger glucose metabolism simulators, potentially providing a valuable simulation tool for the study of PBH management.

Keywords: Health and medicine, Hybrid systems, Predictive control
Abstract: Obstructive sleep apnea (OSA) is a widespread sleep disorder that significantly impacts public health. Continuous positive airway pressure (CPAP) therapy is the gold standard for treating OSA; however, adherence to treatment remains a major challenge, and many patients discontinue CPAP use shortly after initiation. To address this challenge, this paper presents a three-degree-of-freedom Kalman filter-based hybrid model predictive control (3DoF-KF HMPC) framework aimed at improving CPAP adherence and quality of life among patients with OSA. A hypothetical dynamic model (developed with OSA domain experts) and patterned after the SleepWell24 intervention is presented; this model serves as both the simulation and internal model for the 3DoF-KF HMPC, which provides "ambitious yet attainable" CPAP goals while coordinating actions to mitigate reported symptoms. By integrating behavioral strategies with hybrid model predictive control techniques, the framework personalizes CPAP recommendations, aiming to enhance participant engagement and adherence. The results presented in this paper show promise for the effectiveness of a future "just-in-time" control engineering-based adaptive intervention with the goal of optimizing CPAP use and improving health-related outcomes for newly diagnosed OSA patients.

Keywords: Control applications, Health and medicine, Estimation
Abstract: Effective therapeutic neurostimulation requires predictive models that can reliably map how neural activity responds to specific stimuli. While recent advances in closed-loop neurostimulation devices show promise for treating neurological disorders, most modeling approaches neglect the crucial relationship between stimulation input and neural response. A fundamental test of model predictive capability remains unaddressed: given baseline electroencephalographic (EEG) data and subsequent neural responses to stimulation, can we accurately reconstruct information about the injected stimuli? Furthermore, can such reconstruction remain valid across different stimulation parameters and electrode configurations, reflecting the diverse electrode placements required in clinical practice? In this paper, we demonstrate that input reconstruction is achievable using discrete-time linear fractional-order dynamical networks, which capture the rich temporal dependencies characteristic of neural systems. We present a novel minimization-minimization algorithm that generalizes expectation-maximization principles to learn both system parameters and unknown inputs. Through both a pedagogical example and extensive validation on a clinical dataset of simultaneous high-density EEG and intracerebral stimulation recordings, we show successful reconstruction of therapeutic biphasic pulses across varying stimulation locations, intensities, and electrode configurations. Our results demonstrate consistent reconstruction performance using different sizes and locations of electrode arrays, suggesting the robustness of our approach for practical neurostimulation applications.

Keywords: Adaptive control, Health and medicine, Control applications

Keywords: Autonomous systems, Mobile Robots, Robotics
Abstract: Quadratic programs (QP) subject to multiple time-dependent control barrier function (CBF) based constraints have been used to design safety-critical controllers. However, ensuring the existence of a solution at all times to the QP subject to multiple CBF constraints (hereby called compatibility) is non-trivial. We quantify the feasible control input space defined by multiple CBFs at a state in terms of its volume. We then introduce a novel feasible space (FS) CBF that prevents this volume from going to zero. FS-CBF is shown to be a sufficient condition for the compatibility of multiple CBFs. For high-dimensional systems though, finding a valid FS-CBF may be difficult due to the limitations of existing computational hardware or theoretical approaches. In such cases, we show empirically that imposing the feasible space volume as a candidate FS-CBF not only enhances feasibility but also exhibits reduced sensitivity to changes in the user-chosen parameters such as gains of the nominal controller. Finally, paired with a global planner, we evaluate our controller for navigation among other dynamically moving agents in the AWS Hospital gazebo environment. The proposed controller is demonstrated to outperform the standard CBF-QP controller in maintaining feasibility.

