Keywords: Networked control systems, Identification for control, Quantized systems
Abstract: Future autonomous systems ranging from multi-robot networks to distributed satellite constellations must operate reliably under limited, delayed, and lossy communications. Traditional control and learning frameworks, however, assume perfect information flow, leaving a fundamental gap between the theory and the realities of modern cyber-physical systems. This tutorial introduces a unified framework for communication-constrained learning, estimation, and control, where sensing, estimation, learning, and decision-making are tightly coupled through shared, resource-limited communication channels.

We begin with the theoretical foundations of system identi- fication under communication constraints, highlighting recent advances in operator-theoretic and learning-enabled modeling that yield provable performance guarantees even with degraded data. The second part focuses on communication-aware control design, covering sequential and joint control–estimation–communication co-design, adaptive quantization, and risk-aware robust control under bandwidth-limited, delay-prone, and unreliable communication. Finally, we demonstrate how these theories enable resilient guidance and control in robotics applications where nonlinear dynamics, limited bandwidth, and high-latency links pose critical challenges to autonomy.

The tutorial combines theory, algorithms, and case studies to provide attendees with (i) a principled understanding of the quality–quantity trade-off in data-driven control, (ii) practical tools for implementing communication-aware controllers, and (iii) insights into emerging aerospace, autonomous cars, and robotic applications where these ideas are reshaping design paradigms.

Keywords: Aerospace, Control applications
Abstract: This paper presents an input-constrained nonlinear guidance law to address the problem of intercepting a stationary target in contested environments with multiple defending agents. Contrary to prior approaches that rely on explicit knowledge of defender strategies or utilize conservative safety conditions based on a defender's range, our work characterizes defender threats geometrically through engagement zones that delineate inevitable-interception regions. Outside these engagement zones, the interceptor remains invulnerable. The proposed guidance law switches between a repulsive safety maneuver near these zones and a pursuit maneuver outside their influence. To deal with multiple engagement zones, we employ a smooth minimum function (log-sum-exponent approximation) that aggregates threats from all the zones while prioritizing the most critical threats. Input saturation is modeled and embedded in the non-holonomic vehicle dynamics so the controller respects actuator limits while maintaining stability. Numerical simulations with several defenders demonstrate the proposed method’s ability to avoid engagement zones and achieve interception across diverse initial conditions.

Keywords: Adaptive control, Lyapunov methods, Aerospace
Abstract: This paper develops an inversion-free adaptive controller for stabilizing and tracking the trajectory of a bicopter system. In a bicopter system, the inertial parameters, that is, mass and moment of inertia, parameterize the input map. Since the classical adaptive backstepping technique requires inverting the input map, which would contain the parameter estimates, the stability of the closed-loop system cannot be guaranteed because the inversion of these estimates may result in division by zero. This paper proposes a novel technique to circumvent the inversion of parameter estimates in the control law. The resulting controller requires only the sign of the unknown parameters, which are trivially known to be positive in the case of a bicopter. The proposed controller is validated in simulation for a smooth and nonsmooth trajectory-tracking problem.

Keywords: Aerospace, Predictive control for nonlinear systems, Lyapunov methods
Abstract: In this work, a nonlinear adaptive Lyapunov-based model predictive control (NALMPC) framework for attitude regulation of quadcopters is presented. Through backstepping, a Lyapunov controller is synthesized that accounts for angular displacement constraints, while considering uncertain moments of inertia. The derived feedback laws are utilized to form a contractive constraint that is integrated into the optimal control problem (OCP) of the MPC protocol. Additionally, the moment of inertia adaptation laws are integrated into the OCP, thus resulting to a predictive model that is updated in real-time, based on the operating conditions. A numerical comparative case study against a standard nonlinear MPC (NMPC) formulation is presented, where NALMPC outperforms NMPC in terms of tracking error and computational load. Finally, experimental validation on a real-world quadcopter demonstrates the efficacy of the proposed framework in terms of tracking and computational efficiency.

Keywords: Optimal control, Aerospace, Flight control
Abstract: This paper presents an optimal control framework for emergency landing trajectory generation under loss-of-thrust conditions. The method incorporates population density maps to minimize ground risk while ensuring feasibility with a six-degree-of-freedom nonlinear dynamics model. Trajectories are initialized using either discrete search or Dubins paths. Time-dependent and time-independent cost functions are evaluated. Using real population density data from the New York City region, results show that excluding a time cost favors longer trajectories that bypass densely populated regions, while including time in the cost function reduces flight duration at the expense of greater population exposure. Closed-loop simulations confirm feasibility and consistently lower costs compared to discrete search and Dubins solutions. The study highlights the importance of accurate initialization for optimal control as small errors in reference trajectory position or yaw can result in an adverse optimal control gradient and in turn infeasible results. Convergence and computational considerations suggest optimal control be deployed in preflight planning where computational and supervisory requirements can be safely accommodated.

Keywords: Kalman filtering, Estimation, Modeling
Abstract: In environmental monitoring as well as emergency response applications such as wildfires, wind velocity measurement is essential. Quadrotor UAVs have become popular platforms for wind velocity estimation due to their maneuverability, compact size, and cost-effectiveness. Numerous studies use the Extended Kalman Filter (EKF) to estimate the wind velocity based on the quadrotor dynamic model. However, EKF performance is constrained by its reliance on linearized approximations of the nonlinear quadrotor dynamics around current states, limiting accuracy in highly nonlinear scenarios, including windy conditions. This study proposes the use of an Unscented Kalman Filter (UKF), a nonlinear estimator to provide accurate wind estimations while maintaining the trajectory of the quadrotor UAV. The quadrotor is modeled on the Special Euclidean group SE(3) and the approach is evaluated through numerical simulations using a geometric controller to maintain quadrotor flight paths, under different wind conditions. The results indicate that the UKF consistently outperforms the EKF, reducing wind velocity estimation error by 11.8% under near-linear flight conditions, and 25.5% under strongly nonlinear conditions. This demonstrates the potential of the UKF as a reliable estimator for highly nonlinear scenarios, capable of maintaining the trajectory with minimal deviation while providing accurate wind velocity estimations.

Keywords: Constrained control, Aerospace
Abstract: We present a tutorial focused on the applications of control barrier functions (CBFs) to flight-critical aerospace systems. This tutorial is motivated by the need for mathematically justified approaches to state and input limiting on various aerospace systems, much of which is currently accomplished using heuristic techniques. Here, we illustrate how CBFs -- originally motivated by applications in robotics -- provide a natural framework for constrained control of aerospace systems. Our tutorial encompasses both the theoretical foundations of CBFs as well as real-world aerospace applications of CBFs.

Keywords: Game theory, Agents-based systems, Information theory and control
Abstract: In adversarial settings, a mobile agent may strategically plan its motion to influence an opponent’s inference about its intended goal. We study deceptive path planning in a scenario where a mobile agent aims to reach a privately selected goal while an adversarial observer allocates limited defensive resources based on the observed trajectory. Unlike classical path-planning and goal-recognition approaches that model observers as passive inference process, our game-theoretic formulation models them as strategic decision-makers. For the resulting dynamic asymmetric-information game, we develop an efficient solution method that combines a linear programming formulation with the Double Oracle algorithm. To evaluate performance, we introduce metrics that quantify both the risk and the effectiveness of deception and provide illustrative numerical examples.

Keywords: Game theory, Agents-based systems, Network analysis and control
Abstract: Game-theoretic models and solution concepts provide rigorous tools for predicting collective behavior in multi-agent systems. In practice, however, different agents may rely on different game-theoretic models to design their strategies. As a result, when these heterogeneous models interact, the realized outcome can deviate substantially from the outcome each agent expects based on its own local model. In this work, we introduce the game-to-real gap, a new metric that quantifies the impact of such model misspecification in multi-agent environments. The game-to-real gap is defined as the difference between the utility an agent actually obtains in the multi-agent environment (where other agents may have misspecified models) and the utility it expects under its own game model. Focusing on quadratic network games, we show that misspecifications in either (i) the external shock or (ii) the player interaction network can lead to arbitrarily large game-to-real gaps. We further develop novel network centrality measures that allow exact evaluation of this gap in quadratic network games. Our analysis reveals that standard network centrality measures fail to capture the effects of model misspecification, underscoring the need for new structural metrics that account for this limitation. Finally, through illustrative numerical experiments, we show that existing centrality measures in network games may provide a counterintuitive understanding of the impact of model misspecification.

Keywords: Game theory, Autonomous systems, Automata
Abstract: Nash equilibrium is a central solution concept for reasoning about self-interested agents. We study the synthesis of Nash equilibria in two-player deterministic games on graphs, where players have private, partially-ordered preferences on temporal goals. Unlike prior work, which assumes preferences are common knowledge, we develop a communication protocol for equilibrium synthesis in settings where players' preferences are private information. In the protocol, players communicate to synthesize equilibria by exchanging information about when they can force desirable outcomes. We incorporate privacy by ensuring the protocol stops before enough information is revealed to expose players' preferences. We prove completeness by showing that, when no player halts communication, the protocol either returns an equilibrium or certifies that none exist. We then prove privacy by showing that, with stopping, the messages a player sends are always consistent with multiple preference relations and thus do not reveal some given secret regarding a player's true preference ordering. Experiments show that we can synthesize non-trivial equilibria while preserving privacy of preferences, showing the protocol’s potential for applications in strategy synthesis with constrained information sharing.

Keywords: Game theory, Learning, Information theory and control
Abstract: In Bayesian persuasion, an informed sender, who observes a state, commits to a randomized signaling scheme that guides a self-interested receiver's actions. Classical models assume the receiver knows the commitment. Instead, we study a setting where the receiver infers the scheme from repeated interactions. We bound the sender’s performance loss relative to the known-commitment case by a term that grows with the signal space size and shrinks as the receiver’s optimal actions become more distinct. We then lower bound the samples required for the sender to approximately achieve their known-commitment utility in the inference setting. We show that the sender in persuasion requires more samples than the leader in a Stackelberg game, which includes commitment but lacks signaling. Motivated by these bounds, we propose two methods for designing inferable signaling schemes, one being stochastic gradient descent (SGD) on the sender’s inference-setting utility, and the other being optimization with a boundedly-rational receiver model. SGD performs best in low-interaction regimes, but applying the boundedly-rational model and tuning the rationality constant remains a flexible method for designing inferable schemes. Applied in a safety-alert example, SGD finds schemes that have fewer signals and make citizens’ optimal actions more distinct than in the known-commitment case.

Keywords: Markov processes, Game theory, Information theory and control
Abstract: We study decision-making with rational inattention in settings where agents have perception constraints. In such settings, inaccurate prior beliefs or models of others may lead to inattention blindness, where an agent is unaware of its incorrect beliefs. We model this phenomenon in two-player zero-sum stochastic games, where Player 1 has perception constraints and Player 2 deceptively deviates from its security policy presumed by Player 1 to gain an advantage. We formulate the perception constraints as an online sensor selection problem, develop a value-weighted objective function for sensor selection capturing rational inattention, and propose the greedy algorithm for selection under this monotone objective function. When Player 2 does not deviate from the presumed policy, this objective function provides an upper bound on the expected value loss compared to the security value where Player 1 has perfect information of the state. We then propose a myopic decision-making algorithm for Player 2 to exploit Player 1's beliefs by deviating from the presumed policy and, thereby, improve upon the security value. Numerical examples illustrate how Player 1 persistently chooses sensors that are consistent with its priors, allowing Player 2 to systematically exploit its inattention.

Keywords: Game theory, Networked control systems, Iterative learning control
Abstract: This work investigates a two-player coordination game in which the players exhibit heterogeneous levels of bounded rationality. We analyze the log-linear learning dynamics, where the probability distribution used to select which of the agents gets to revise its strategy is fixed but not necessarily uniform. The stationary distribution of the resulting Markov chain on the strategy profile space is derived in closed-form as a function of the rationalities and the agent selection probabilities. We proceed by showing that adjusting the selection probabilities can be used to bias the stationary distribution toward the potential-maximizing state. However, this optimization comes at the cost of a reduced convergence rate, whereas the uniform selection probabilities uniquely maximizes the convergence speed irrespective of the players’ rationality levels. To address this trade-off, a Pareto-optimal probability selection rule is proposed, balancing the distributional bias with convergence rate. Moreover, it is shown that in coordination games, high levels of rationality sometimes accelerate convergence, whereas in other cases they may paradoxically hinder the convergence rate of the log-linear learning dynamics.

Keywords: Vision-based control, Autonomous systems, Constrained control
Abstract: Robotic systems navigating in real-world settings require a semantic understanding of their environment to properly determine safe actions. This work aims to develop the mathematical underpinnings of such a representation---specifically, the goal is to develop safety filters that are risk-aware. To this end, we take a two step approach: encoding an understanding of the environment via Poisson’s equation, and associated risk via Laplace guidance fields. That is, we first solve a Dirichlet problem for Poisson’s equation to generate a safety function that encodes system safety as its 0-superlevel set. We then separately solve a Dirichlet problem for Laplace’s equation to synthesize a safe guidance field that encodes variable levels of caution around obstacles---by enforcing a tunable flux boundary condition. The safety function and guidance fields are then combined to define a safety constraint and used to synthesize a risk-aware safety filter which, given a semantic understanding of an environment with associated risk levels of environmental features, guarantees safety while prioritizing avoidance of higher risk obstacles. We demonstrate this method in simulation and discuss how a priori understandings of obstacle risk can be directly incorporated into the safety filter to generate safe behaviors that are risk-aware.

Keywords: Constrained control, Robust adaptive control, Autonomous robots
Abstract: Planning safe trajectories under model uncertainty is a fundamental challenge. Robust planning ensures safety by considering worst-case realizations, yet ignores uncertainty reduction and leads to overly conservative behavior. Actively reducing uncertainty on-the-fly during a nominal mission defines the dual control problem. Most approaches address this by adding a weighted exploration term to the cost, tuned to trade off the nominal objective and uncertainty reduction, but without formal consideration of when exploration is beneficial. Moreover, safety is enforced in some methods but not in others. We propose a framework that integrates robust planning with active exploration under formal guarantees as follows: The key innovation and contribution is that exploration is pursued only when it provides a verifiable improvement without compromising safety. To achieve this, we utilize our earlier work on gatekeeper as an architecture for safety verification, and extend it so that it generates both safe and informative trajectories that reduce uncertainty and the cost of the mission, or keep it within a user-defined budget. The methodology is evaluated via simulation case studies on the online dual control of a quadrotor under parametric uncertainty.

Keywords: Stochastic systems, Statistical learning, Robust control
Abstract: We study safety for stochastic systems under sampled–data control and learned dynamics. We develop Bayesian Risk‑Aware Control Barrier Functions (RA‑CBFs) for discrete time. First, we give two guarantees for barrier crossing over a finite horizon: (i) a time‑uniform martingale bound using Ville’s inequality and a predictable variance budget, and (ii) a tighter pathwise bound that recovers the continuous‑time RA‑CBF margin via DDS time change and the reflection principle. Second, we propagate posterior uncertainty in drift and diffusion into an upper confidence bound (UCB) on the generator and derive a closed-form inter-sample robustness margin. Third, we synthesize a minimally invasive controller via a convex QP that enforces risk and uncertainty thresholds jointly. The result is a high-confidence, finite-horizon safety bound and a practical sampled-data controller that, under the stronger pathwise noise assumption, recovers the continuous-time RA-CBF margin while avoiding supermartingale-based S-CBF conditions.

Keywords: Neural networks, Delay systems, Uncertain systems
Abstract: We present a risk-aware safety certification method for autonomous, learning-enabled control systems. Focusing on two realistic risks, state/input delays and interval matrix uncertainty, we model the neural network (NN) controller with local sector bounds and exploit positivity structure to derive linear, delay-independent certificates that guarantee local exponential stability across admissible uncertainties. To benchmark performance, we adopt and implement a state-of-the-art IQC NN-verification pipeline. In representative cases, our positivity-based tests run orders of magnitude faster than SDP-based IQC while certifying regimes the latter cannot, providing scalable safety guarantees that complement risk-aware control.

Keywords: Reinforcement learning, Decentralized control, Cooperative control
Abstract: Risk sensitivity has become a central theme in reinforcement learning (RL), where convex risk measures and robust formulations provide principled ways to model preferences beyond expected return. Recent extensions to multi-agent RL (MARL) have largely emphasized the risk-averse setting, prioritizing robustness to uncertainty. In cooperative MARL, however, such conservatism often leads to suboptimal equilibria, and a parallel line of work has shown that optimism can promote cooperation. Existing optimistic methods, though effective in practice, are typically heuristic and lack theoretical grounding. Building on the dual representation for convex risk measures, we propose a principled framework that interprets risk-seeking objectives as optimism.We introduce optimistic value functions, which formalize optimism as divergence-penalized risk-seeking evaluations. Building on this foundation, we derive a policy-gradient theorem for optimistic value functions, including explicit formulas for the entropic risk/KLpenalty setting, and develop decentralized optimistic actorcritic algorithms that implement these updates. Empirical results on cooperative benchmarks demonstrate that riskseeking optimism consistently improves coordination over both risk-neutral baselines and heuristic optimistic methods. Our framework thus unifies risk-sensitive learning and optimism, offering a theoretically grounded and practically effective approach to cooperation in MARL.

Keywords: Autonomous systems, Automotive systems, Automotive control
Abstract: This paper presents a hybrid control framework with a risk-budgeted monitor for safety-certified autonomous driving. A sliding-window monitor tracks insufficient barrier residuals and triggers switching from a relaxed control barrier function (R-CBF) to a more conservative conditional value-at-risk CBF (CVaR-CBF) when the safety margin deteriorates. Two real-time triggers are considered: feasibility-triggered (FT), which activates CVaR-CBF when the R-CBF problem is reported infeasible, and quality-triggered (QT), which switches when the residual falls below a prescribed safety margin.

The framework is evaluated with model predictive control (MPC) under vehicle localization noise and obstacle position uncertainty across multiple AV-pedestrian interaction scenarios with 1,500 Monte Carlo runs. In the most challenging case with 5 m pedestrian detection uncertainty, the proposed method achieves a 94--96% collision-free success rate over 300 trials while maintaining the lowest mean cross-track error (CTE = 3.2--3.6 m), indicating faster trajectory recovery after obstacle avoidance and a favorable balance between safety and performance.

Keywords: Distributed control, Linear systems, Optimal control
Abstract: This paper designs H2 and H-Infinitydistributed controllers with local communication and local disturbance rejection. We propose a two-step procedure: first, select closed-loop poles; then, optimize over parameterized controllers. We build on the system level synthesis (SLS) parameterization --- primarily used in the discrete-time setting --- and extend it to the general continuous-time setting. We verify our approach in simulation on a 9-node grid governed by linearized swing equations, where our distributed controllers achieve performance comparable to that of optimal centralized controllers while facilitating local disturbance rejection.

Keywords: Distributed control, Feedback linearization, Large-scale systems
Abstract: We study distributed control for a network of nonlinear, differentially flat subsystems subject to dynamic coupling. Although differential flatness simplifies planning and control for isolated subsystems, the presence of coupling can destroy this property for the overall joint system. Focusing on subsystems in pure-feedback form, we identify a class of compatible lower-triangular dynamic couplings that preserve flatness and guarantee that the flat outputs of the subsystems remain the flat outputs of the coupled system. Further, we show that the joint flatness diffeomorphism can be constructed from those of the individual subsystems and, crucially, its sparsity structure reflects that of the coupling. Exploiting this structure, we synthesize a distributed tracking controller that computes control actions from local information only, thereby ensuring scalability. We validate our proposed framework on a simulated example of planar quadrotors dynamically coupled via aerodynamic downwash, and show that the distributed controller achieves accurate trajectory tracking.

Keywords: Distributed control, Cooperative control, Decentralized control
Abstract: Classical Distributed Model Predictive Control (DiMPC) requires multiple iterations to achieve convergence, leading to high computational and communication burdens. This work focuses on the improvement of an iteration-free distributed MPC methodology that minimizes computational effort and communication load. The aforementioned methodology leverages multiparametric programming to compute explicit control laws offline for each subsystem, enabling real-time control without iterative data exchanges between subsystems. Extending our previous work on iteration-free DiMPC, here we introduce a FAcet-based Critical region Exploration Technique for iteration-free DiMPC (FACET-DiMPC) that further reduces computational complexity by leveraging facet properties to do targeted critical region exploration. Simulation results demonstrate that the developed method achieves comparable control performance to centralized methods, while significantly reducing communication overhead and computation time. In particular, the proposed methodology offers substantial efficiency gains in terms of the average computation time reduction of 98% compared to classic iterative DiMPC methods and 42% compared to iteration-free DiMPC methods, making it well-suited for real-time control applications with tight latency and computation constraints.

Keywords: Distributed control, Machine learning, Adaptive control
Abstract: Distributed model predictive control (DMPC) has emerged as a computationally efficient alternative to centralized model predictive control (CMPC) by decomposing the centralized optimal control problem into a collection of constituent subsystem controllers. The subsystem controllers can be solved in parallel and coordinated to reach a consensus solution, significantly reducing computational cost at the potential expense of control quality relative to CMPC. In this paper, we demonstrate that state measurements and operational objectives, which serve as parameters in the optimal control problem of each distributed subsystem controller, are strongly correlated with the quality of control achieved by a particular DMPC architecture. We propose a novel graph-based classifier that selects the most suitable DMPC architecture based on these time-varying parameters and integrate this selection process into the online control loop. This adaptive approach dynamically responds to process variations, enhancing control performance and mitigating the potential loss in control quality typically associated with DMPC.

Keywords: Distributed control, Output regulation, LMIs
Abstract: Recent work [1] by the authors has considered robust cooperative output regulation of discrete-time uncertain heterogeneous (in dimension) multi-agent systems (MASs) using the distributed internal model approach, assuming that each follower has full access to its own state. This paper extends our discrete-time robust cooperative output regulation results to the case wherein followers have partial access to their states. Specifically, a detectability condition is added, and each follower employs a local state observer. We first show that the solvability of the robust cooperative output regulation problem (RCORP) with the internal model-based distributed dynamic measurement output feedback control law of interest reduces to the solvability of stabilization problems, including a structured stabilization problem. Then, we leverage our recent results, which cover the existence and design of the corresponding structured control gain, and extend both global and agent-wise local conditions to solve the RCORP via distributed dynamic measurement output feedback. Finally, we provide a numerical example of a MAS comprising three quadrotors and one uninhabited ground vehicle to illustrate the resulting agent-wise local design method.

Keywords: Decentralized control, Distributed control, Communication networks
Abstract: Distributed systems require fusing heterogeneous local probability distributions into a global summary over sparse and unreliable communication networks. Traditional consensus algorithms, which average distributions in Euclidean space, ignore their inherent geometric structure, leading to misleading results. Wasserstein barycenters offer a geometry-aware alternative by minimizing optimal transport costs, but their entropic approximations via the Sinkhorn algorithm typically require centralized coordination. This paper proposes a fully decentralized Sinkhorn algorithm that reformulates the centralized geometric mean as an arithmetic average in the log-domain, enabling approximation through local gossip protocols. Agents exchange log-messages with neighbors, interleaving consensus phases with local updates to mimic centralized iterations without a coordinator. To optimize bandwidth, we integrate event-triggered transmissions and b-bit quantization, providing tunable trade-offs between accuracy and communication while accommodating asynchrony and packet loss. Under mild assumptions, we prove convergence to a neighborhood of the centralized entropic barycenter, with bias linearly dependent on consensus tolerance, trigger threshold, and quantization error. Complexity scales near-linearly with network size. Simulations confirm near-centralized accuracy with significantly fewer messages, across various topologies and conditions.

Keywords: Kalman filtering, Estimation, Energy systems
Abstract: This work investigates the use of dense extended Kalman filter (DEKF) to simultaneously estimate both the state-of-charge (SOC) and equivalent-circuit parameters of serial-connected, lithium-ion battery cells under limited voltage measurement. Stability analyses are conducted to certify the reliability of the DEKF for simultaneous state-parameter estimation under certain mild assumptions. To begin, the local asymptotic stability of the expected dense error in the presence of nonlinearity is verified by selection of an appropriate Lyapunov function. Then the expected sparse error is shown to be marginally stable in the absence of nonlinearities by proving that the spectral radius of the sparse expected error dynamics is exactly equal to 1. Finally, a sufficient condition for sparse stability in the presence of nonlinearity is obtained via selection of an appropriate Lyapunov function. Simulation trials under several different scenarios are executed, confirming that the associated error dynamics are indeed stable for the duration of the simulation time.

Keywords: Neural networks, Optimization, Adaptive control
Abstract: Accurate remaining useful life (RUL) prediction is essential for battery safety and reliability, yet point forecasts can be misleading under nonstationarity and domain shift. We propose QE–RULMamba, a probabilistic, transformer-based state-space forecaster that ensembles multiple experts under a fixed parameter budget and enforces quantile diversity to capture prognostic uncertainty. From capacity forecasts, we compute actionable RUL via a lower-confidence-bound (LCB) search that respects simple temporal and throughput constraints and exposes a tunable risk knob for operations. Evaluated on the NASA and TJU datasets, QE–RULMamba achieves consistently lower MAE/RMSE and higher R^2 than competitive sequence models and a non-quantile ensemble, while yielding calibrated intervals, tight in-distribution and appropriately wider under shift. The resulting confidence-aware RUL curves track oracle behavior in stable regimes and become deliberately conservative when degradation dynamics are uncertain, enabling practical triggers (e.g., early inspection when bounds approach the end-of-life threshold) and underscoring the value of uncertainty-aware forecasting for real-world battery management.

Keywords: Control applications, Automotive control, Hybrid systems
Abstract: This paper proposes a novel multi-objective optimization based energy management system (EMS) for battery/ Supercapacitor (SC) Electric Vehicles (EVs). Two conflicting objectives, viz. minimization of battery state of charge (SOC) degradation and reduction of SC power loss are considered for allocating power optimally between battery and SC. The choice of objective functions is based on the fact that battery is the main energy source of the EV and SC is used for supporting the battery during the journey. The primary objective is to minimize battery degradation and reduce overall system losses by minimizing SC power loss. This will prevent overburdening of the auxiliary energy source i.e. SC and improve system performance. The proposed optimal control problem is solved by applying Dynamic Programming (DP) method. A Pareto front is generated displaying the trade-offs between the two objective functions which can play an important role in benchmarking other real-time control techniques. A throughput based dynamic model is also utilized for studying the battery State of Health (SOH) degradation in varying conditions. Simulation study and numerical investigations are conducted by altering the priority of the two objective functions and performance of the proposed technique is evaluated. The proposed Energy Management System (EMS) is compared with the EMS using Filter based Strategy (FBS) and battery only system.

Keywords: Energy systems, Estimation, Neural networks
Abstract: This paper presents a data-driven framework for accurate battery capacity estimation, addressing challenges of large storage requirements and time-scale discrepancies. The proposed method converts high-frequency time-series data into compact snapshots using a histogram-based data mapping strategy and optimal bin-sizing, reducing data storage significantly. A recurrent neural network (RNN) is employed to model capacity evolution, and the network architecture is optimized using a random search method. Evaluation on an open source NASA dataset achieved less than 1 % validation error, and the benefit scales across other battery chemistries. These results confirm the scalability and adaptability of the method for diverse applications.

Keywords: Energy systems, Optimization algorithms, Machine learning
Abstract: The global transition towards electrified transportation has increased the demand for fast and reliable charging technologies, which are essential for the widespread adoption of electric vehicles. However, enabling rapid charging without accelerating battery degradation remains a major technical challenge. Conventional methods for optimizing charging protocols, such as model-based approaches and grid search, are constrained by inaccurate battery models and the high cost of extensive experimental characterization. Bayesian optimization offers an efficient alternative, as it can learn and optimize unknown objectives with limited sampling. Nevertheless, standard acquisition functions are often designed for general-purpose tasks and may not align with the performance oriented nature of battery fast charging. This research gap motivates the development of the acquisition function better suited to high performance charging applications. We therefore propose a novel acquisition function that prioritizes best case outcomes by identifying and exploring high risk, high reward regions in the objective landscape. Experimental results on the mathematical benchmark function and the PET based battery simulator validate the effectiveness of our proposed approach.

Keywords: Control applications, Power electronics, Automotive systems
Abstract: Multilevel inverters have recently emerged as a promising alternative for traction and stationary applications, offering advantages such as reduced losses and lower current distortion compared to conventional two-level inverters. To fully harness these benefits, suitable control strategies are required to drive the system towards desired operating states, along with a battery management system to preserve the health of the battery pack. In this work, we consider a cascaded multilevel inverter that directly integrates batteries as the energy source and propose a control technique based on a geometric approach. The method simultaneously achieves accurate machine control and balances the state of charge across modules, thereby mitigating non-uniform aging of the battery pack. The effectiveness of the proposed approach is validated through simulation studies.

Keywords: Linear parameter-varying systems, Optimal control, Optimization algorithms
Abstract: Parameter tuning in structured control design faces inherent challenges due to nonconvex feasible sets, sensitivity to initialization, and the need for stabilizing starting points. We revisit this problem by recasting parameter tuning as a structured state-feedback synthesis that expands the gain over sparsity-preserving bases. We formulate a quadratic objective in the state and augmentation terms and develop two augmented-Lagrangian algorithms that efficiently compute optimal parameters in the presence of application-driven bound constraints. This framework accommodates additional convex constraints when needed, delivers high-quality warm starts for subspace gradient methods, and applies broadly whenever the dynamics depend affinely on the unknown parameters. We provide examples that illustrate the effectiveness of our approach and the proposed algorithms in solving parameter tuning and structured control design problems.

Keywords: Linear systems, Computational methods
Abstract: We propose a novel linear algebraic approach to construct a minimal kernel representation of a discrete LTI behavior from an arbitrary kernel representation. We also present an algorithm to generate data for a restricted behavior from its kernel representation. Using this data, we construct certain Hankel matrices and obtain a minimal kernel representation from left kernels of these Hankel matrices via a modified algorithm of [Markovsky and Dorfler, IEEE Transactions on Automatic Control, vol. 68, no. 3, pp. 1667–1677, 2022]. As a consequence, we give a constructive algorithm to compute a minimal basis of a finitely generated module over a single variable polynomial ring.

Keywords: Linear systems, Networked control systems, Identification
Abstract: This work studies the limitations of uniquely identifying the structure (i.e., topology) of a networked linear system from partial measurements of its nodal dynamics. In general, many networks can be consistent with these measurements; this is a consideration often neglected by standard network inference methods. We show that the space of these networks are related through the nullspace of the observability matrix for the true network. We establish relevant metrics to investigate this space, including an analytic characterization of the most structurally dissimilar network that can be inferred, as well as the possibility of mis-inferring presence or absence of edges. In simulations, we find that when observing over 6% of nodes in random network models (e.g., Erdos-Renyi and Watts-Strogatz), approximately 99% of edges are correctly classified. Extending this discussion, we construct a family of networks that keep measurements epsilon-close to each other, and connect the identifiability of these networks to the spectral properties of an augmented observability Gramian.

Keywords: Hybrid systems, Linear systems, Stability of hybrid systems
Abstract: In this work, an output feedback control scheme is proposed for the stabilization of a continuous-time LTI system characterized by non-simultaneous sensing and actuation. Using a Gramian-based framework, a periodically repeating time window is divided into a sensing phase, where the state is reconstructed by filtering the output, and an actuation phase, where the state is driven to a target point balancing the control effort and the stabilization goal. The overall control architecture is cast as a well-posed hybrid dynamical system, providing a convenient tool for proving robust global exponential stability.

Keywords: Linear systems, Predictive control for linear systems, Autonomous systems
Abstract: Robust Controllable (RC) sets enable safe control of dynamical systems under constraints and uncertainty. Existing approaches typically rely on polytopic representations for the computation of these sets, which suffer from conservativeness and scalability issues. Recently, Constrained Convex Generators (CCGs) were proposed to allow set-based control and analysis in presence of both ellipsoidal and polytopic state-input constraints. However, the computation of RC sets using CCGs is currently hindered because the Pontryagin difference set operation has not been developed for CCGs. In this paper, we provide theory and algorithms to address this challenge, and enable safe control under uncertainty using RC sets and CCGs. Specifically, we propose an inner approximation for the RC set using a CCG description. We show in simulations that the proposed approach improves accuracy and memory usage when compared to computations with polytopic approximations.

Keywords: Mechanical systems/robotics, Estimation, Differential-algebraic systems
Abstract: Physics-based simulation models are essential for the development of vapor compression cycles (VCCs), the core technology behind most modern refrigeration and air conditioning systems. These models enable robust control, monitoring, fault detection and diagnostics, and digital twin technologies. However, nonlinear dynamics, high-dimensional parameter and state spaces, numerical stiffness, and the limited integration of conventional modeling environments with scientific machine learning workflows create significant challenges for efficient, transferable, and automated calibration. Existing approaches typically rely on deterministic methods and lack mechanisms for principled knowledge transfer across calibration tasks, while also failing to quantify epistemic and aleatoric uncertainty. To address these limitations, we propose a Bayesian calibration framework for VCC systems that explicitly quantifies various sources of uncertainty in model predictions. The framework supports transferability across datasets by sequentially updating informative priors from previously inferred posteriors. We implement and evaluate this approach on a commercially available high-fidelity Julia based dynamic VCC model, demonstrating its ability to successfully estimate key system parameters.

Keywords: Mechanical systems/robotics, Robotics, Human-in-the-loop control
Abstract: Human-robot collaboration (HRC) requires robots to adapt their motions to human intent to ensure safe and efficient cooperation in shared spaces. Although large language models (LLMs) provide high-level reasoning for inferring human intent, their application to reliable motion planning in HRC remains challenging. Physical human-robot interaction (pHRI) is intuitive but often relies on continuous kinesthetic guidance, which imposes burdens on operators. To address these challenges, a contact-informed adaptive motion-planning framework is introduced to infer human intent directly from physical contact and employ the inferred intent for online motion correction in HRC. First, an optimization-based force estimation method is proposed to infer human-intended contact forces and locations from joint torque measurements and a robot dynamics model, thereby reducing cost and installation complexity while enabling whole-body sensitivity. Then, a torque-based contact detection mechanism with link-level localization is introduced to reduce the optimization search space and to enable real-time estimation. Subsequently, a contact-informed adaptive motion planner is developed to infer human intent from contacts and to replan robot motion online, while maintaining smoothness and adapting to human corrections. Finally, experiments on a 7-DOF manipulator demonstrate the accuracy of the proposed force estimation method with a mean absolute joint-torque estimation error of 0.665 Nm, and verify the effectiveness of the contact-informed adaptive motion planner under perception uncertainty in HRC.

Keywords: Mechanical systems/robotics, Mechatronics, Flexible structures
Abstract: A cable-driven continuum robot arm is an underactuated mechanism and may suffer residual vibration at the end of a rest-to-rest maneuver. In this work, a time-delay filter is applied as an input shaper to the system to eliminate the excitation of vibratory modes. A non-robust and a robust time-delay filter are designed based on a linear system model and demonstrate improved response compared to a velocity-driven pulse input. Experimental results using the continuum robot validate the application of the input shaper, with reduced overshoot and settling time exemplifying the reduction in oscillation at the end of the maneuver. It is also shown that utilizing the robust shaper further improves the response of the arm in comparison to applying the non-robust shaper. These results are significant towards the precise and robust implementation of continuum robots in applications involving arbitrary end-effector trajectories.