Keywords: Nonlinear systems, Algebraic/geometric methods
Abstract: Collision-free motion is a fundamental requirement for many autonomous systems. This paper develops a safety-critical control approach for the collision-free navigation of polytope-shaped agents in polytope-shaped environments. A systematic method is proposed to generate control barrier function candidates in closed form that lead to controllers with formal safety guarantees. The proposed approach is demonstrated through simulation, with obstacle avoidance examples in 2D and 3D, including dynamically changing environments.

Keywords: Robotics, Predictive control, Mobile Robots
Abstract: Human-robot collision avoidance plays an essential role in facilitating the integration of robotic systems. However, many state-of-the-art approaches do not consider the future movements of dynamic obstacles or neglect the importance of allowing trajectory tracking behavior. To this end, a method based on model predictive control (MPC) and dynamic obstacle prediction is proposed that features trajectory tracking with minimal error in obstacle-free cases. Otherwise, collisions with dynamic obstacles are avoided by either only adapting the speed or by additional local deviations from the path while quickly recovering to a low tracking error afterwards. This MPC formulation is applied to an omnidirectional mobile robot within simulations and real-world experiments that demonstrate the effectiveness of the proposed approach with respect to tracking accuracy and human safety.

Keywords: Autonomous systems, Automotive applications, Transportation systems
Abstract: This work presents a new shared control framework for teleoperated vehicles, targeting critical safety challenges arising from the control communication latency and the correctness of driver warnings. The proposed delay-compensated shared control architecture integrates two key components: a conformal prediction-based warning system that proactively alerts remote drivers of potential hazards and an onboard safety filter that combines a delay compensator, a disturbance observer, and a control barrier function-based quadratic program. The proposed design framework generates real-time safe control commands at the human-operation level despite delayed human inputs. A high-fidelity simulation platform was developed for semi-autonomous vehicle teleoperation using Chrono, a multi-physics-based simulator. Through extensive experiments in diverse scenarios, the proposed approach demonstrates robust performance and reliable safety maintenance under aggressive maneuvers and communication delays.

Keywords: Autonomous systems, Cyberphysical systems, Robotics
Abstract: Autonomous mobile robots are actively applied to execute complex tasks, such as package delivery, autonomous taxiing, and search-and-rescue. Signal Temporal Logic (STL) offers a powerful formalism for such complex tasks. However, designing plans (trajectories) that satisfy tasks formalized by STL grammar, particularly for nonholonomic systems like a car-like robot or a fixed-wing aircraft, is a challenging problem. This paper proposes a method to generate trajectories for a multi-robot system with car-like robots to perform complex tasks specified with STL grammar. The proposed method solves a nonlinear program (NLP) to construct trajectories with several constant curvature curves that satisfy the specification. In doing so, it also guarantees the kinematic feasibility of the solution trajectories. Extensive simulation studies show that the proposed method finds satisfying solutions 4X faster than a model predictive control baseline. Additionally, it is able to construct trajectories for complex STL specifications that the baseline fails to satisfy.

Keywords: Aerospace applications, Control applications
Abstract: This paper aims to mitigate the vulnerability of an aircraft to radar detection arising from inherent design limitations by proposing a technique to generate maneuvers that reduce the aircraft’s radar cross-section (RCS) below a specified threshold when exposed to dynamic (mobile) radar threats. The proposed approach utilizes control barrier functions and exploits the relationship between control inputs and the RCS, with the required RCS database for the baseline aircraft generated through numerical analyzes. The outcomes are contrasted with a virtual path that omits RCS constraints and maintains the aircraft’s rotational attitude stationary. Simulation results validate the effectiveness of the proposed methodology, as evidenced by a comparison with the virtual trajectory, where the RCS reaches up to 25 dBsm. In contrast, the CBF-pilot approach keeps the RCS at the predefined threshold of 0 dBsm throughout the maneuver.

Keywords: Renewable Energy, Modeling, Reduced order modeling
Abstract: In recent years, there has been a growing focus on wave-to-wire (W2W) control of wave energy converters, aiming to fully harness the potential of wave energy devices. The W2W optimal control problem presents significant challenges in the marine energy field due to the complexity involved in developing detailed mathematical models for power take-off units and other supporting systems. In this context, oscillating water columns (OWCs) stand out as devices with the most robust technology, with all the power take-off equipment implemented above the water line and no need to move excessively to generate energy, unlike other point absorbers. The modeling of these systems is straightforward, as they have been extensively researched across multiple engineering fields, and both empirical data and performance maps are readily available. This paper investigates two simple optimal control formulations for OWCs. The first formulation is turbine efficiency maximization (TEM) control, and the second is W2W control. Both control methods are implemented in the presence of mooring systems, nonlinear excitation, and nonlinear hydrostatic forces. The study then compares the two control methods and highlights the impact of considering the entire power conversion chain within the control method on the performance of the turbine and generator. The results from this study show that a 30% increase in electrical power generation is expected when a W2W approach is implemented compared with the more traditional TEM approach. The TEM model used in this study is publicly available at: WEC-Sim Applications GitHub repository

Keywords: Renewable Energy, Predictive control, Energy Systems
Abstract: The deployment and operation of wind turbines can have a negative environmental impact on bird population. To preserve avian wildlife, several studies have been done to understand and reduce the risks of collisions between birds and wind turbines. This work proposes a novel control scheme for preventing bird collisions with rotating blades of wind turbines. The scheme consists of a nonlinear model predictive control (NMPC) based on obstacle avoidance. The core idea relies on the assumption that bird flight trajectories can be predicted, and then, the turbine speed is modified by a small amount to avoid a potential collision with the bird. The proposed NMPC scheme defines a bird-blade distance constraint to keep a safety distance between the wind turbine blades and birds approaching the rotor swept area. Numerical simulation results are presented to illustrate the behavior of the control scheme and its ability to prevent collisions.

Keywords: Process control, Identification, Gain scheduling
Abstract: Classical stability analysis can be difficult to apply to high-order systems due to its reliance on the availability of explicit analytical models. An example of high-order systems that stand the risk of instability with potentially catastrophic consequences but where little formal stability analysis has been applied is the pressure dynamics in deepwater oil and gas wells. This work examines the frequency response of realistic deepwater wells and analyzes how high-order interactions, well configuration, and operating points affect stability. We present a workflow to empirically determine interpretable stability margins for choke pressure control systems on challenging wells, and validate the approach through high-fidelity simulations. Finally, we demonstrate how the workflow can be exploited to generate a robust gain-scheduled choke pressure controller.

Keywords: Estimation, Renewable Energy, Linear parameter-varying systems
Abstract: Wave excitation force estimation for wave energy converters remains a significant challenge, since the vast majority of the state-of-the-art methods are not able to fully address all the requirements for a reliable estimate under nonlinear conditions, particularly when paired with optimal controllers. To address this issue, we leverage the features of zonotopic Kalman filters to deal with nonlinear effects in the system dynamics, while maintaining a reduced computational cost, demonstrating suitable optimality properties. For the optimal design of the filter, we introduce a novel quasi-LPV formulation of wave energy converters, which effectively takes into account nonlinear effects in the dynamics of the device. For estimation purposes, the wave excitation force is modelled according to the internal model principle, while the filter also provides an estimate of the system state, necessary for energy maximising optimal control. The proposed filter estimation quality is demonstrated extensively via a numerical case study, under realistic irregular wave conditions.