Keywords: Mechanical systems/robotics, Hierarchical control, Adaptive control
Abstract: This paper presents a hierarchical model-free control framework for underactuated mechanical systems aimed at generating stable limit cycles. Our proposed control architecture consists of a high-level controller that generates reference trajectories while the low-level controller guarantees accurate tracking of these trajectories on the unactuated coordinates. Our proposed controller has desirable robustness and stability properties and we show its effectiveness through demonstrations on various underactuated mechanical systems such as rotary inverted pendulum (RIP), a segway, an acrobot and a leg-foot model on deformable ground.

Keywords: Mechanical systems/robotics, Robotics, Numerical algorithms

Keywords: Robotics, Mechatronics, Control applications
Abstract: We introduce a new class of attitude control laws for rotational systems; the proposed framework generalizes the use of the Euler axis–angle representation beyond quaternion-based formulations. Using basic Lyapunov stability theory and the notion of extended class K function, we developed a method for determining and enforcing the global asymptotic stability of the single fixed point of the resulting closed-loop (CL) scheme. In contrast with traditional quaternion-based methods, the introduced generalized axis–angle approach enables greater flexibility in the design of the control law, which is of great utility when employed in combination with a switching scheme whose transition state depends on the angular velocity of the controlled rotational system. Through simulation and real-time experimental results, we demonstrate the effectiveness of the developed formulation. According to the recorded data, in the execution of high-speed tumble-recovery maneuvers, the new method consistently achieves shorter stabilization times and requires lower control effort relative to those corresponding to the quaternion-based and geometric-control methods used as benchmarks.

Keywords: Smart grid, Human-in-the-loop control
Abstract: Demand charge, a utility fee based on an electricity customer's peak power consumption, often constitutes a significant portion of costs for commercial electric vehicle (EV) charging station operators. This paper explores control methods to reduce peak power consumption at workplace EV charging stations in a joint price and power optimization framework. We optimize a menu of price options to incentivize users to select controllable charging service. Using this framework, we propose a model predictive control approach to reduce both demand charge and overall operator costs. Through a Monte Carlo simulation, we find that our algorithm outperforms a state-of-the-art benchmark optimization strategy and can significantly reduce station operator costs.

Keywords: Transportation networks, Energy systems, Smart grid
Abstract: The deployment of medium-range battery electric aircraft is a promising pathway to improve the environmental footprint of air mobility. Yet such a deployment would be accompanied by significant electric power requirements at airports due to aircraft charging. Given the growing prevalence of electric vehicles and their bi-directional charging capabilities—so-called vehicle-to-grid (V2G)—we study energy buffer capabilities of parked electric vehicles to alleviate pressure on grid connections. To this end, we present energy management strategies for airports providing cost-optimal apron and landside V2G charge scheduling. Specifically, we first formulate the optimal energy management problem of joint aircraft charging and landside V2G coordination as a linear program, whereby we use partial differential equations to model the aggregated charging dynamics of the electric vehicle fleet. Second, we consider a shuttle flight network with a single hub of a large Dutch airline, real-world grid prices, and synthetic parking garage occupancy data to test our framework. Our results show that V2G at even a single airport can indeed reduce energy costs to charge the aircraft fleet: Compared to a baseline scenario without V2G, the proposed concept yields cost savings of up to 32 %, depending on the schedule and amount of participating vehicles, and has other potential beneficial effects on the local power grid, e.g., the reduction of potential power peaks.

Keywords: Optimization, Multivehicle systems, Large-scale systems
Abstract: The rapid transition to electric mobility requires not only the deployment of charging infrastructure but also the strategic allocation of power within urban electricity networks. While most existing studies address charger siting and network coverage, this work focuses on the complementary problem of how much power should be distributed across different areas of a city — a challenge we define as the Surface Power Allocation Problem (SPAP). We introduce a macroscopic modeling framework in which the urban landscape is partitioned into Urban Surface Units (USUs), each treated as a virtual charging node with aggregated demand. Using an origin–destination graph of mobility flows and a dynamic nonlinear representation of electric vehicle state-of-charge evolution, we estimate spatial power requirements consistent with long-term sustainability objectives. The resulting optimization problem is formulated as a large-scale nonlinear program, which we solve through a relaxation and gradient-based approach to provide computationally efficient solutions.

Keywords: Transportation networks, Optimization algorithms, Game theory
Abstract: The transition to electric mobility calls for charging infrastructure that is both efficient and socially equitable. This paper examines fairness in electric vehicle (EV) charging station pricing and capacity through a game-theoretic perspective. We model a non-cooperative market in which competing charging service providers set prices and capacities while customers choose stations based on generalized cost, leading to a market equilibrium. We then benchmark this decentralized outcome against an idealized planner solution that jointly optimizes efficiency and equity. To align market outcomes with socially desirable goals, we design targeted incentives that guide operators toward more fair charger placement. Case studies demonstrate that unregulated competition tends to exacerbate disparities in charger access across demographic groups, whereas carefully calibrated incentives can reduce inequities without significant efficiency loss. The framework provides insights for policymakers on reconciling free-market dynamics with the broader societal goals of fairness in electrified mobility systems.

Keywords: Manufacturing systems, Hierarchical control, Optimal control
Abstract: Demand-side energy management, such as the real-time pricing (RTP) program, offers manufacturers opportunities to reduce energy costs by shifting production to low-price hours. However, this strategy is challenging to implement when machine degradation is considered, as degraded machines have decreased processing capacity and increased energy consumption. Proactive maintenance (PM) can restore machine health but requires production downtime, creating a challenging trade-off: scheduling maintenance during low-price periods sacrifices energy savings opportunities, while deferring maintenance leads to capacity losses and higher energy consumption. To address this challenge, we propose a hierarchical bi-level control framework that jointly optimizes PM planning and runtime production scheduling, considering the machine degradation. A higher-level optimization, with the lower-level model predictive control (MPC) embedded as a subproblem, determines PM plans that minimize total operational costs under day-ahead RTP. At runtime, the lower-level MPC executes closed-loop production scheduling to minimize energy costs under realized RTP, meeting delivery targets. Simulation results from a lithium-ion battery pack assembly line case study demonstrate that the framework strategically shifts PM away from bottlenecks and high-price hours, meeting daily production targets while reducing energy costs.

Keywords: Optimization, Optimization algorithms, Energy systems
Abstract: This paper presents an optimization framework for scheduling electric vehicle (EV) charging at public stations, with the aim of minimizing overall user dissatisfaction while taking into account arrival time, charging duration, and power demand. To promote fairness, the framework differentiates between high-priority users (those without access to home or workplace charging) and low-priority users (those with such access). In addition, it ensures contiguous charging time slots by incorporating trigger functions into the optimization model. The problem is formulated as a Mixed Integer Linear Program (MILP). The effectiveness of the proposed approach is demonstrated through simulations using EV charger data from the UC Merced campus parking lot.

Keywords: Autonomous systems, Constrained control, Switched systems
Abstract: Control Barrier Functions (CBFs) have been proposed to ensure safety of autonomous systems. This paper considers control policies that switch between CBF constraints. Under this approach, we represent a complex non-convex safe region as a union of sets that are computationally tractable to verify. We denote this framework as union-CBFs and make the following contributions. First, considering switching CBF-QP controllers, we propose a sufficient condition that ensures (i) the system undergoes a finite number of switches in any finite time interval and ensures (ii) the forward invariance of the closed-loop system in between switches. Second, we consider two types of switching strategies and propose union-CBFs conditions for each strategy to satisfy (i) and (ii). Third, we formulate Sum-of-Squares (SOS) algorithms to verify the conditions. The experiments show that our union-CBFs framework results in a larger safe region compared to high-degree polynomial CBFs. We also show the efficiency of the verification algorithms using a polynomial system model.

Keywords: Autonomous systems, Formal verification/synthesis, Hybrid systems
Abstract: Certifying safety for nonlinear systems with polytopic input constraints is challenging because control barrier function (CBF) synthesis must ensure control admissibility under input saturation. We propose an approximation-verification pipeline that performs convex barrier synthesis on piecewise-affine (PWA) surrogates and certifies safety for the original nonlinear system through facet-wise verification. To reduce conservatism while preserving tractability, we use a two-slope Leaky ReLU surrogate for the extended class-K function alpha(.) and combine multiple certificates using Union of Invariant Sets (UIS). Counterexamples are handled through local uncertainty updates. Simulations on pendulum and cart-pole systems with input saturation show larger certified invariant sets than linear-alpha designs while maintaining tractable computation time.

Keywords: Estimation, Autonomous robots, Kalman filtering
Abstract: This paper focuses on multi-robot navigation in GPS-denied environments, where accurate localization is essential for tasks such as environmental monitoring and search-and-rescue. While Visual–Inertial Navigation Systems (VINS) provide lightweight 6-DoF state estimation for single robots, extending them to multi-robot settings introduces the critical challenge of inter-robot frame alignment. Existing approaches often assume ground-truth initialization or overlook this issue, limiting their practicality. We propose a cooperative multi-robot VINS framework that augments the inter-robot transformation as part of the state and jointly estimates it online together with each robot’s trajectory. Crucially, our method leverages common observations as additional constraints, ensuring consistent frame alignment and enabling effective cross-robot fusion. Monte Carlo simulations on 3-D trajectories demonstrate that the proposed framework achieves lower error and improved consistency compared with single-robot baselines, validating its effectiveness for real-world multi-robot localization.

Keywords: Emerging control applications, Mechanical systems/robotics, Intelligent systems
Abstract: Rapid robot motion generation is critical in Human-Robot Collaboration (HRC) systems, as robots need to respond to dynamic environments in real time by continuously observing their surroundings and replanning their motions to ensure both safe interactions and efficient task execution. Current sampling-based motion planners face challenges in scaling to high-dimensional configuration spaces and often require post-processing to interpolate and smooth the generated paths, resulting in time inefficiency in complex environments. Optimization-based planners, on the other hand, can incorporate multiple constraints and generate smooth trajectories directly, making them potentially more time-efficient. However, optimization-based planners are sensitive to initialization and may get stuck in local minima. In this work, we present a novel learning-based method that utilizes a Flow Matching model conditioned on a single-view point cloud to learn near-optimal solutions for optimization initialization. Our method does not require privileged knowledge of the environment, such as obstacle locations and geometries, and can generate feasible trajectories directly from single-view depth camera input. Simulation studies on a UR5e robotic manipulator in cluttered workspaces demonstrate that the proposed generative initializer achieves a high success rate on its own, significantly improves the success rate of trajectory optimization compared with traditional and learning-based benchmark initializers, requires fewer optimization iterations, and exhibits strong generalization to unseen environments.

Keywords: Autonomous systems, Uncertain systems, Optimization
Abstract: Autonomous control systems face significant challenges in executing complex tasks under latent risks—risks arising from occluded agents or partially observable dynamics. To address this, we propose a framework that integrates Large Language Models (LLMs), numerical optimization, and optimization-based control to enable context-aware subtask generation and refinement under latent risks. The framework operates across multiple timescales: (i) in-context learning with LLMs for leveraging past experience, (ii) multi-turn Chain-of-Thought reasoning with numerical optimization for refining subtask parameters, and (iii) real-time feedback control via model predictive control (MPC). Complex tasks are decomposed into subtasks that explicitly account for latent risks, refined through multi-turn optimization in physics-based simulation, and improved over time by accumulating failure-related examples for in-context learning. We validate the framework through simulation-based case studies involving robots and autonomous vehicles using GPT-4o, demonstrating its feasibility and potential to complement MPC with language-driven reasoning for safer decision-making in latent-risk scenarios.

Keywords: Autonomous systems, Robotics
Abstract: Recent advancements in large language models (LLMs) have shown significant promise in various domains, especially robotics. However, most prior LLM-based work in robotic applications either directly predicts waypoints or applies LLMs within fixed tool integration frameworks, offering limited flexibility in exploring and configuring solutions best suited to different tasks. In this work, we propose a framework that leverages LLMs to select appropriate planning and control strategies based on task descriptions, environmental constraints, and system dynamics. These strategies are then executed by calling the available comprehensive planning and control APIs. Our approach employs iterative LLM-based reasoning with performance feedback to refine the algorithm selection. We validate our approach through extensive experiments across tasks of varying complexity, from simple tracking to complex planning scenarios involving spatiotemporal constraints. The results demonstrate that using LLMs to determine planning and control strategies from natural language descriptions significantly enhances robotic autonomy while reducing the need for extensive manual tuning and expert knowledge. Furthermore, our framework maintains generalizability across different tasks and notably outperforms baseline methods that rely on LLMs for direct trajectory, control sequence, or code generation.

Keywords: Data driven control, Machine learning, Constrained control
Abstract: This article presents a novel method for environmental exploration that takes safety into account in unknown areas by using recursive Gaussian process regression (RGPR). Safety in unknown environments is ensured by an RGPR-based control barrier function, which is constructed non-parametrically from the posterior mean of RGPR. Simultaneously, an exploration function inspired by a control Lyapunov function is presented. This function utilizes the posterior variance of RGPR to achieve efficient exploration. Moreover, the use of RGPR mitigates the cubic growth in computational cost inherent to conventional Gaussian process regression, thereby enabling online learning. To validate the effectiveness of the proposed method, an experiment using differential-drive robots equipped with a LiDAR sensor is conducted.

Keywords: Data driven control, Optimal control, Networked control systems
Abstract: This paper studies the resilience of cyber-physical systems under denial-of-service attacks. We develop a novel framework for resilient control that avoids the need for detailed information about the system or attacker dynamics by treating the plant–attacker interaction as an interconnected system. Using small-gain analysis and switching systems theory, we derive explicit resilience conditions, and employ reinforcement learning to synthesize an optimal policy directly from input–state data, estimating the required small-gain bounds in a data-driven manner. A numerical example illustrates the effectiveness of the proposed approach.

Keywords: Data driven control, Optimal control, Predictive control for linear systems
Abstract: This letter presents a robust data-driven receding-horizon control framework for the discrete-time linear quadratic regulator (LQR) with input constraints. Unlike earlier data-driven approaches that design a controller from initial data and apply it unchanged throughout the trajectory, our method exploits all available execution data in a receding-horizon manner, thereby capturing additional information about the unknown system and enabling less conservative performance. Existing data-driven LQR model predictive control methods rely on over-approximations of the consistency set and ell_2 descriptions of noise. In contrast, the proposed approach uses exact descriptions of the consistency set under ell_infty-bounded noise, leveraging duality to recast the problem into a tractable convex optimization. Further, the proposed controller renders the closed-loop system input-to-state stable. Simulation results demonstrate the effectiveness of the method.

Keywords: Data driven control, Predictive control for nonlinear systems, Linear parameter-varying systems
Abstract: Model predictive control is a well established control technology for trajectory tracking. Its use requires the availability of an accurate model of the plant, but obtaining such a model is often time consuming and costly. Data-Enabled Predictive Control (DeePC) addresses this shortcoming in the linear time-invariant setting, by skipping the model building step and instead relying directly on input-output data. Unfortunately, many real systems are nonlinear and exhibit strong operating-point dependence. Building on classical linear parameter-varying control, we introduce DeePC-GS, a gain-scheduled extension of DeePC for unknown, regime-varying systems. The key idea is to allow DeePC to switch between different local Hankel matrices—selected online via a measurable scheduling variable—thereby uniting classical gain scheduling tools with identification-free, data-driven MPC. We test the effectiveness of our DeePC-GS formulation on a nonlinear ship-steering benchmark, demonstrating that it outperforms state-of-the-art data-driven MPC while maintaining tractable computation.

Keywords: Data driven control, Robotics, Predictive control for linear systems
Abstract: This paper introduces a Data-Fused Model Predictive Control (DFMPC) framework that combines physics-based models with data-driven representations of unknown dynamics. Leveraging Willems’ Fundamental Lemma and an artificial equilibrium formulation, the method enables tracking of changing, potentially unreachable setpoints while explicitly handling measurement noise through slack variables and regularization. We provide guarantees of recursive feasibility and practical stability under input–output constraints for a specific class of reference signals. The approach is validated on the iRonCub flying humanoid robot, integrating analytical momentum models with data-driven turbine dynamics. Simulations show improved tracking and robustness compared to a purely model-based MPC, while maintaining real-time feasibility.

Keywords: Compartmental and Positive systems, Data driven control, Constrained control
Abstract: This paper investigates data informativity of positive systems using linear programming (LP). The concept called data informativity represents the sufficiency of a given dataset to solve analysis/design problems. In this paper, we provide the necessary and sufficient conditions for the data-driven analysis and design problems of positive systems to be solvable. Moreover, we clarify that these conditions are characterized by LP problems.

Keywords: Lyapunov methods, Constrained control, Optimal control
Abstract: Control Barrier Functions (CBFs) are widely used in guaranteeing safety for nonlinear systems. They can conservatively map a constrained optimal control problem into a sequence of point-wise Quadratic Programs (QPs) for affine control systems. One of the challenges with the CBF-based point-wise optimization is its myopic property: optimality and safety are only considered at the current time instant. This could make the system overly aggressive and susceptible to infeasibility in the optimization. To address this issue, we propose a sampling method for CBFs to formulate a receding horizon QP. Specifically, at each iteration, we first sample a trajectory based on the current state by a tractable planning method, and then incorporate it into nonlinear dynamics and constraints to obtain piecewise linear approximations that are then enforced by CBFs within the planning horizon. The eventual optimization problem is reformulated into a sequence of receding horizon QPs instead of complex nonlinear programs. We show that the proposed framework can still guarantee safety while staying close to the (optimal) planning trajectory with minimum deviation. We illustrate our approach on traffic merging and robot obstacle avoidance, and compare it with other methods to show its advantage in efficiency and guarantees.