Keywords: Control applications, Gain scheduling, Renewable Energy
Abstract: As Floating Offshore Wind Turbines (FOWTs) take on a growing share in global renewable energy generation, they will eventually be required to support the grid through ancillary services. This will be achieved through down-regulating (DR) the FOWT by varying the output electrical power through Active Power Control (APC). The FOWT control schemes for power maximization, derived from land-based wind-turbine (LWT) have bandwidth limitations due to the existence of right-half-plane zeros (RHPZs) in Region 3. Conventionally this problem is solved by detuning the land-based controllers but this significantly impacts the power quality. This paper proposes a novel gain-schedule design methodology for Collective Pitch Control (CPC), based on linear model dynamics in Region 3 for DR operation of FOWT. Parallel compensation through feedback of nacelle fore-aft velocity in generator torque is used to address the RHPZs and to achieve control bandwidth similar to LWT counterparts. Two operational strategies (OS) for power command tracking are compared for distributing demanded power, i.e., constant generator torque or angular speed references. System performance is evaluated through both linear and nonlinear analysis in Region 3 considering up to 20% DR. OpenFAST toolbox in MATLAB/SIMULINK is used under varying wind conditions, using RegA and RegD standard power signals by the PJM regional transmission organization for nonlinear analysis. Results demonstrate the proposed solutions’ effectiveness in FOWTs for achieving power quality alongside system stability.

Keywords: Renewable Energy, Identification, Control applications
Abstract: Achieving economic feasibility in wave energy conversion relies on the development of efficient and reliable control strategies, which require accurate control-oriented models of wave energy systems. However, deriving such models from first-principles-based hydrodynamic approaches can be challenging, as linearisation and small-motion assumptions often lead to discrepancies between theoretical predictions and real-world behaviour. Recognising these limitations, this paper presents a frequency-domain system identification approach applied to a multiple-degree-of-freedom wave energy converter, SWINGO. Experimental tests on a scaled prototype were conducted in a wave basin to obtain system response data under different operating conditions. The identification approach captures the coupled dynamics between the floater and gyropendulum, improving upon boundary-element-method-based models. The identified best linear approximation models are validated across multiple sea states, demonstrating their effectiveness in characterising the system response. Validation results are analysed in terms of normalised mean average percentage error, confirming the approach suitability for control design.

Keywords: Optimization, Hybrid systems, Reinforcement learning
Abstract: This paper presents an optimization-based approach for energy management and trajectory planning in hybrid electric fuel cell unmanned aerial vehicles (UAVs). The objective is to minimize mission completion time while efficiently managing the energy distribution among fuel cells, batteries, and super capacitors in a hybrid power system. The energy management system (EMS) is based on twin delayed deep deterministic policy gradient (TD3), to optimally allocate power while respecting operational constraints and enhancing system longevity. Dynamic programming (DP) is employed to generate a velocity lookup table, ensuring feasible trajectories that minimize travel time within energy limits and dynamic constraints and using a propulsion power consumption model. An iterative method addresses the coupled optimization of trajectory planning and EMS, accounting for power losses in the hybrid electric system. Simulation results validate the proposed approach, highlighting its capability to optimize mission performance while effectively managing energy in the system.

Keywords: Control applications, Renewable Energy, Optimization
Abstract: Solar photovoltaic power plants are inherently intermittent, and their variations in power output create challenges for grid integration. Over the years, numerous studies have explored methods to regulate ramp rates under different operational scenarios. However, few have addressed this issue while considering economic viability. This paper proposes a novel hybrid switching-based control methodology combined with battery size optimization to maximize annual profit while ensuring operational compliance. Simulation results demonstrate the economic benefits and effectiveness of the proposed approach in eliminating ramp rate violations, improving the integration of intermittent energy sources, and reducing the impact of power curtailment.

Keywords: Renewable Energy, Optimization, Energy Systems
Abstract: The integration of agriculture and photovoltaic energy production has given rise to agrivoltaic technology, where solar panels are installed over cultivated land to enable simultaneous crop growth and energy generation. While this approach increases land productivity, it may also reduce plant light exposure, potentially affecting yields. This paper investigates the interaction between crops and photovoltaic modules, introducing a novel optimization-based method to balance light availability for crops with power output. We present the underlying mathematical model, leading to a non-convex mixed-integer optimization problem, and then propose an approximate convex formulation with formally established error bounds. Numerical simulations using real-world data from an agrivoltaic site in Southern Italy are provided and analyzed.