Keywords: Lyapunov methods, Formal verification/synthesis, Neural networks
Abstract: We present a two-stage approach for learning stability-certified neural controllers that achieves a reduction of up to ~95% in training time compared to the state-of-the-art baseline, which introduced monotonic neural Lyapunov architectures. Our method combines monotonic neural Lyapunov functions with fulfillment priority logic (FPL) to efficiently initialize controllers before formal verification. Traditional approaches for jointly learning controllers and neural Lyapunov functions require computationally expensive mixed-integer linear programming (MILP) or satisfiability modulo theory (SMT) solvers at each training iteration, often taking several hours to converge. We address this bottleneck by leveraging FPL to perform early joint initialization of the controller and Lyapunov networks. Building on the monotonic neural network architecture from the baseline, which guarantees non-negativity and a unique global minimum by construction, our method focuses on efficiently satisfying the remaining property of decreasing along trajectories. Existing works focus on maximizing the region of attraction/convergence of the learned controller. In contrast, leveraging FPL allows us to (1) increase learning efficiency substantially and (2) focus on complementary performance metrics, such as convergence rate and control effort minimization, thereby adding significant specification flexibility. In this paper, we encode an approximate Lyapunov-decrease condition in FPL to pre-train the controller and Lyapunov networks, then apply a MILP-based verification/refinement step. This decouples efficient learning from certificate enforcement and allows the FPL specification to include auxiliary objectives (e.g., convergence rate and control effort), whose influence persists through the final MILP pass. The resulting controllers converge rapidly while admitting formal Lyapunov certificates on standard nonlinear control benchmarks.

Keywords: Lyapunov methods, Neural networks, Learning
Abstract: Lyapunov functions are pivotal for analyzing the stability of dynamical systems. While neural networks have emerged as effective approximators, existing methods often struggle with hard-to-fit regions and scale poorly to high-dimensional systems. To address these limitations, this paper introduces the Augmented Lyapunov-Net, a novel framework that integrates a modified CounterExample Guided Inductive Synthesis (CEGIS) architecture. Our framework features a Learner, implemented by the Lyapunov-Net, and an Adversarial Verifier, which actively employs gradient-free optimization to pinpoint states that violate the Lie derivative condition. These strategically discovered counterexamples then guide a focused sampling procedure, enabling the Learner to efficiently refine the candidate function in critical regions. Numerical experiments on several benchmark systems demonstrate that the proposed method achieves higher approximation efficiency to higher-dimensional systems compared to the original Lyapunov-Net.

Keywords: Lyapunov methods, Stability of nonlinear systems, Constrained control
Abstract: This paper presents a control law for stabilization and trajectory tracking of a multicopter subject to safety constraints. The proposed approach guarantees forward invariance of a prescribed safety set while ensuring smooth tracking performance. Unlike conventional control barrier function methods, the constrained control problem is transformed into an unconstrained one using state-dependent mappings together with carefully constructed Lyapunov functions. This approach enables explicit synthesis of the control law, instead of requiring a solution of constrained optimization at each step. The transformation also enables the controller to enforce safety without sacrificing stability or performance. Simulation results for a polytopic reference trajectory confined within a designated safe region demonstrate the effectiveness of the proposed method.

Keywords: Lyapunov methods, Stability of nonlinear systems, Neural networks
Abstract: Deep neural networks (DNNs) are a function approximation tool that are capable of approximating complex dynamic functions by learning complex relationships between the input-output data through training and optimization. Classical DNN training methods utilize numerical optimization tools to improve the function approximation performance. However, in most cases, DNN-based controllers are implemented in an openloop manner, meaning that the DNN model in the controller does not adapt its weights online and therefore fails to take into consideration any unexpected behavior that could hinder system stability and robustness. To tackle this issue, the authors recently leveraged Lyapunov-based methods to develop innovative realtime update laws that train a DNN’s weights in real-time, ensuring stability and improved performance. In this paper, the performance and stability result for DNN-based controllers are further improved by incorporating a Lyapunov-based concurrent learning (CL) inspired term within the real-time DNN update law. A rigorous Lyapunov-based stability analysis was performed to ensure that the proposed DNN-based controller and CL augmented update law results in global exponential convergence of the tracking errors and the DNN weight estimation errors towards an ultimate bound. Thus, for the first time, the proposed approach will ensure both trajectory tracking and convergence of the DNN weights towards their optimal values. Furthermore, simulations were performed that demonstrate the significant potential of the developed DNN-based control system.

Keywords: Lyapunov methods
Abstract: This work presents the design and corresponding analysis of a robust tracking controller for a class of nonlinear systems subject to hysteresis. Specifically, the hysteresis effects are assumed to be in the form of the Bouc-Wen model and despite the lack of accurate knowledge of system dynamics and hysteresis model parameters, a continuous robust controller is designed via the use of the robustness of integral of sign of the error feedback. The stability of the closed--loop system and the convergence of the tracking error are ensured via Lyapunov based arguments. The tracking performance of the proposed control framework is investigated by numerical simulation studies for a second order system.

Keywords: Reinforcement learning, Machine learning, Emerging control applications
Abstract: Large Language Models (LLMs) have demonstrated remarkable capabilities in knowledge acquisition, reasoning, and tool use, making them promising candidates for autonomous agent applications. However, training LLM agents for complex multi-turn task planning faces significant challenges, including sparse episode-wise rewards, credit assignment across long horizons, and the computational overhead of reinforcement learning in multi-turn interaction settings. To this end, this paper introduces a novel approach that transforms multi-turn task planning into single-turn task reasoning problems, enabling efficient policy optimization through Group Relative Policy Optimization (GRPO) with dense and verifiable reward from expert trajectories. Our theoretical analysis shows that GRPO improvement on single-turn task reasoning results in higher multi-turn success probability under the minimal turns, as well as the generalization to subtasks with shorter horizons. Experimental evaluation on the complex task planning benchmark demonstrates that our 1.5B parameter model trained with single-turn GRPO achieves superior performance compared to larger baseline models up to 14B parameters, with success rates of 70% for long-horizon planning tasks with over 30 steps. We also theoretically and empirically validate the strong cross-task generalizability that the models trained on complex tasks can lead to the successful completion of all simpler subtasks.

Keywords: Reinforcement learning, Machine learning, Optimization algorithms
Abstract: We consider a federated reinforcement learning setting involving M agents, all of whom interact with a common Markov Decision Process (MDP). The agents exchange information via a central server to learn the optimal value function. Our goal is to understand to what extent one can hope for collaborative sample-complexity speedups in such a setting, when a small fraction of the agents are adversarial and can act arbitrarily. To that end, we propose Robust Fed-Q, a federated Q-learning algorithm that blends ideas from both model-based and model-free RL, along with the median-of-means device from robust statistics. We prove that despite corruption, with high-probability, Robust Fed-Q (i) guarantees emph{exact} convergence to the optimal value function in the limit of infinite samples, and (ii) enjoys near-optimal finite-time rates that benefit from collaboration. In addition, our approach requires just tilde{O}(1) rounds of communication to achieve each of the above guarantees, a feature of independent interest in FL where communication is the major bottleneck.

Keywords: Reinforcement learning, Markov processes
Abstract: In reinforcement learning (RL) scenarios where agents must preserve privacy about their objectives, deceptive behavior becomes essential to prevent adversarial observers from inferring the agent’s true intentions. Unlike existing approaches that often assume Markovian reward functions or require model-based planning, we propose a model-free deceptive reinforcement learning framework using reward machines (RMs) that operates under adversarial settings. We introduce a belief-induced reward mechanism for deceptive RL using RMs, where agents balance ground truth task completion with entropy maximization over observer beliefs about candidate reward functions. RMs allow us to encode complex temporal dependencies in non-Markovian reward structures and guide the deceptive learning process through structured task representation. Our methods introduce Q-learning and actor-critic algorithms with reward machines to learn optimal deceptive policies that distract observers while maintaining task completion objectives, ensuring effective privacy preservation without compromising primary goals. We establish theoretical guarantees, demonstrating that our algorithms converge to optimal deceptive policies with privacy preservation properties. We further evaluate our methods against baselines in experimental case studies to demonstrate their effectiveness in balancing deception and task performance.

Keywords: Reinforcement learning, Mechanical systems/robotics, Robotics
Abstract: Meta-Reinforcement Learning (Meta-RL) commonly generalizes via smoothness in the task encoding. While this enables local generalization around each training task, it requires dense coverage of the task space and leaves richer task space structure untapped. In response, we develop a geometric perspective that endows the task space with a "hereditary geometry" induced by the inherent symmetries of the underlying system. Concretely, the agent reuses a policy learned at the train time by transforming states and actions through actions of a Lie group. This converts Meta-RL into symmetry discovery rather than smooth extrapolation, enabling the agent to generalize to wider regions of the task space. We show that when the task space is inherited from the symmetries of the underlying system, the task space embeds into a subgroup of those symmetries whose actions are linearizable, connected, and compact--properties that enable efficient learning and inference at the test time. To learn these structures, we develop a differential symmetry discovery method. This collapses functional invariance constraints and thereby improves numerical stability and sample efficiency over functional approaches. Empirically, on a two-dimensional navigation task, our method efficiently recovers the ground-truth symmetry and generalizes across the entire task space, while a common baseline generalizes only near training tasks.

Keywords: Reinforcement learning
Abstract: Multi-agent reinforcement learning with temporal logic presents unique challenges, particularly in competitive settings where agents pursue conflicting objectives. This paper presents NASTL-CIRL (Nash Signal Temporal Logic for Causal Inference in Reinforcement Learning), a novel approach that combines Nash Q-Learning with Causal Signal Temporal Logic (Causal STL) inference to guide agent behavior in competitive environments. Our approach enables agents to infer causal relationships through STL formulas that explain agent behavior and environmental dynamics, providing strategic advantages. Pacman Game scenario and Factory Assembly scenario are conducted to show the superior performance of the proposed NASTL-CIRL algorithm compared to baseline methods. Our results show that integrating causal STL inference with Nash equilibrium concepts leads to more structured and interpretable agent behaviors while maintaining competitive performance.

Keywords: Reinforcement learning, Neural networks, Optimal control
Abstract: This paper develops a novel Lipschitz-aware safe exploration framework for reinforcement learning in environments with abrupt, unmodeled safety variations. Local Lipschitz constants of the safety function are estimated online using kernel density estimation (KDE), providing a data-driven measure of rapid changes in the safety landscape. These estimates are incorporated into a robust quadratic program (QP) with Lipschitz-aware control barrier function (CBF) constraints, yielding a safe exploration law that guarantees forward invariance of an enlarged safe set under probing noise. The exploration mechanism is then coupled with a safety-aware learning stage to obtain a unified safe RL framework. Simulations on an inverted pendulum illustrate the efficacy of the proposed approach.

Keywords: Biological systems
Abstract: A key component of intraneuronal communication is the modulation of postsynaptic firing frequencies by stochastic transmitter release from presynaptic neurons. The time interval between successive postsynaptic firings is called the inter-spike interval (ISI), and understanding its statistics is integral to neural information processing. We start with a model of an excitatory chemical synapse with postsynaptic neuron firing governed as per a classical integrate-and-fire model. Using a first-passage time framework, we derive exact analytical results for the ISI statistical moments, revealing parameter regimes driving precision in postsynaptic action potential timing. Next, we extended this analysis to include both an excitatory and an inhibitory presynaptic connection onto the same postsynaptic neuron. We consider both a fixed postsynaptic-firing threshold and a threshold that adapts based on the postsynaptic membrane potential history. Our analysis shows that the latter adaptive threshold can result in scenarios where increasing the inhibitory input frequency increases the postsynaptic firing frequency. Moreover, we characterize parameter regimes where ISI noise is hypo-exponential or hyper-exponential based on its coefficient of variation being less than or higher than one, respectively.

Keywords: Human-in-the-loop control, Markov processes, Biomedical
Abstract: This work addresses the challenge of designing haptic feedback for high-dimensional motor learning tasks, which is difficult due to the latent evolution of motor skill and the redundancy inherent in high-dimensional control spaces. We propose to formulate the feedback design problem as a Partially Observable Markov Decision Process (POMDP) and to compute an optimal nudging policy. In our framework, the POMDP's latent states represent the evolution of the learner's internal motor skill during a high-dimensional, de-novo (novel) motor learning task performed with a hand exoskeleton. Results from a human study (N=10) on novel motor skill acquisition show faster improvement in task performance in the participant group trained with our POMDP-derived nudging policy than in a group trained with no feedback. These results demonstrate the potential of skill-informed feedback design to accelerate learning and performance.

Keywords: Biomedical, Model Validation, Human-in-the-loop control
Abstract: Abstract--- Functional electrical stimulation (FES) with multi-electrode systems enables more flexible control of muscle activation but is complicated by nonlinear, redundant input–output relationships that often lead to unintended co-activation. We address this as a control allocation problem and introduce a selectivity-aware framework that integrates Dynamic Active Subspaces (DAS) with Model Predictive Control (MPC). By reducing the input space to dominant, task-relevant directions, the method simplifies optimization, reducing the computational intensity of the control problem, while preserving biological richness. We demonstrate the approach using simulations from an invasive sciatic nerve cuff in a pig model.

Keywords: Control of networks, Large-scale systems, Stochastic optimal control
Abstract: Controlling complex, nonlinear systems, such as the brain, presents a fundamental challenge that requires simplified models for practical controller design. Traditional approaches often fail when these systems operate far from steady states under noise and changing inputs. Different control strategies drive systems into distinct behavioral regimes, each requiring a specific approximation. Rather than imposing a single approximation, this work introduces ergodic quasilinearization (EQL), which automatically identifies the appropriate linear model for each operating scenario. EQL generates adaptive linear models whose parameters adjust based on the system's long-term statistical behavior under varying inputs and noise levels. These statistics are derived analytically from the steady-state equalities, eliminating the need for repeated computation of the full nonlinear dynamics. The effectiveness of EQL is demonstrated on large-scale brain network models, where traditional methods encounter difficulties due to complex nonlinearities and high dimensionality. Conventional linearization methods perform well under fixed conditions but lose accuracy when control strategies change the operating regime. In contrast, EQL maintains accuracy across diverse operating scenarios, supporting robust controller design for systems that rarely reach simple steady states. We demonstrate the power of EQL in predicting brain-model responses to complex stimulation protocols and in identifying an optimal open-loop control for reproducing target brain-activity patterns.

Keywords: Healthcare and medical systems, Biomedical
Abstract: Blood pressure (BP) management is a critical component of blood transfusion, but no mature technology capable of predicting BP response to blood transfusion exists. This paper concerns the development and preliminary in vivo testing of a BP prediction method applicable to hemorrhage and blood transfusion. Key obstacles are (i) large inter-individual variability in the BP response to blood transfusion, (ii) unknown hemorrhage, and (iii) input/state-dependent observability. To cope with these challenges, we developed a multi-modal sequential inference-enabled BP prediction method built upon a mathematical model of patient physiology parameterized by population-informed prior. The method infers patient-specific physiological state and hemorrhage, and uses them to predict future BP in a patient receiving blood transfusion. The in vivo testing of the method using the data collected from large animals undergoing hemorrhage and blood transfusion showed that it could adequately predict mean arterial BP with median absolute errors for 5-min and 15-min predictions of 3.1 mmHg and 7.4 mmHg as well as adequately infer physiological state and hemorrhage: all the hemorrhage events were detected with <3.5 min delay, with median F1 score of 85%. In sum, the prediction of BP response to blood transfusion may be feasible, even in the presence of unknown hemorrhage.

Keywords: Constrained control, Biomedical, Neural networks
Abstract: Proton Pump Inhibitors (PPIs) are the standard of care for gastric acid disorders but carry significant risks when administered chronically at high doses. Precise long-term control of gastric acidity is challenged by the impracticality of invasive gastric acid monitoring beyond 72 hours and wide inter-patient variability. We propose a noninvasive, symptom-based framework that tailors PPI dosing solely on patient-reported reflux and digestive symptom patterns. A Bayesian Neural Network prediction model learns to predict patient symptoms and quantifies its uncertainty from historical symptom scores, meal, and PPIs intake data. These probabilistic forecasts feed a chance-constrained Model Predictive Control (MPC) algorithm that dynamically computes future PPI doses to minimize drug usage while enforcing acid suppression with high confidence—without any direct acid measurement. In silico studies over diverse dietary schedules and virtual patient profiles demonstrate that our learning-augmented MPC reduces total PPI consumption by 65% compared to standard fixed regimens, while maintaining acid suppression with at least 95% probability. The proposed approach offers a practical path to personalized PPI therapy, minimizing treatment burden and overdose risk without invasive sensors.

Keywords: Constrained control, Optimization
Abstract: This paper proposes a novel approach to safe navigation in environments with static and dynamic obstacles by embedding control barrier functions (CBFs) within the model predictive control (MPC) framework. Unlike conventional methods that rely on unbounded additive slack variables, the proposed approach enforces each CBF constraint separately, allowing individual flexibility through dedicated slack variables with bounded relaxation weights. These weights modulate the permissible degree of constraint relaxation, ensuring that any safety softening remains quantitatively bounded, systematically tunable, and theoretically consistent with the CBF-based safety guarantees. Furthermore, the feasibility of the proposed approach is guaranteed, and the effectiveness of our method is demonstrated through the simulation results.

Keywords: Constrained control, Lyapunov methods, Nonlinear output feedback
Abstract: First-order lags are commonly used in control system design to approximate system delays, which can significantly compromise system safety. To maintain safety under lags, this paper introduces the concept of lag-compensating control barrier function (LCCBF). First, we propose a lag-eliminating state transformation that removes first-order lag dynamics from linear systems. Then, we use this transformation to construct an LCCBF from the CBF designed for the lag-free system. We prove that controllers synthesized via the LCCBF ensure safety under actuator lags, while preserving feasibility for the same input bounds that are used to synthesize controllers via the original CBF. Finally, we simulate a double integrator with actuator limits to validate that the LCCBF maintains both safety and feasibility under lags.

Keywords: Constrained control, Lyapunov methods, Optimal control
Abstract: In safety-critical control systems, ensuring both system safety and smooth control input is essential for practical deployment. Existing Control Barrier Function (CBF) frameworks, especially High-Order CBFs (HOCBFs), effectively enforce safety constraints, but also raise concerns about the smoothness of the resulting control inputs. While smoothness typically refers to continuity and differentiability, it does not by itself ensure bounded input variation. In contrast, Lipschitz continuity is a stronger form of continuity that not only is necessary for the theoretical guarantee of safety, but also bounds the rate of variation and eliminates abrupt changes in the control input. Such abrupt changes can degrade system performance or even violate actuator limitations, yet current CBF-based methods do not provide Lipschitz continuity guarantees. This paper introduces Filtered Control Barrier Functions (FCBFs), which extend HOCBFs by incorporating an auxiliary dynamic system—referred to as an input regularization filter—to produce Lipschitz continuous control inputs. The proposed framework ensures safety, control bounds, and Lipschitz continuity of the control inputs simultaneously by integrating FCBFs and HOCBFs within a unified quadratic program (QP). Theoretical guarantees are provided and simulations on a unicycle model demonstrate the effectiveness of the proposed method compared to standard and smoothness-penalized HOCBF approaches.

Keywords: Constrained control, Lyapunov methods
Abstract: Guaranteeing safety in the presence of unmatched disturbances---uncertainties that cannot be directly canceled by the control input---remains a key challenge in nonlinear control. This paper presents a constructive approach to safety-critical control of nonlinear systems with unmatched disturbances. We first present a generalization of the input-to-state safety (ISSf) framework for systems with these uncertainties using the recently developed notion of an Optimal Decay CBF, which provides more flexibility for satisfying the associated Lyapunov-like conditions for safety. From there, we outline a procedure for constructing ISSf-CBFs for two relevant classes of systems with unmatched uncertainties: i) strict-feedback systems; ii) dual-relative-degree systems, which are similar to differentially flat systems. Our theoretical results are illustrated via numerical simulations of an inverted pendulum and planar quadrotor.

Keywords: Constrained control, Lyapunov methods
Abstract: This paper investigates the problem of composing multiple control barrier functions (CBFs)---and matrix control barrier functions (MCBFs)---through logical and combinatorial operations. Standard CBF formulations naturally enable conjunctive (AND) combinations, but disjunctive (OR) and more general logical structures introduce nonsmoothness and possibly a combinatorial blow-up in the number of logical combinations. We introduce the framework of combinatorial CBFs that addresses p-choose-r safety specifications and their nested composition. The proposed framework ensures safety for the exact safe set in a scalable way, using the original number of primitive constraints. We establish theoretical guarantees on safety under these compositions, and we demonstrate their use on a patrolling problem in a multi-agent system.

Keywords: Constrained control, Optimization
Abstract: Optimization-based controllers often lack regularity guarantees, such as Lipschitz continuity, when multiple constraints are present. When used to control a dynamical system, these conditions are essential to ensure the existence and uniqueness of the system's trajectory. Here we propose a general method to convert a Quadratic Program (QP) into a Second Order Cone Program (SOCP), which is shown to be Lipschitz continuous. Key features of our approach are that (i) the regularity of the resulting formulation does not depend on the structural properties of the constraints, such as the linear independence of their gradients; and (ii) it admits a closed-form solution under some assumptions, which is not available for general QPs with multiple constraints, enabling faster computation. We support our method with rigorous analysis and examples.

Keywords: Predictive control for linear systems, Constrained control, Optimal control
Abstract: This paper considers the formulation and implementation of model predictive control (MPC) laws with cost functions defined as Minkowski functions of convex, compact sets. We propose constructive procedures for selecting stabilizing terminal elements based on a lambda-contractive set. In addition, we describe a set-based implementation of the exact MPC policy based on constrained zonotopes and a suboptimal MPC policy based on a polytopic set approximation. Our numerical simulations demonstrate that the proposed suboptimal set-based policy achieves performance comparable to the standard trajectory-based optimal control formulation, while requiring less computational effort for low-dimensional systems.

Keywords: Constrained control, Hybrid systems, Lyapunov methods
Abstract: This paper presents discrete control barrier proximal dynamics (D-CBPD), a computationally-light safe control method for sampled-data systems. Ensuring safety in such systems requires bridging the gap between continuous-time control design and discrete implementation. D-CBPD is a discretized contracting dynamics that solves a control barrier function (CBF)-based quadratic program (QP). We first characterize how discretized parametric contracting dynamics track their fixed point and derive an explicit upper bound on the tracking error. We then analyze their use as discrete controllers for continuous-time systems. Leveraging Lipschitz properties, we show that trajectories remain bounded under constant inputs at each interval and exhibit linear growth in passive and weakly contracting systems. This allows us to bound the deviation of the discrete control signal from the optimal continuous-time controller, both at sampling instants and during inter-sample evolution. Interpreting this deviation as an input disturbance enables analysis of the discrete-time controller as a continuous-time system subject to disturbance. Building on this, we introduce D-CBPD, a safe and scalable discrete controller that guarantees safety of a continuously evolving system with a bounded and adjustable violation margin, provide explicit condition for step size to ensure convergence, and demonstrate its effectiveness in thermal management of a simplified lithium-ion battery.

Keywords: Predictive control for nonlinear systems, Constrained control, Formal verification/synthesis
Abstract: Robust model predictive control (MPC) generally relies on Pontryagin difference calculations to tighten constraints. In cases where the dynamics model or disturbances are not known a priori, it is desirable to perform these constraint tightening calculations online. This paper presents a constraint tightening method based on reachability analysis of error dynamics that is tractable to implement online. Constrained zonotopes are used to bound the error dynamics using polyhedral envelopes to account for dynamic nonlinearities. To leverage a recently developed method for inner-approximating Pontryagin differences where the subtrahend is a zonotope, a novel optimization-free method for over-approximating constrained zontopes as zonotopes is presented. The method is shown to consistently produce tighter over-approximations than existing optimization-free methods. The developed online constraint tightening procedure is evaluated in a numerical example where the MPC uses a linear time-varying model.

Keywords: Constrained control, Optimal control, Lyapunov methods
Abstract: Control barrier functions (CBFs) and Hamilton-Jacobi reachability (HJR) are central frameworks in safe control. Traditionally, these frameworks have been viewed as distinct, with the former focusing on optimally safe controller design and the latter providing sufficient conditions for safety. A previous work introduced the notion of a control barrier value function (CB-VF), which is defined similarly to the other value functions studied in HJR but has certain CBF-like properties. In this work, we proceed the other direction by generalizing CBFs to non-differentiable ``viscosity'' CBFs. We show the connection between viscosity CBFs and CB-VFs, bridging the CBF and HJR frameworks. Through this bridge, we characterize the viscosity CBFs as precisely those functions which provide CBF-like safety guarantees (control invariance and smooth approach to the boundary). We then further show nice theoretical properties of viscosity CBFs, including their desirable closure under maximum and limit operations. In the process, we also extend CB-VFs to non-exponential anti-discounting and update the corresponding theory for CB-VFs along these lines.

Keywords: Constrained control, Identification, Pattern recognition and classification
Abstract: Positive invariant (PI) sets are essential for ensuring safety, i.e. constraint adherence, of dynamical systems. With the increasing availability of sampled data from complex (and often unmodeled) systems, it is advantageous to leverage these data sets for PI set synthesis. This paper uses data driven geometric conditions of invariance to synthesize PI sets from data. Where previous data driven, set-based approaches to PI set synthesis used deterministic sampling schemes, this work instead synthesizes PI sets from any pre-collected data set. Beyond a data set and Lipschitz continuity, no additional information about the system is needed. A tree data structure is used to partition the space and select samples used to construct the PI set, while Lipschitz continuity is used to provide deterministic guarantees of invariance. Finally, probabilistic bounds are given on the number of samples needed for the algorithm to determine a PI set of a certain volume.

Keywords: Nonlinear output feedback, Observers for nonlinear systems, Uncertain systems
Abstract: We introduce a guaranteed privacy-preserving controller for nonlinear discrete-time systems with bounded uncertainties. Moving beyond stochastic differential privacy, our design offers deterministic privacy through hard bounds on the proximity of set-valued estimates. The solution involves synthesizing a stabilizing controller for a perturbed framer system, where control gains and a privacy-inducing noise factor are co-optimized via semi-definite programming. This integrated approach ensures both input-to-state stable closed-loop dynamics and certified privacy. We also formalize the inherent performance-privacy trade-off by quantifying the accuracy loss due to privacy constraints. Simulations confirm that our method outperforms standard differential privacy techniques.

Keywords: Optimization, Finance, Stochastic optimal control
Abstract: We consider the problem of optimally allocating a limited number of resources across time to maximize revenue under stochastic demands. This formulation is relevant in various areas of control, such as supply chain, ticket revenue maximization, healthcare operations, and energy allocation in power grids. We propose a bisection method to solve the static optimization problem and extend our approach to a shrinking horizon algorithm for the sequential problem. The shrinking horizon algorithm computes future allocations after updating the distribution of future demands by conditioning on the observed values of demand. We illustrate the method on a simple synthetic example with jointly log-normal demands, showing that it achieves performance close to a bound obtained by solving the prescient problem.

Keywords: Optimization, Hybrid systems, Linear systems
Abstract: Feedback optimization algorithms compute inputs to a system using real-time output measurements, which helps mitigate the effects of disturbances. However, existing work often models both system dynamics and computations in either discrete or continuous time, which may not accurately model some applications. In this work, we model linear system dynamics in continuous time, and we model the computations of inputs in discrete time. Therefore, we present a novel hybrid systems model of feedback optimization. We first establish the well-posedness of this hybrid model and establish completeness of solutions while ruling out Zeno behavior. Then we show the state of the system converges exponentially fast to a ball of known radius about a desired goal state. Next we analytically show that this system is robust to perturbations in (i) the values of measured outputs, (ii) the matrices that model the linear time-invariant system, and (iii) the times at which inputs are applied to the system. Simulation results confirm that this approach successfully mitigates the effects of disturbances.

Keywords: Optimization, Linear systems, Large-scale systems
Abstract: We address resource-aware optimal sensor scheduling for linear gaussian systems over a finite horizon, aiming to minimize the total number of sensor activations while ensuring a target estimation accuracy, measured by the log determinant of the posterior error covariance. Due to the NP hard nature of the problem and the temporal coupling over the time step, prior work has mainly focused on a time-invariant approximation of optimal sensor selection while comparing with greedy scheduling. In this paper, we extend this to time varying sensor scheduling in order to approximate the total number of sensor activations over the finite time horizon. We build a sequential analysis aligned with Kalman recursion and derive a theoretical utilization bound using greedy scheduling under a sequence-supermodular objective. Our simulations across diverse system settings demonstrate that the proposed bound reliably captures sensor usage while ensuring the desired accuracy.

Keywords: Optimization, Optimization algorithms, Lyapunov methods
Abstract: Fixed-point dynamical systems provide a powerful framework for representing and analyzing a wide variety of problems in control, optimization, and learning. Many fundamental tasks—such as solving nonlinear equations, minimizing objective functions, enforcing constraints, and coordinating distributed agents—can be formulated as fixed-point problems. Traditional methods often rely on asymptotic convergence and may not guarantee performance within a finite horizon. In this work, we first establish a framework for analyzing prescribed-time convergence of fixed-point dynamical systems, ensuring that trajectories reach a fixed point within a user-defined and uniformly bounded time, independent of initial conditions. Building on this foundation, we extend the framework to proximal dynamics for equilibrium problems, thereby unifying prescribed-time fixed-point analysis with proximal operator-based methods. Rigorous analysis confirms prescribed-time stability under smooth assumptions, and numerical experiments illustrate the effectiveness of the proposed approach.

Keywords: Optimization, Optimization algorithms, Computational methods
Abstract: In this paper we propose a Zeroth-order gradient descent algorithm for cost functions with Low-rank structure, i.e., the gradients of the cost function exists in a lower-dimensional space. Low-rank structures have emerged as a recurring phenomenon in modern machine learning, deep learning, and recommendation systems. In higher-dimensional problems with costly gradient computations, such Low-rank structures have been exploited to design first-order algorithms with reduced computational cost. We propose a Zeroth-order algorithm to cater for scenarios where the cost function is not explicitly available and only function evaluations are possible by querying a noisy oracle. The algorithm first identifies the lower dimensional subspace where the gradients exists mostly. Then in order to compute an approximate gradient the queries to the oracle are done only in the directions corresponding to this lower dimensional subspace thereby reducing the oracle complexity. We provide convergence analysis for the case of strongly convex and non-convex functions.

Keywords: Optimization algorithms, Optimization
Abstract: We study online Riemannian optimization on Hadamard manifolds under the framework of horospherical convexity (h-convexity). Prior work mostly relies on the geodesic convexity (g-convexity), often leading to regret bounds scaling poorly with the manifold curvature. To address this limitation, we analyze Riemannian online gradient descent for h-convex and strongly h-convex functions and establish O(sqrt{T}) and O(log(T)) regret guarantees, respectively. These bounds are curvature-independent and match the results in the Euclidean setting. We validate our approach with experiments on the manifold of symmetric positive-definite (SPD) matrices equipped with the affine-invariant metric. In particular, we investigate online Tyler's M-estimation and online Fréchet mean computation, showing the application of h-convexity in practice.

Keywords: Delay systems, Machine learning, Lyapunov methods
Abstract: In this work, we present the first stability results for approximate predictors in multi-input non- linear systems with distinct actuation delays. We show that if the predictor approximation satisfies a uniform (in time) error bound, semi-global practical stability is correspondingly achieved. For such approximators, the required uniform error bound depends on the desired region of attraction and the number of control inputs in the system. The result is achieved through transforming the delay into a transport PDE and con- ducting analysis on the coupled ODE-PDE cascade. To highlight the viability of such error bounds, we demonstrate our results on a class of approximators - neural operators - showcasing sufficiency for satisfying such a universal bound both theoretically and in simulation on a mobile robot experiment.

Keywords: Differential-algebraic systems, Numerical algorithms, Simulation
Abstract: This paper investigates the effect of green infrastructures in mitigating air pollution in urban environments. This complex issue needs a thorough examination of traffic-related emissions and the atmospheric dispersion of chemical species. A macroscopic second-order model is employed to simulate traffic conditions, followed by a microscopic approach to estimate emission levels. Finally, partial differential equations-based atmospheric dispersion, driven by chemistry ordinary differential equations for particulate/nitrogen oxides/ozone interactions, handles vertical transport under varying conditions. Numerical results show that different configurations of green barriers yield varying pollutant concentrations in metropolitan regions.

Keywords: Distributed parameter systems, Machine learning, Computer-aided control design
Abstract: The control of high-dimensional distributed parameter systems (DPS) remains a challenge when explicit coarse-grained equations are unavailable. Classical equation-free (EF) approaches rely on fine-scale simulators treated as black-box timesteppers. However, repeated simulations for steady-state computation, linearization, and control design are often computationally prohibitive, or the microscopic timestepper may not even be available, leaving us with data as the only resource. We propose a data-driven alternative that uses local neural operators, trained on spatiotemporal microscopic/mesoscopic data, to obtain efficient short-time solution operators. These surrogates are employed within Krylov subspace methods to compute coarse steady and unsteady-states, while also providing Jacobian information in a matrix-free manner. Krylov–Arnoldi iterations then approximate the dominant eigenspectrum, yielding reduced models that capture the open-loop slow dynamics without explicit Jacobian assembly. Both discrete-time Linear Quadratic Regulator (dLQR) and pole-placement (PP) controllers are based on this reduced system and lifted back to the full nonlinear dynamics, thereby closing the feedback loop. The framework is validated by stabilizing an unstable steady-state of the Liouville–Bratu PDE, demonstrating consistent performance between the learned surrogate and the true system, with quantified degradation under plant-model mismatch.

Keywords: Distributed parameter systems, Identification, Uncertain systems
Abstract: We study an inverse problem arising in the estimation of breath alcohol concentration (BrAC) from transdermal alcohol concentration (TAC) measured by wearable biosensors. The forward model is a parabolic PDE–ODE system describing ethanol diffusion through the skin, with random parameters reflecting inter-individual and environmental variability. To reduce uncertainty in BrAC estimation, we develop a framework that conditions the distribution of skin parameters on subject-specific covariates by exploiting the dependence of BrAC model parameters on those covariates. The key analytical tools are Banach space versions of the Implicit Function Theorem and sensitivity results expressed in terms of Frechet derivatives of parameter dependent semigroups. A central contribution is the rigorous proof of convergence of second derivatives under Galerkin approximation, ensuring that sensitivity and uncertainty quantification computed in finite dimensions faithfully approximate the infinite-dimensional setting. Together, these results establish a mathematically rigorous basis for population calibration and real-time BrAC reconstruction from TAC, with applications to clinical monitoring, addiction treatment, and digital health.

Keywords: Distributed parameter systems, Kalman filtering
Abstract: This paper introduces economic aspects in the state estimator design for a class of infinite dimensional systems. At one end a spatially distributed sensor providing the best possible filter performance and associated with a highly reliable sensor is identified as a very expensive choice. At the other end, a network of inexpensive sensors with pointwise spatial distribution and characterized by reduced reliability are utilized as an alternative to the expensive sensor. Using a modification to centroidal Voronoi Tessellations to arrive at a network of pointwise sensors that collectively approximate the spatially varying expensive sensor, both the total price and the associated filter performance are examined to provide a joint economic and filter performance metric for sensor design in spatially distributed systems.

Keywords: Control applications, Computational methods, Fluid flow systems
Abstract: This paper studies inf-sup stable finite element discretizations for control of viscous incompressible flows with a grad-div stabilization. The proposed grad-div stabilization method augments the mixed Galerkin finite element with a term enhancing mass conservation and improving the performance of the algorithm. We analyse a fully discrete optimality system with grad-div stabilized finite element method. An optimal order error estimate of the approximations of the optimality system is derived. We formulate and solve computationally a control problem that involves velocity matching in a closed cavity. Numerical experiments show the feasibility and applicability of the grad-div stabilized finite element method for control of incompressible flows.