Keywords: Predictive control, Control applications
Abstract: District heating networks are increasingly being deployed in modern cities to provide space heating and hot water to buildings. Due to the intermittency of energy sources, these networks can face challenges under extreme conditions, such as very cold outdoor temperatures, which may lead to thermal unfairness and reduced efficiency. Therefore, a coordination unit is crucial for effectively managing these challenges and maintaining network performance. By abstracting district heating networks as cascade control systems, this paper introduces a novel coordination scheme for district heating networks. This scheme, which is called CARE-HEAT, is designed to improve thermal fairness and efficiency within these networks, while also addressing high-level cross-building policies. We evaluate CARE-HEAT and demonstrate that it improves thermal fairness by ~84% compared to uncoordinated networks, and by ~67% compared to networks where hot water flow is distributed equally among buildings.

Keywords: Energy Systems, Predictive control, Stochastic/uncertain systems
Abstract: Energy systems with a significant share of renewable energy production require control strategies that can handle the effects of uncertainty to ensure stable and reliable operation. Simultaneously, liberalized energy markets balance supply and demand on a large scale. In an effort towards renewable energy integration, this study proposes a stochastic model predictive control framework to handle the energy management of an offshore wind utility system while trading on the day-ahead spot and intraday markets. Furthermore, this study extends the framework with the entropic risk measure (ERM) to handle risk-sensitive decision-making. The uncertainty is modeled in discrete scenarios, which propagate uncertainty through a nonlinear model. The proposed framework is validated in simulation with historical data. Results show that the proposed method balances demand and supply while minimizing operational costs through market participation. Furthermore, the framework is compared to itself with different degrees of risk aversion to illustrate the effects of risk sensitivity on the system.

Keywords: Energy Systems, Renewable Energy, Predictive control
Abstract: This study proposes a general forecast and control framework for constrained stochastic nonlinear optimal control of isolated gas-renewable energy systems with energy storage. Typically, these systems require control strategies that can handle significant uncertainties in produced renewable energy due to forecast uncertainty from meteorological forecasts. To address the uncertainty in meteorological forecasts, data-driven stochastic grey-box models of the renewable energy source are modelled and probabilistically forecast (PF) with stochastic differential equations (SDEs). The PF scheme improves upon the meteorological forecasts by forming a probability distribution in time with the meteorological forecasts and past data as input. Based on these distributions, a multi-stage (MS) nonlinear model predictive control (NMPC) formulation is utilised, resulting in a tractable control formulation. The proposed framework is validated in simulation with real-life data for a hybrid gas-wind energy system, which shows that the proposed method is real-time capable despite using standard solvers and outperforms standard methods, such as certainty-equivalent NMPC when relying on meteorological forecasts. Though motivated by the energy sector, the proposed method can be extended to any stochastic system since SDEs are a general class of stochastic processes.

Keywords: Estimation, Aerospace applications, Kalman filtering
Abstract: In this paper a noise covariance model structure is proposed to constrain the solutions to an Autocovariance Least Squares (ALS) problem. These constraints enable the identification of process and measurement noise covariances simultaneously across multiple operating conditions. This approach benefits real world systems where data at a specific operating condition are limited, by utilizing data across a larger set of conditions. One such system is aircraft flight throughout a range of dynamic pressures. The structured noise covariance constraints result in identified noise covariance models as a function of the operating condition. The identified model can be used to approximate the noise covariances in between the design conditions where datasets are available. The approach is demonstrated first in simulation with a mass-spring-damper and then experimentally with the University of Washington's gust load alleviation wind tunnel test-bed. The wind tunnel ALS problem is solved simultaneously at multiple dynamic pressures with an affine noise covariance model structure. A linear Kalman filter, designed with the constrained ALS identified noise covariances, is demonstrated at a design condition and at a condition where no design data were available.