Keywords: Neural networks, Machine learning, Reinforcement learning
Abstract: Learning-based methods for synthesizing controllers have gained popularity due to their high expressiveness and strong empirical performance. However, in safety-critical scenarios such as autonomous driving, robotics, and power systems, empirical performance alone is insufficient, and formal verification of controller properties such as stability and safety is highly desirable. Unfortunately, many prior verification approaches are either tied to specific structural assumptions on the system or certificate, making them difficult to transfer across settings, or suffer from poor scalability on higher-dimensional neural network systems. In this tutorial, we present a unified framework that aims at mitigating this gap via bridging control with the state-of-the-art neural network verifier α,β-CROWN (alpha-beta-CROWN). At its core, α,β-CROWN is a general-purpose bounding engine for general computation graphs: given an input domain, it can automatically produce certified bounds and explicit linear relaxation of the objective function. These certified bounds are useful on their own for tasks such as reachability analysis and local linear approximation, and they also provide the foundation for more complex routines that perform verification, satisfiability checking, and optimization. More specifically, many control problems reduce to verifying real-valued inequalities over a state domain (e.g., Lyapunov theory and barrier functions). Consequently, α,β-CROWN enables scalable verification of such conditions by computing tight bounds on these conditions and recursively partitioning and pruning subdomains based on the bounds. Thanks to GPU parallelization, this pipeline demonstrates superior scalability on verification and optimization problems that are challenging for traditional approaches. In this tutorial, we discuss the basics of α,β-CROWN, and introduce its application to various control-related tasks. Finally, we will discuss training frameworks for co-synthesizing controllers and corresponding verifier-friendly certificates.

Keywords: Communication networks, Flight control, Stability of nonlinear systems
Abstract: In unmanned aerial vehicle (UAV) air-to-air (A2A) communication with rigidly mounted directional antennas, flight-induced attitude variations can cause beam deflections, leading to degraded link quality or even complete signal loss. Existing studies typically assume level flight, thereby overlooking this critical effect. In this paper, we formulate communication reliability as a yaw alignment and pitch-constrained control problem, where yaw regulation ensures beam pointing and pitch is restricted within the half-power beamwidth (HPBW) to preserve link reliability. To address this formulation, we design a time-varying gain command-filtered backstepping controller that achieves accurate trajectory tracking while rigorously enforcing attitude constraints. Theoretical analysis shows that the closed-loop signals converge within a prescribed time and ultimately decay to zero, thus avoiding large transient responses. Simulation results demonstrate that the proposed method sustains stable and reliable communication under representative flight scenarios.

Keywords: Flight control, Robust adaptive control, Fault tolerant systems
Abstract: Autopilots for fixed-wing aircraft typically use a combination of inertial sensors, which include accelerometers and rate gyros, and non-inertial sensors, which include GPS and air-data sensors. By measuring airspeed, angle of attack, and sideslip angle, air-data sensors determine the magnitude and direction of the aircraft velocity vector in the aircraft frame. Since lift and drag are aerodynamic effects, air-data sensors are essential for estimating the aerodynamic forces and moments applied to the vehicle. Unfortunately, air-data sensors are susceptible to failure, with potentially catastrophic consequences. For a fixed-wing flight control system based on adaptive model predictive control, this paper assesses the performance degradation due to the failure of an angle-of-attack sensor.

Keywords: Flight control, Robotics
Abstract: This paper addresses the problem of obstacle nav- igation and path planning for UAVs. This is accomplished by finding a path from start to end points through an obstacle field. To address this problem, a modified Rapidly Exploring Random Trees (RRT) algorithm is used to randomly create a tree of rectangular Safe Flight Corridors (SFCs). When a valid series of SFCs has been found from the start to the goal, a flight path that stays within that series is generated using a Basis Spline (B-Spline), which provides continuous position, velocity, and acceleration control. The path is generated by adjusting the B-Spline control points to minimize the path length objective function with additional constraints on the curvature of the B- Spline. Simulation results demonstrate the effectiveness of this algorithm for UAV path planning.

Keywords: Flight control, Predictive control for nonlinear systems, Robotics
Abstract: This paper presents an attitude tracking control method for unmanned aerial vehicles (UAVs) using a unit-quaternion representation. Initially, a prescribed performance controller (PPC) is developed to ensure closed-loop stability, demonstrating that tracking errors are exponentially converging within predefined bounds. Subsequently, the control performance is enhanced through the application of the Lyapunov-based model predictive control (LMPC). Compared with existing methods, the combination of the LMPC and PPC (i) optimizes the tracking performance, (ii) accounts for design constraints, and (iii) addresses the nonlinearities of the system's dynamics. The recursive feasibility of LMPC is rigorously proved, and the asymptotic stability of the closed-loop system can be shown using Lyapunov analysis. Finally, the applicability and validity of the theoretical results are demonstrated in simulations. To highlight the performance of the LMPC controller, it compares to two baseline controllers, PPC and an output feedback controller (OFC) that is widely studied for UAVs. The simulation results demonstrate the advantages of LMPC in attitude tracking applications. In particular, the combination of LMPC and PPC reduces the transient tracking error by 60.0% compared to OFC and 8.2% compared to PPC and improves the steady-state error by 92.0% compared to OFC and 46.4% compared to PPC.

Keywords: Flight control, Indirect adaptive control, Optimal control
Abstract: This paper presents a centralized predictive cost adaptive control (PCAC) strategy for the position and attitude control of quadrotors. PCAC is an optimal, prediction-based control method that uses recursive least squares (RLS) to identify model parameters online, enabling adaptability in dynamic environments. Addressing challenges with black-box approaches in systems with complex couplings and fast dynamics, this study leverages the unique sparsity of quadrotor models linearized around hover points. By identifying only essential parameters related to nonlinear couplings and dynamics, this approach reduces the number of parameters to estimate, accelerates identification, and enhances stability during transients. Furthermore, the proposed control scheme removes the need for an attitude setpoint, typically required in conventional cascaded control designs.

Keywords: Control over communications, Communication networks, Optimal control
Abstract: Efficient mobile jamming against eavesdroppers in wireless networks necessitates accurate coordination between mobility and antenna beamforming. We study the coordinated beamforming and control problem for a UAV that carries two omnidirectional antennas, and which uses them to jam an eavesdropper while leaving a friendly client unaffected. The UAV can shape its jamming beampattern by controlling its position, its antennas' orientation, and the relative phasing for each antenna. We derive a closed-form expression for the antennas' phases that guarantees zero jamming impact on the client. In addition, we determine the antennas’ orientation and the UAV’s position that maximizes jamming impact on the eavesdropper through an optimal control problem, optimizing the orientation pointwise and the position through the UAV’s control input. Simulations show how this coordinated beamforming and control scheme enables directional GPS denial while guaranteeing zero interference towards a friendly direction.

Keywords: Manufacturing systems, Robotics, Machine learning
Abstract: Remanufacturing is fundamentally more challenging than traditional manufacturing due to the significant uncertainty, variability, and incompleteness inherent in end-of-life (EoL) products. At the same time, it has become increasingly essential and urgent for facilitating a circular economy, driven by the growing volume of discarded electronic products and the escalating scarcity of critical materials. In this paper, we review the existing literature and examine the key challenges as well as emerging opportunities in intelligent automation for EoL electronics remanufacturing, providing a comprehensive overview of how robotics, control, and artificial intelligence (AI) can jointly enable scalable, safe, and intelligent remanufacturing systems. This paper starts with the definition, scope, and motivation of remanufacturing within the context of a circular economy, highlighting its societal and environmental significance. Then it delves into intelligent automation approaches for disassembly, inspection, sorting, and component reprocessing in this domain, covering advanced methods for multimodal perception, decision-making under uncertainty, flexible planning algorithms, and force-aware manipulation. The paper further reviews several emerging techniques, including large foundation models, human-in-the-loop integration, and digital twins that have the potential to support future research in this area. By integrating these topics, we aim to illustrate how next-generation remanufacturing systems can achieve robust, adaptable, and efficient operation in the face of complex real-world challenges.

Keywords: Game theory, Networked control systems, Stochastic systems
Abstract: This paper addresses the challenge of modeling and control in hierarchical, multi-agent systems, known as holonic systems, where local agent decisions are coupled with global systemic outcomes. We introduce the Bayesian Holonic Equilibrium (BHE), a concept that ensures consistency between agent-level rationality and system-wide emergent behavior. We establish the theoretical soundness of the BHE by showing its existence and, under stronger regularity conditions, its uniqueness. We propose a two-time scale learning algorithm to compute such an equilibrium. This algorithm mirrors the system's structure, with a fast timescale for intra-holon strategy coordination and a slow timescale for inter-holon belief adaptation about external risks. The convergence of the algorithm to the theoretical equilibrium is validated through a numerical experiment on a continuous public good game. This work provides a complete theoretical and algorithmic framework for the principled design and analysis of strategic risk in complex, coupled control systems.

Keywords: Game theory, Optimization, Emerging control applications
Abstract: Considering a swarm of Unmanned Aerial Vehicles (UAVs) carrying sensors with nondeterministic detection and noisy localization measurements, while sharing observations with neighboring UAVs, we address the problem of localization of a hidden target in continuous space and discrete time. The goal is to coordinate UAVs to maximize the information gathered while minimizing their individual energy costs. We formulate the problem as a time-varying non-cooperative game with coupling constraints. We show that the target is localized in finite time with probability one and that the game has a generalized potential structure. Further, we provide an exact best-response algorithm for UAVs to iteratively compute their trajectories. Finally, we numerically compare the potential game to the team-based approach, demonstrating comparable performance under different communication graph structures and assessing the impact of the swarm size on various metrics.

Keywords: Game theory, Optimal control, Adaptive control
Abstract: Game-theoretic approaches and Nash equilibrium have been widely applied across various engineering domains. However, practical challenges such as disturbances, delays, and actuator limitations can hinder the precise execution of Nash equilibrium strategies. This work investigates the impact of such implementation imperfections on game trajectories and players' costs in the context of a two-player finite-horizon linear quadratic (LQ) nonzero-sum game. Specifically, we analyze how small deviations by one player, measured or estimated at each stage affect the state trajectory and the other player’s cost. To mitigate these effects, we construct a compensation law for the influenced player by augmenting the nominal game with the measurable deviation dynamics. The resulting policy is shown to be optimal within a causal affine policy class, and, for sufficiently small deviations, it locally outperforms the uncompensated equilibrium-derived feedback. Rigorous analysis and proofs are provided, and the effectiveness of the proposed approach is demonstrated through a representative numerical example.

Keywords: Game theory, Reinforcement learning, Cooperative control
Abstract: Game theory provides the gold standard for analyzing adversarial engagements, offering strong optimality guarantees. However, these guarantees often become brittle when assumptions such as perfect information are violated. Reinforcement learning (RL), by contrast, is adaptive but can be sample-inefficient in large, complex domains. This paper introduces a hybrid approach that leverages game-theoretic insights to improve RL training efficiency. We study a border defense game with limited perceptual range, where defender performance depends on both search and pursuit strategies, making classical differential game solutions inapplicable. Our method employs the Apollonius Circle (AC) to compute equilibrium in the post-detection phase, enabling early termination of RL episodes without learning pursuit dynamics. This allows RL to concentrate on learning search strategies while guaranteeing optimal continuation after detection. Across single- and multi-defender settings, this early termination method yields 10–20% higher rewards, faster convergence, and more efficient search trajectories.

Keywords: Game theory, Reinforcement learning, Iterative learning control
Abstract: Multi-agent reinforcement learning (MARL) optimizes strategic interactions in non-cooperative dynamic games, where agents have misaligned objectives. However, data-driven methods such as multi-agent policy gradients (MA-PG) often suffer from instability and limit-cycle behaviors. Prior stabilization techniques typically rely on entropy-based exploration, which slows learning and increases variance. We propose a model-based approach that incorporates approximate priors into the reward function as regularization. In linear quadratic (LQ) games, we prove that such priors stabilize policy gradients and guarantee local exponential convergence to an approximate Nash equilibrium. We then extend this idea to infinite-horizon nonlinear games by introducing Multi-agent Guided Policy Search (MA-GPS), which constructs short-horizon local LQ approximations from trajectories of current policies to guide training. Experiments on nonlinear vehicle platooning and a six-player strategic basketball formation show that MA-GPS achieves faster convergence and more stable learning than existing MARL methods.

Keywords: Game theory, Smart grid, Optimization
Abstract: This paper studies the Nash equilibrium of a timeof- use electricity pricing game. Existing formulations in this literature are typically solved as Stackelberg problems, with the company posting prices and the user responding by adjusting load. In our prior work, we studied both Stackelberg orders of play and introduced a fairness penalty to prevent opportunistic price increases when the company acts as the follower. Here we formulate the simultaneous-move game directly, establish existence and uniqueness of the Nash equilibrium under the adopted model, derive the players’ best responses, and compare the resulting equilibrium numerically with the two Stackelberg benchmarks.

Keywords: Uncertain systems, Sampled-data control, Machine learning
Abstract: Control Barrier Functions provide a principled framework for ensuring safety, yet their reliance on accurate system models is a critical limitation for practical deployment. Moreover, the majority of theory relies on continuous-time guarantees that may not hold for discrete controllers. To address these challenges, we employ Gaussian processes to learn the components of the CBF safety condition, which depend on the unknown system dynamics, directly from data. Since controllers typically need to be implemented in a sampled-data fashion, we develop robust CBF conditions that account for inter-sampling effects to ensure the safety of unknown control-affine systems. The Gaussian Process model is updated online under a bimodal control framework that switches between control-focused and learning-focused phases, and its uncertainty is leveraged to guarantee satisfaction of the CBF conditions with high probability. The effectiveness of the proposed framework is demonstrated in simulations.

Keywords: Data driven control, Robust control, Game theory
Abstract: We consider the problem of robust control of an unknown but minimal linear time-invariant system under input and output disturbances and adversarial manipulation. An attacker can (i) corrupt sensor measurements (deception attacks) and (ii) perturb the control channel (actuation attacks). To address the lack of model knowledge, we adapt the Data-enabled Predictive Control (DeePC) framework, which constructs predictors directly from input–output data with bounded disturbance. We formulate a finite-horizon open-loop control problem as a two-player zero-sum game with asymmetric information: the defender selects control inputs based on measured data and only knows an upper bound on disturbances, whereas the attacker has access to the true disturbance realization and can remain stealthy by hiding within this uncertainty set. The main contributions are (i) sufficient conditions for the existence of a Nash equilibrium corresponding to saddle-point policies for this game, and (ii) an analysis of the defender’s security strategy against deception and actuation attacks. Simulation studies on finite-horizon control demonstrate the effectiveness of the proposed approach.

Keywords: Network analysis and control, Control of networks, Fault tolerant systems
Abstract: This work studies resilient leader-follower consensus with a bounded number of adversaries. Existing approaches typically require robustness conditions of the entire network to guarantee resilient consensus. However, the behavior of such systems when these conditions are not fully met remains unexplored. To address this gap, we introduce the notion of partial leader-follower consensus, in which a subset of non-adversarial followers successfully tracks the leader’s reference state despite insufficient robustness. We propose a novel distributed algorithm - the Bootstrap Percolation and Mean Subsequence Reduced (BP-MSR) algorithm - and establish sufficient conditions for individual followers to achieve consensus via the BP-MSR algorithm in arbitrary time-varying graphs. We validate our findings through simulations, demonstrating that our method guarantees partial leader-follower consensus, even when standard resilient consensus algorithms fail.

Keywords: Cooperative control, Networked control systems, Agents-based systems
Abstract: This paper presents a novel framework for ensuring safety in dynamically coupled multi-agent systems through collaborative control. Drawing inspiration from ecological models of altruism, we develop collaborative control barrier functions that allow agents to cooperatively enforce individual safety constraints under coupling dynamics. We introduce an altruistic safety condition based on the so-called Hamilton’s rule, enabling agents to trade off their own safety to support higher-priority neighbors. By incorporating these conditions into a distributed optimization framework, we demonstrate increased feasibility and robustness in maintaining system-wide safety. The effectiveness of the proposed approach is illustrated through simulation in a simplified formation control scenario.

Keywords: Networked control systems, Network analysis and control, Control applications
Abstract: We develop a robust safety-critical control framework for networked susceptible–infected–recovered (SIR) dynamics using control barrier functions (CBFs). Each node applies a local intervention to keep infections below a prescribed threshold despite coupling with neighbors and uncertainty in epidemic parameters and measurements. We derive a closed-form nominal CBF controller and extend it to robust CBFs via compensation terms. We study two uncertainty models: a constant bounded disturbance and a low-prevalence–amplified model that increases conservatism when infections are small. Simulations on a small network show nominal safety under low uncertainty and formal robustness under higher uncertainty, with a clear safety–effort trade-off.

Keywords: Statistical learning, Optimization algorithms, Reinforcement learning
Abstract: Safe Bayesian optimization (BO) with Gaussian processes is an effective tool for tuning control policies in safety-critical real-world systems, specifically due to its sample efficiency and safety guarantees. However, most safe BO algorithms assume homoscedastic sub-Gaussian measurement noise, an assumption that does not hold in many relevant applications. In this article, we propose a straightforward yet rigorous approach for safe BO across noise models, including homoscedastic sub-Gaussian and heteroscedastic heavy-tailed distributions. We provide a high-probability bound on the measurement noise via the scenario approach, integrate these bounds into high probability confidence intervals, and prove safety and optimality for our proposed safe BO algorithm. We deploy our algorithm in synthetic examples and in tuning a controller for the Franka Emika manipulator in simulation.

Keywords: Aerospace, Multivehicle systems, Cooperative control
Abstract: This paper introduces the concept of trajectory encryption in cooperative simultaneous target interception, wherein heterogeneity in guidance principles across a team of unmanned autonomous systems is leveraged as a strategic design feature. By employing a mix of heterogeneous time-to-go formulations leading to a cooperative guidance strategy, the swarm of vehicles is able to generate diverse trajectory families. This diversity expands the feasible solution space for simultaneous target interception, enhances robustness under disturbances, and enables flexible time-to-go adjustments without predictable detouring. From an adversarial perspective, heterogeneity obscures the collective interception intent by preventing straightforward prediction of swarm dynamics, effectively acting as an encryption layer in the trajectory domain. Simulations demonstrate that the swarm of heterogeneous vehicles is able to intercept a moving target simultaneously from a diverse set of initial engagement configurations.

Keywords: Aerospace, Cooperative control
Abstract: Consensus over networked agents is typically studied using undirected or directed communication graphs. Undirected graphs enforce symmetry in information exchange, leading to convergence to the average of initial states, while directed graphs permit asymmetry but make consensus dependent on root nodes and their influence. Both paradigms impose inherent restrictions on achievable consensus values and network robustness. This paper introduces a theoretical framework for achieving consensus over a class of network topologies, termed pseudo-undirected graphs, which retains bidirectional connectivity between node pairs but allows the corresponding edge weights to differ, including the possibility of negative values under bounded conditions. The resulting Laplacian is generally non-symmetric, yet it guarantees consensus under connectivity assumptions, to expand the solution space, which enables the system to achieve a stable consensus value that can lie outside the convex hull of the initial state set. We derive admissibility bounds for negative weights for a pseudo-undirected path graph, and show an application in the simultaneous interception of a moving target.

Keywords: Aerospace, Control applications
Abstract: This paper addresses the problem of engaging a maneuvering target using a mobile pursuer equipped with a rotating turret. Unlike conventional point-capture formulations, the objective is to maintain a prescribed standoff distance between the pursuer and the target, while ensuring that the turret’s heading aligns with the instantaneous line of sight (LOS) with respect to the mobile target. The pursuer is modeled as a non-holonomic vehicle with lateral acceleration capability, whereas the turret can have free angular maneuverability. We develop a nonlinear guidance strategy that coordinates the pursuer’s motion with the turret’s rotation to neutralize the target. The proposed method explicitly accounts for the nonlinear coupled kinematics of the pursuer-turret team, ensuring stability and convergence without linearization. Simulation results demonstrate the effectiveness of the guidance strategy against a general maneuvering target.

Keywords: Reinforcement learning, Networked control systems
Abstract: Conventional multi-agent reinforcement learning (MARL) methods rely on time-triggered execution, where agents sample and communicate actions at fixed intervals. This approach is often computationally expensive and communication-intensive. To address this limitation, we propose ET-MAPG (Event-Triggered Multi-Agent Policy Gradient reinforcement learning)-- a framework that jointly learns an agent's control policy and its event-triggering policy. Unlike prior work that decouples these mechanisms, ET-MAPG integrates them into a unified learning process, enabling agents to learn what action to take and when to execute it. For scenarios with inter-agent communication, we introduce AET-MAPG, an attention-based variant that leverages a self-attention mechanism to learn selective communication patterns. AET-MAPG empowers agents to determine when to trigger an action, and also with whom to communicate and what information to exchange, thereby optimizing coordination. Both methods can be integrated with any policy gradient MARL algorithm. Extensive experiments across diverse MARL benchmarks demonstrate that our approaches achieve performance comparable to state-of-the-art, time-triggered baselines while significantly reducing both computational load and communication overhead.

Keywords: Optimization, Optimization algorithms, Autonomous systems
Abstract: A cooperative routing problem involving two vehicles, a sensing agent and a relay agent, is considered. The sensing agent must visit a given set of targets to collect data and transmit it to a ground station. Hardware limitations restrict the transmission range, requiring the second agent to act as a relay between the sensing agent and the ground station. The collected data is transmitted through a two-hop network: from the sensing agent to the relay agent, and then to the ground station. Given a set of target locations, the routing problem aims to determine the optimal sequence of targets to be visited by the sensing agent, along with the corresponding positions of the relay agent, to minimize the combined travel cost of both vehicles. This problem is formulated as a mixed-integer convex program utilizing the framework of graphs of convex sets. The formulation avoids ``big-M" constraints, resulting in a tighter convex relaxation. Furthermore, this approach is shown to be applicable to the neighborhood traveling salesman problem and can be extended to a three-agent scenario with two relay agents. Computational experiments are conducted to evaluate the proposed formulations and validate the tightness of the lower bounds obtained via convex relaxation.

Keywords: Game theory, Optimal control, Information theory and control
Abstract: We examine a Tactical Asset Allocation Game (TAAG) in which two opposing players attempt to place offensive and defensive assets to optimize their respective expected value upon initiation of an engagement. The Defender player selects an initial Target position to maximize the expected distance between the Target and a Penetrator agent. The Defender also positions an Interceptor to intercept the Penetrator agent and, incurring a corresponding cost, selects an obfuscation tactic for the sensor network. Based on the sensor data, the Attacker then selects an incursion point for the Penetrator with the objective of minimizing that same expected distance between Target and Penetrator. The values of the possible engagements emerging from the placement strategies are determined using a variation of the Target-Penetrator-Interceptor (TPI) game. The TAAG is then resolved by utilizing a linear program based on the matrix game established by combinations of different allocation and deception opportunities and further demonstrates the potential of differential games to be utilized in solving more complex problems in control and game theory.

Keywords: Optimization, Energy systems, Control applications
Abstract: Simultaneous peak shaving for multi-load microgrids is an important problem with significant financial impact on industrial and commercial energy users. Cost savings can be realized by strategically reducing peak demand across multiple loads through the use of solar photovoltaic (PV) and battery energy storage systems (BESS). A mixed integer linear programming (MILP) optimization problem formulation for PV-BESS multi-load dispatch is outlined. It serves as a standard against which live controller performance can be compared. A heuristic dispatch control strategy is developed leveraging the nature of the 15 minute average demand billing structure. The practical controller has no need of complex live weather and demand forecasting and is developed for a multi-load application. Industrial electric utility data from a manufacturing facility in Houston, Texas is used for testing of the heuristic control strategy. Results show net savings within as near as 3.3% when comparing the heuristic controller and MILP on the industrial data.

Keywords: Control applications, Energy systems
Abstract: This work explores methods to identify energy system designs for infeasible control co-design optimization problems. Control co-design, or CCD, has been recognized as a powerful tool to maximize energy system capabilities through simultaneous determination of plant and controller parameters. However, due to the inherent nonlinearities, complexity, and conflicting criteria of energy systems, CCD optimization problems are susceptible to infeasibility and can lack potential solutions. While transforming the optimization problem by relaxing constraints has been developed for optimal control infeasibility challenges, solution feasibility for CCD is relatively unexplored. This paper proposes a framework to convert infeasible optimization problems into solvable forms for a class of CCD problems. The framework introduces a procedure to rank metric bounds from least likely to most likely to cause infeasibility. This provides guidance to algorithmically relax a limited number of constraints, leaving others intact. The proposed framework is applied to a CCD problem for designing a battery within a microgrid. Comparison against a baseline approach for relaxing optimization problems shows the framework requires only a reduced number of iterations to determine a solution.

Keywords: Energy systems, Machine learning, Optimal control
Abstract: Fast charging accelerates lithium-ion battery degradation and can trigger lithium plating, posing serious safety risks. While existing charging strategies such as Constant-Current Constant-Voltage (CCCV) or equivalent circuit model (ECM)-based charging methods are computationally efficient, they lack access to internal electrochemical states and cannot enforce degradation-aware safety limits. This work proposes a hybrid Model Predictive Control (MPC) framework that augments an ECM with a Gaussian Process (GP) surrogate trained on high-fidelity Doyle-Fuller-Newman (DFN) simulations. The key novelty lies in a GP surrogate that estimates lithium plating overpotential at the anode-separator interface, enabling real-time enforcement of physically meaningful safety constraints without the computational burden of electrochemical models. Simulation results on an LG-M50 cell show that the proposed GP-ECM MPC reduces plating violations by 85-95% compared to CCCV, ECM-MPC, and SPMe-MPC methods, with less than a 10% increase in charging time and only 7 ms of additional computation per optimization step. These results demonstrate a scalable, safety-conscious fast-charging strategy suitable for practical deployment in future battery management systems.

Keywords: Energy systems, Control applications
Abstract: This paper presents a unified optimization framework for phase change material (PCM) based cooling systems. Thermal management is critical in applications such as photovoltaic (PV) modules, battery packs, and power electronics, where excessive heat reduces performance and lifespan. Designing such systems is challenging because energy dynamics, capacity, heat rejection, and structural constraints must all be considered. Although prior studies have investigated PCM applications and heat transfer enhancement, there are limited efforts that unify such diverse performance objectives through formalized design methods. This paper develops a framework that formulates the PCM design problem using critical energy-based terms, with static and dynamic objectives capturing the PCM physical design and control aspects. Two case studies are used to validate the approach: the first explores passive cooling, and the second implements an active cooling configuration. The results compare the design and control of these systems, showing improvement in individual performance metrics between the two options.

Keywords: Power systems, Reinforcement learning, Stochastic optimal control
Abstract: We study the optimal dynamic wireless charging (DWC) for electric vehicles in coupled power and transportation systems with stochastic renewable energy supply. We formulate the problem as a Markov decision process (MDP) that captures the interdependencies between traffic flows, charging demand, and grid operations. Our key contribution is a path-dependent threshold mechanism that reduces the high-dimensional charging decision space by exploiting the structural properties of optimal policies along vehicles’ paths. To approximate optimal thresholds, we develop a customized deep reinforcement learning framework that achieves efficient learning through intelligent action space reduction. The proposed method dynamically adjusts charging decisions by jointly considering traffic conditions, renewable availability, and grid capacity constraints. Experiments on realistic traffic-grid scenarios demonstrate up to 44% operational cost reduction compared to state-of-the-art benchmarks while reducing training time by 22%.

Keywords: Power systems, Smart grid, Optimization
Abstract: This paper studies the strategic behavior of a monopolistic storage aggregator in electricity markets, focusing on how coordinated operation of geographically dispersed grid-scale storage units can deliberately alter power flows to create artificial transmission congestion, i.e., congestion driven by strategic actions rather than fundamental supply-demand patterns. We model the interaction between the aggregator, who earns from both energy arbitrage and financial transmission rights (FTRs), and the system operator as a Stackelberg game. In a two-bus setting, we analytically characterize equilibrium outcomes and identify necessary and sufficient conditions for artificial congestion, expressed as a threshold on the total energy that the aggregator can strategically mobilize. We further show that FTR ownership strongly shapes incentives: Congestion is more likely when the aggregator owns/longs FTRs, but can be avoided if the aggregator is required to short them. For general networks, we present methods to solve for the equilibrium of the Stackelberg game and screen for artificial congestion.

Keywords: Linear systems, Predictive control for linear systems, Biologically-inspired methods
Abstract: Wing-assisted inclined running (WAIR) observed in some young birds, is an attractive maneuver that can be extended to legged aerial systems. This study proposes a control method using a modified Variable Length Inverted Pendulum (VLIP) by assuming a fixed zero moment point and thruster forces collocated at the center of mass of the pendulum. A QP MPC is used to find the optimal ground reaction forces and thruster forces to track a reference position and velocity trajectory. Simulation result of this VLIP model on a slope of 40 degrees is maintained and shows thruster forces that can be obtained through posture manipulation. The simulation also provides insight to how the combined efforts of the thrusters and the tractive forces from the legs make WAIR possible in thruster-assisted legged systems.

Keywords: Linear systems, Robust control, Observers for Linear systems
Abstract: This paper investigates bounds on the estimation error of a linear system affected by norm-bounded disturbances and full sensor attacks. The system is equipped with a detector that evaluates the norm of the innovation signal to detect faults, and the attacker wants to avoid detection. We utilize induced L∞ system norms, also called peak-to-peak norms, to compare the estimation error bounds under nominal operations and under attack. This leads to a sufficient condition for when the bound on the estimation error is smaller during an attack than during nominal operation. This condition is independent of the attack strategy and depends only on the attacker's desire to remain undetected and (indirectly) the observer gain. Therefore, we investigate both an observer design method, that seeks to reduce the error bound under attack while keeping the nominal error bound low, and detector threshold tuning. As a numerical illustration, we show how a sensor attack can deactivate a robust safety filter based on control barrier functions if the attacked error bound is larger than the nominal one. We also statistically evaluate our observer design method and the effect of the detector threshold.

Keywords: Linear systems, Switched systems, Behavioural systems
Abstract: In this paper, we consider the problem of directly synthesizing a control input from the past input-output data to achieve precision tracking for switched linear systems. To this end, we proposed a data-driven direct input synthesis controller that generates the control input from the past input-output data directly. An initialization process is designed to collect the information of transient response, and it's shown that when the past data is rich enough, i.e., satisfying the persistently excitation condition, the transient response can be precisely calculated. The control input is generated by the data-driven direct input synthesis control method, the tracking performance is guaranteed under the finite preview condition, and the tracking error is bounded by the summation of a monotonic increasing function, {alpha}_1(cdot), and a monotonic decreasing function, {alpha}_2(cdot). The proposed method is illustrated through a numerical simulation on a switched linear system.

Keywords: Linear systems
Abstract: This paper presents a graph-based procedure for computing the generic number of finite poles and zeros. The two key ingredients are the path-cycle families and the separator-based decomposition. The former generically encodes degree and valuation of minors. The latter gives an equivalent system matrix that eases system zeros characterization within graphs. Using graph connectivity properties, the generic number of zeros located at s=0 are computed. The remaining finite structure is then derived.

Keywords: Observers for Linear systems, Robust control, Uncertain systems
Abstract: This paper proposes a linear input-output observer design methodology for a population of systems in which each observer uses knowledge of the linear time-invariant dynamics of the particular device. Observers are typically composed of a known model of the system and a correction mechanism to produce an estimate of the state. The proposed design procedure characterizes the population variation in the frequency domain and synthesizes a single robust correction filter. The correction filter guarantees a level of estimation performance for all systems compatible with the uncertainty characterization. This is accomplished by posing a robust performance problem using the observer error dynamics and solving it using DK-iteration. The design procedure is experimentally demonstrated on a flexible joint robotic manipulator with varied joint stiffnesses. It is shown that the proposed robust correction filter achieves comparable estimation performance to a method using a correction gain tailored toward each joint stiffness configuration.

Keywords: Output regulation, Linear systems, Constrained control
Abstract: This paper addresses robust output regulation for linear systems driven by exogenous signals represented as composite Bernstein/Bézier polynomials. Building on prior work for single-segment signals, we develop a continuous internal-model-based controller that guarantees exponential convergence of the regulation error without hybrid switching laws. We rigorously characterize steady-state trajectories induced by composite polynomials and analyze transient errors at segment transitions. Based on this analysis, we propose a replanning strategy for future segments that mitigates switching effects while satisfying state and input constraints. Numerical simulations on navigation tasks demonstrate the effectiveness of the approach.

Keywords: Model/Controller reduction, Large-scale systems, Computational methods
Abstract: We develop the theoretical formulation for a non-intrusive, quadrature-based method for approximate balanced truncation (QuadBT) of linear systems with quadratic outputs, thus extending the applicability of QuadBT, which was originally designed for data-driven balanced truncation of standard linear systems with linear outputs only. The new approach makes use of the time-domain and frequency-domain quadrature-based representation of the system's infinite Gramians, only implicitly. We show that by sampling solely the extended impulse responses (kernels) of the original system and their derivatives (or the corresponding transfer functions), we construct a reduced-order model that mimics the approximation quality of the intrusive (projection-based) balanced truncation. Although the sampling of the required kernels via input/output simulations or physical experiments is still an open question, we demonstrate a proof of concept for the proposed framework on an example using numerically evaluated data.

Keywords: Modeling, Automotive control, Supervisory control
Abstract: Human driver modeling is fundamental for analyzing connected mixed traffic, where human-driven and autonomous vehicles will coexist, and real-time advisory guidance is increasingly available. Traditional car-following models capture physical interactions with the lead vehicle but often fail to represent how drivers integrate advisory speed inputs into their decision-making. This paper proposes an optimization-based human driver model that imitates car-following behavior while accounting for advisory speed guidance. The model combines front-vehicle states with advisory information through a confidence mechanism that reflects driver trust, allowing it to reproduce both compliance with and rejection of advisories. Model parameters are optimized offline using naturalistic data collected in a mixed-reality digital twin driving simulator, enabling precise calibration to individual driving styles. Experimental results in congestive traffic scenarios show that the proposed model achieves sub–0.5 m/s trajectory error and outperforms the existing naive models.

Keywords: Modeling, Control applications, Estimation
Abstract: Interactive Multiple Model (IMM) has been widely applied in motion tracking tasks due to its capability in handling multiple macroscopic behavior hypotheses. In the context of human-driven vehicle motion tracking, drivers often exhibit diverse and personalized driving preferences, which challenge conventional IMM algorithms that rely on predefined models. Nevertheless, the adaptation of multiple models to match observations may induce an overfitting issue, which compromises the ability to identify the real behavior. To address these challenges, this paper presents a regularized online adaptation approach against overfitting for IMM-based vehicle motion tracking. We introduce an advanced motion modeling framework in the Frenet coordinate frame, which naturally decouples longitudinal and lateral motions. Then, a differentiable parametric motion policy is incorporated for enhancing the model adaptability based on simultaneous state and parameter estimation using extended Kalman filtering (EKF). Furthermore, the overfitting problem is addressed by imposing regularization constraints on the model parameters in the filtering correction step based on the Alternating direction method of multipliers (ADMM) iteration. In simulation, the proposed method shows advantages in accurate motion tracking while remaining computationally efficient.