Keywords: Estimation, Sensors, Biosystems
Abstract: Acoustic telemetry systems equipped with multiple receiver units can determine the 3D positions of individual fish by employing the time difference of arrival (TDoA) method. However, these datasets are often temporally sparse and exhibit uncertain positioning accuracy, which limits their utility in describing continuous fish behaviour. State estimation methods can address these limitations by enhancing data accuracy through mathematical modelling. This study employs the Kalman filter to estimate fish trajectories and velocities in open-net sea cages. We used a kinematic constant velocity model as the motion model, and the observation model reflects the estimator’s assimilation of Cartesian position measurements. The measurement noise of the state estimator was calculated as the sample covariance of a position dataset recorded using an acoustic telemetry tag positioned statically relative to the receivers in a cage at a commercial fish farm. The process noise was tuned so that the average normalized innovations squared (ANIS) of the estimator satisfied a 95% confidence interval, ensuring that the estimator was consistent. The state estimator was validated using position datasets for two free-swimming salmon tracked with acoustic telemetry in the same cage as the stationary tag. The resulting estimated trajectories and velocities resemble those observed for real salmon in open-net sea cages, and the estimated speeds were close to those recorded in a previous study using Doppler shift measurements. This work represents a natural progression towards improving the accuracy with which individual fish trajectories and velocities may be estimated through acoustic telemetry in aquaculture.

Keywords: Estimation, Navigation, Control applications
Abstract: This paper presents two optimal measurement pro- jection schemes for the factor-graph-based Global Navigation Satellite System (GNSS) positioning with Real-Time-Kinematics (RTK). While factor graph optimization (FGO) has demonstrated improved accuracy and robustness in GNSS positioning compared to conventional filtering-based methods, the improvement has a cost of increased computational complexity due to the fact that FGO processes the batch of historical data simultaneously. Two measurement projection schemes are proposed to alleviate the computational burden of FGO by optimally projecting the GNSS measurements into a lower-dimensional subspace. Thereby, the dimensionality of the factor graph optimization is significantly reduced with only minimally performance loss. Monte Carlo simulation results demonstrate that the proposed measurement reduction schemes can achieve a significant computational speedup for the FGO-based GNSS-RTK positioning while retaining high-precision positioning performance.

Keywords: Kalman filtering, Observers, Filtering
Abstract: We consider a Kalman filter that estimates the states of a system where measurements are taken at discrete times. The measurement frequency can be adjusted, and it is assumed that the measurements become more precise with increasing measurement period. Hence, we investigate the inherent trade-off between accurate measurement and measuring frequently enough to achieve the best possible state estimate using the continuous-discrete KF.

Keywords: Kalman filtering, Fault detection/accomodation, Sensor fusion
Abstract: GNSS integrity is essential for safe and robust navigation, especially in automated and autonomous systems. A key component to achieve integrity within filtering frameworks is the accurate modeling of uncertainties in the estimation problem, enabling the use of prior information to assess the integrity of new observations. Leveraging the Lie group extended Kalman filters (EKF) capability to capture the uncertainties inherent in loosely coupled GNSS-aided inertial navigation, this work presents a position measurement outlier detection method derived explicitly for the Lie group EKF. The method uses a Gaussian mixture to provide a more accurate approximation of the non-Gaussian distributed innovation that arises from the use of the Lie group EKF. A planar simulation demonstrates that the proposed method fully utilizes the Lie group EKF's uncertainty information, enabling more accurate detection and rejection of false position measurements, even in scenarios with high orientation uncertainty.

Keywords: Estimation, Optimization, Stochastic/uncertain systems
Abstract: The state and output estimation accuracy depends on both the observer and the sensor locations. This paper focuses on this co-design problem in lithography applications. A theoretical formulation of this co-design problem is presented and solved for discrete-time linear stochastic models. We compare the optimal solution of two variants of the estimation problem. The first variant minimizes the transient estimation error whereas the second one minimizes the steady-state estimation error. In both cases, the Kalman filter is optimal. While solving these problems for a lithography application formulated as a 3D thermoelastic model, we observe a significant difference between the optimal sensor placements. Our results highlight the importance of designing a sensor layout in line with the desired transient or steady-state estimation performance.

Keywords: Cooperative control, Ships and offshore vessels, Nonlinear robust control
Abstract: 在本文中，我们讨论了合作的任务 多个欠驱动无人驾驶的路径跟踪控制 受到干扰的水面车辆 （USV）。我们提出了一个 生成 n 的分布式路径协作框架 n 个 USV 的所需路径，保持所需的 编队。 此外，还设计了一种用于 USV 的路径跟踪控制器 基于反步方法，确保 位置跟随误差收敛于任意 小 neighborhood 并实现 闭环系统的错误。此外，我们还设计了一个 饱和 函数和干扰观察器来增强 鲁棒性 控制器。要验证 提出 controller 中，我们呈现并分析 模拟 和实验测试。

Keywords: Game theory, Reinforcement learning, Intelligent systems
Abstract: This paper explores the trade-off between efficiency and complexity in mean-field game (MFG) theory for large-scale multi-agent systems (LS-MAS). While MFG reduces computational burden by avoiding the curse of dimensionality and achieving Nash equilibrium, its solutions often yield inefficient social cost compared to the centralized McKean-Vlasov control. To address this, we propose an extended MFG (EMFG) framework that improves efficiency with minimal complexity increase by introducing a decomposed mean field term using probability density functions (PDFs). Agents are grouped based on terminal PDF constraints, and an actor-critic-decomposed mass (ACDM) algorithm is developed to solve the resulting forward-backward PDE system. Although the PDF decomposition enlarges the neural network, the algorithm effectively balances control efficiency and computational cost. We provide an induction-based proof to bound the inefficiency gap between EMFG and centralized control and establish Lyapunov stability for the convergence of ACDM.

Keywords: Cooperative control, Control applications, Distributed control
Abstract: Pulsed light sources are a critical component of modern lithography, with fine light beam wavelength control paramount for wafer etching accuracy. We study optimal wavelength control by casting it as a decentralized linear quadratic Gaussian (LQG) problem in presence of time-delays. In particular, we consider the multi-optics module (optics and actuators) used for generating the requisite wavelength in light sources as cooperatively interacting systems defined over a directed graph. We show that any measurement and other continuous time-delays can be exactly compensated, and the resulting optimal controller implementation at the individual optics-level outperforms existing wavelength control techniques.

Keywords: Autonomous systems, Cooperative control, Aerospace applications
Abstract: Manned-Unmanned Teaming (MUM-T) is of increasing importance in the modern aerospace arena. In this context, task-sharing between humans and autonomous systems withstands as an open research area. Motivated by this challenge, this study proposes a novel task-sharing concept, i.e. radar cross-section (RCS) shadowing—aimed at reducing the vulnerability of a non-stealth manned aircraft in a contested, hostile environment by concealing it with a stealth unmanned autonomous wing-man. For this purpose, the leader (manned) and wing-man (unmanned) aircraft are primarily represented as triangles approximating their shapes. The wing-man’s projected shadow onto the leader’s triangular representation defines the masked region. To ensure full coverage, a reference command generation strategy is designed for the position channels (north, east, and altitude), guiding the wing-man’s motion. Additionally, the leader’s reference commands are determined based on the center of the shadow triangle to enhance coverage reliability. As a final step, the control architecture employs identical Lyapunov-based position controllers for both aircraft. The efficacy of the proposed strategy is demonstrated through two distinct scenarios, showing that the leader is successfully shadowed during the mandatory radar penetration maneuver, thereby significantly reducing radar visibility.

Keywords: Control Technology, Control applications
Abstract: Irrigation channels operating under practical decentralized controllers can exhibit string instabilities in the form of undesirable amplification of flow transients as they propagate spatially. To limit the propagation of such instability, the application of an existing decentralized water-level balancing control scheme along a mid section of multiple pools is considered. The main design trade-offs are illustrated by numerical example.