Keywords: Modeling, Lyapunov methods, Optimal control
Abstract: The ability to manipulate and interlace cables using aerial vehicles can greatly improve aerial transportation tasks. Such interlacing cables create hitches by winding two or more cables around each other, which can enclose payloads or can further develop into knots. Dynamic modeling and control of such hitches are key to mastering inter-cable interactions in the context of cable-suspended aerial manipulation. This paper introduces an ellipsoid-based kinematic model to connect the geometric nature of a hitch created by two cables and the dynamics of the hitch driven by four aerial vehicles, which reveals the control-affine form of the system. As the constraint for maintaining tension of a cable is also control-affine, we design a quadratic programming-based controller that combines Control Lyapunov and High-Order Control Barrier Functions (CLF-HOCBF-QP) to precisely track a desired hitch position and system shape while enforcing safety constraints like cable tautness. We convert desired geometric reference configurations into target robot positions and introduce a composite error into the Lyapunov function to ensure a relative degree of one to the input. Numerical simulations validate our approach, demonstrating stable, high-speed tracking of dynamic references.

Keywords: Reduced order modeling, Stochastic systems, Uncertain systems
Abstract: Polynomial chaos expansions provide surrogate models for stochastic systems, with coefficients typically derived using Galerkin projection, stochastic collocation, or least squares approximation. These traditional approaches often fail to accurately capture statistical moments without resorting to high-order approximations. We propose a constrained optimization framework that modifies standard techniques to determine polynomial chaos coefficients that precisely recover the first two statistical moments. The effectiveness of our approach is demonstrated on several candidate algebraic functions of random variables, showing significant improvements in statistical accuracy even with low-order approximations.

Keywords: Mechatronics, Mechanical systems/robotics, Switched systems
Abstract: This paper describes a control algorithm and its implementation within a model predictive framework for a special motion rectifier power take-off (PTO). Commonly used for wave energy converters, the motion rectifier PTO has advantages in preventing unnecessary generator velocity reversals. However, control of this type of PTO is hard to design, especially when direct control optimization is needed in model predictive control (MPC). A modified dynamic programming (DP) algorithm is then developed to perform open-loop optimization of both the generator torque and the PTO switching. This DP optimizer is embedded inside a rolling horizon MPC style controller to enable real-time control of the PTO. Due to the sequential nature of the DP algorithm, unidirectional power constraint can be easily enforced in the optimization, allowing power evaluation of a passive power conversion system, which can use simpler components but is traditionally challenging for optimal control. Simulation results demonstrate the effectiveness of the proposed control algorithm. It is found that the rectifier PTO increases power by 120% for the selected large generator inertia, and unidirectional power flow constraint only decreases power by 8%. Finally, a hardware-in-loop experiment is conducted to demonstrate the control algorithm in real-time. Experiment results show a 50% power increase by enabling active switching control while keeping the generator damping fixed.

Keywords: Energy systems
Abstract: Effective control strategies are essential for maximizing the power output of wave energy converters. This study applies port-based modeling for biconjugate impedance matching to the HERO WEC (hydraulic and electric reverse osmosis wave energy converter), a small-scale, modular point absorber for water desalination, exploring the benefits and limitations for this type of reactive control with a hydraulic system. Despite its foundational role in electrical engineering, biconjugate impedance matching has had limited use in wave energy and has not yet been demonstrated in wave-powered desalination systems. Representing this complex multidisciplinary system as an equivalent circuit enables analysis of frequency-based relations. This paper discusses important considerations of port-based modeling for hydraulic systems and identifies optimal parameters to reduce impedance mismatch by 99.93%, providing a clear framework for biconjugate impedance matching in wave-powered water desalination systems.

Keywords: Energy systems, Predictive control for nonlinear systems, Reduced order modeling
Abstract: This paper describes a multi-region control framework for floating offshore wind farms. Specifically, we propose a novel generator torque controller that regulates rotor speed in Region 2, corresponding to wind speeds between the cut-in and rated values. In Region 3 (wind speeds at or above rated but below cut-out speed) we employ a PI-LQR for collective blade pitch. Control blending across the transitional wind speeds (Region 2.5) employs a sigmoid weighting function applied to the control variables. Two modeling paradigms are proposed for farm-level power tracking with rotor speed regularization: a nonlinear model predictive controller (NL-MPC) with a dynamic wake model, and a reduced order model predictive controller based on linear parameter varying turbine models with a time delay representation of wake advection (LPVTD-MPC). These approaches are evaluated over three wind inlet conditions using the PJM ancillary service certification criteria for participation in a secondary frequency regulation market. Results show that both approaches achieve scores of at least 89.9% for the three different testing scenarios, which are well above the qualification threshold of 75%. However, the LPVTD-MPC approach solves the problem in under half the time versus NL-MPC but with slightly larger fluctuations in farm-level power output, highlighting the trade-off between performance and computational tractability. The control framework is among the first to address multi-region wind turbine dynamics together with market driven power tracking objectives for floating offshore wind farms. Such multi-region control becomes increasingly necessary in the floating turbine setting where large (region spanning) wind speed variations are common due to wave induced platform pitching.

Keywords: Energy systems, Maritime control, Emerging control applications
Abstract: This paper presents a hierarchical approach for optimal mission planning of a mobile, anchorless wave-powered mobile desalination device in a spatiotemporally varying environment. The goal is to fill an on-board freshwater tank in minimum time by strategically targeting areas of maximum wave energy resource. In such an environment, the complexity of mission planning lies in the varying nature of the wave resource. In this paper, we first present the use of indirect methods to derive a continuous heading trajectory for the system. Owing to the fact that this optimization problem is ill-posed under many circumstances, we separate the problem into the optimization of a discrete set of waypoints and an effective speed at which to proceed between waypoints (while the system has very limited control over its absolute speed, it can "zig-zag" in order to achieve reduced speeds when advantageous). Here, an upper-level optimizer iteratively uses a genetic algorithm (GA) to generate populations of waypoints, and a lower-level optimizer fully optimizes the effective speed trajectories for each member of the population. We present a series of simulation results for a specified device design, based on wave energy data from the Hawaiian islands.

Keywords: Control applications, Reinforcement learning, Observers for nonlinear systems
Abstract: Core power control of pressurized water reactor (PWR) is challenging as it involves highly nonlinear dynamics characterized by time-varying system parameters, modeling uncertainties, internal and external disturbances. Linear extended state observer (LESO) in the framework of active disturbance rejection control (ADRC) offers a simple yet robust paradigm for control of uncertain systems under diverse disturbances. This work proposes a maximum sensitivity constrained region based graphical approach, termed graphical LESO (GLESO), to tune the controller parameters of linear ADRC. Furthermore, a Q-learning based strategy, called Q-GLESO, is developed to adaptively tune the controller parameters between scaled multiples of those obtained from GLESO. Effectiveness of proposed schemes is demonstrated through comprehensive simulations on the nonlinear mathematical model of PWR which illustrate the superiority of Q-GLESO over GLESO and LESO for various tracking and disturbance rejection test scenarios.

Keywords: Autonomous systems, Optimization, Control applications
Abstract: Many robotic systems must follow planned paths yet pause safely and resume when people or objects intervene. We present an output–space method for systems whose tracked output can be feedback-linearized to a double integrator (e.g., manipulators). The approach has two parts. Offline, we perform a pre-run reachability check to verify that the motion plan respects speed and acceleration magnitude limits. Online, we apply a quadratic program to track the motion plan under the same limits. We use a one-step reachability test to bound the maximum disturbance the system is capable of rejecting. When the state coincides with the reference path we recover perfect tracking in the deterministic case, and we correct errors using a KKT-inspired weight. We demonstrate that safety stops and unplanned deviations are handled efficiently, and the system returns to the motion plan without replanning. We demonstrate our system's improved performance over pure pursuit in simulation.

Keywords: Autonomous vehicles, Observers for nonlinear systems, Sensor fusion
Abstract: Accurate and robust attitude estimation is a central challenge for autonomous vehicles operating in GNSS-denied or highly dynamic environments. In such cases, Inertial Measurement Units (IMUs) alone are insufficient for reliable tilt estimation due to the ambiguity between gravitational and inertial accelerations. While auxiliary velocity sensors, such as GNSS, Pitot tubes, Doppler radar, or visual odometry, are often used, they can be unavailable, intermittent, or costly. This work introduces a barometer-aided attitude estimation architecture that leverages barometric altitude measurements to infer vertical velocity and attitude within a nonlinear observer on SO(3). The design cascades a deterministic Riccati observer with a complementary filter, ensuring Almost Global Asymptotic Stability (AGAS) under a uniform observability condition while maintaining geometric consistency. The analysis highlights barometer-aided estimation as a lightweight and effective complementary modality.

Keywords: Autonomous systems, Predictive control for linear systems, Networked control systems
Abstract: In this paper, the problem of safely driving an autonomous vehicle for long range missions despite malicious actions on the data shared with the remote side is considered. This operating scenario poses two key questions: the capability of the underlying control algorithm to be feasible until the mission is accomplished; the effectiveness of adequate countermeasures to contrast the anomalous dynamical system behavior resulting from unpredictable attack phenomena. These requirements are here addressed by designing a layered control architecture that takes advantage of three methodologies: model predictive control, data encryption and perturbation analysis.

Keywords: Autonomous systems, Information theory and control, Sensor fusion
Abstract: We study autonomous path planning in an unknown, time-varying forest fire threat field, where an ego vehicle must minimize temperature exposure using partial observations from small unmanned aerial vehicle (SUAV) sensors. Threat estimates are obtained indirectly from overhead imagery processed by a U-Net segmentation model, producing noisy, biased measurements of the evolving field.

To address this problem, we apply an Actively Coupled Sensor Configuration and Path-planning (A-CSCP) framework that jointly optimizes sensor placement and ego trajectory. Four sensor reconfiguration strategies are compared: random placement, lawnmower sweep, standard mutual information (SMI), and context-relevant mutual information (CRMI). To remain tractable for the large state dimension, SMI and CRMI are implemented in approximate diagonal form. Simulation results show that CRMI provides the best performance, eliminating exposure and reducing mission duration by aligning sensing actions with the ego vehicle’s future path. These results demonstrate the practical value of approximate information-driven sensing and planning in dynamic hazardous environments.

Keywords: Autonomous systems, Vision-based control, Nonlinear output feedback
Abstract: Small unmanned aircraft must avoid mid-air collisions under strict size, weight, and power limits that constrain sensing and estimation. We propose a sensor-minimal collision-avoidance controller that uses only monocular bearing, apparent angular diameter, and the sign of bearing rate, without estimating range, time-to-collision, or relative velocity. A constructive geometric proof establishes forward invariance of the safety set and finite-time recovery under realistic kinematic and control bounds for planar encounters. Monte Carlo trials (10,000 per scenario) across intruder sizes from small UAS to Boeing 737, including non-spherical silhouettes, achieved 100% success for constant-velocity encounters and remained robust to pixel-level noise and modest intruder maneuvers. The results indicate that camera-level cues alone can support provably safe and computationally lightweight avoidance for resource-constrained aircraft.

Keywords: Autonomous systems, Optimization, Estimation
Abstract: We study informative path planning (IPP) with travel budgets in cluttered environments, where an agent collects measurements of a latent field modeled as a Gaussian process (GP) to reduce uncertainty at target locations. Graph-based solvers provide global guarantees but assume pre-selected measurement locations, while continuous trajectory optimization supports path-based sensing but is computationally intensive and sensitive to initialization in obstacle-dense settings. We propose a hierarchical framework with three stages: (i) graph-based global planning, (ii) segment-wise budget allocation using geometric and kernel bounds, and (iii) spline-based refinement of each segment with hard constraints and obstacle pruning. By combining global guidance with local refinement, our method achieves lower posterior uncertainty than graph-only and continuous baselines, while running faster than continuous-space solvers (up to 9x faster than gradient-based methods and 20x faster than black-box optimizers) across synthetic cluttered environments and Arctic datasets.

Keywords: Robotics, Nonlinear systems identification, Optimal control
Abstract: This paper presents a data-driven model predictive control framework for mobile robots navigating in dynamic environments, leveraging Koopman operator theory. Unlike the conventional Koopman-based approaches that focus on the linearization of system dynamics only, our work focuses on finding a global linear representation for the optimal path planning problem that includes both the nonlinear robot dynamics and collision-avoidance constraints. We deploy extended dynamic mode decomposition to identify linear and bilinear Koopman realizations from input–state data. Our open-loop analysis demonstrates that only the bilinear Koopman model can accurately capture nonlinear state–input couplings and quadratic terms essential for collision avoidance, whereas linear realizations fail to do so. We formulate a quadratic program (QP) for the robot path planning in the presence of moving obstacles in the lifted space and determine the optimal robot action in an MPC framework. Our approach is capable of finding the safe optimal action 320 times faster than a nonlinear MPC counterpart that solves the path planning problem in the original state space. Our work highlights the potential of bilinear Koopman realizations for linearization of highly nonlinear optimal control problems subject to nonlinear state and input constraints to achieve computational efficiency similar to linear problems.

Keywords: Identification, Stochastic systems, Estimation
Abstract: System identification and Koopman spectral analysis are crucial for uncovering physical laws and understanding the long-term behaviour of stochastic dynamical systems governed by stochastic differential equations (SDEs). In this work, we propose a novel method for estimating the Koopman generator of systems of SDEs, based on the theory of resolvent operators and the Yosida approximation. This enables both spectral analysis and accurate estimation and reconstruction of system parameters. The proposed approach relies on only mild assumptions about the system and effectively avoids the error amplification typically associated with direct numerical differentiation. It remains robust even under low sampling rates or with only a single observed trajectory, reliably extracting dominant spectral modes and dynamic features. We validate our method on two systems and compare it with existing techniques as benchmarks. The experimental results demonstrate the effectiveness and improved performance of our approach in system parameter estimation, spectral mode extraction, and overall robustness.

Keywords: Identification for control, Lyapunov methods, Optimization
Abstract: Although Koopman operators provide a way to represent autonomous nonlinear systems as linear operators in a high-dimensional space, they do not immediately apply to non-autonomous systems with inputs. State (or output) feedback controller design therefore remains nonconvex in typical formulations, even after the use of existing approximations via bilinear control-affine terms. We address this gap by introducing the Koopman Control Parametrization, a novel parameterization of control-affine dynamical systems that directly includes a feedback controller defined as a linear combination of nonlinear measurements. With this choice, the Koopman operator of the closed-loop system is a bilinear combination of two matrices: one representing the system, and the other the controller. We propose a set of sufficient conditions such that the parametrization holds. Then, we present an algorithm that calculates the feedback matrix via semi-definite programming, producing a Lyapunov-stable closed-loop system with convex optimization. We evaluate the proposed controllers on two canonical examples of control-affine nonlinear systems (inverted pendulums), showing that our approach successfully stabilize both with a proper choice of basis functions. In summary, this manuscript introduces a control synthesis method for stabilization of nonlinear systems that is broadly applicable, can be computed efficiently, is verifiably stable and data-driven.

Keywords: Observers for nonlinear systems, Machine learning, Nonlinear systems identification
Abstract: The signal of system states needed for feedback controllers is estimated by state observers. A generic observer design is the Kazantzis--Kravaris/Luenberger (KKL) observer, an extension of the Luenberger observer for linear to nonlinear systems. The main challenge in the KKL design is the construction of an injective mapping of states, which requires solving PDEs based on a first-principles model. This paper proposes a data-driven, Koopman operator-based method for the construction of KKL observers for planar limit cycle systems. Specifically, for such systems, the KKL injection is guaranteed to be a linear combination of Koopman eigenfunctions. Hence, the determination of such an injection is reduced to a least-squares regression problem, and the inverse of the injective mapping is then approximated via kernel ridge regression. The entire synthesis procedure uses solely convex optimization. We apply the proposed approach to the Brusselator system, demonstrating accurate estimations of the system states.

Keywords: Predictive control for nonlinear systems, Flight control, Learning
Abstract: This work investigates the trajectory tracking of quadrotors using data-driven model predictive control (DDPC). We first propose a novel deep learning-enabled Koopman operator to establish the linear representation of the quadrotor systems. Rather than employing dynamic mode decomposition, this approach leverages the physics-informed neural network (PINN) to learn both observable functions and model parameters, thereby integrating the strengths of both model-based and model-free techniques. As a result, the bluetext{proposed PINN-based deep Koopman operator} can accurately capture both the dynamic behaviours and physical constraints of actual quadrotor systems. In addition, we develop a DDPC strategy based on the learned Koopman model to stabilize the quadrotor without relying on prior knowledge of the system dynamics. The proposed deep Koopman-based MPC strategy (DK-MPC) yields a quadratic programming formulation, enabling efficient computational solutions. Finally, two numerical examples and the comparison study are provided to verify the efficacy of the proposed method.

Keywords: Nonlinear systems identification, Modeling, Identification
Abstract: This paper proposes the parameter-interpolated extended dynamic mode decomposition (piEDMD) method that explicitly incorporates parameter dependence into the Koopman operator formulation by exploiting parameter-affine system structure. The approach enables interpolation across parameter values without retraining, providing a pathway toward generalizable surrogate models. The proposed method is evaluated on two nonlinear systems, namely a damped pendulum and a Van der Pol oscillator, demonstrating accuracy comparable to standard extended dynamic mode decomposition (EDMD) while eliminating the need for repeated retraining. In testing, the standard EDMD achieved an average RMSE of 8.9% for the pendulum and 11.11% for the Van der Pol oscillator, while the piEDMD achieved 7.8% and 12.69%, respectively.

Keywords: Nonholonomic systems, Autonomous robots, Feedback linearization
Abstract: This paper develops an invariant-set motion-planner (ISMP) for vehicles with unicycle-like dynamics. ISMPs operate by determining a sequence of obstacle-free positively invariant (PI) sets for the closed-loop system that connect the initial and target state. In this paper, we derive PI sets under a dynamic feedback linearization law. We demonstrate that the PI sets are simple in geometry and we provide an efficient algorithm for obstacle-free scaling. The PI sets are then used to develop a graph-based search for finding an obstacle-free path between two states. The approach is demonstrated in an obstacle-rich simulated environment.

Keywords: Nonholonomic systems, Lyapunov methods, Adaptive control
Abstract: The recent development of globally strict control Lyapunov functions (CLFs) for the challenging unicycle parking problem provides a foundation for pursuing optimality. We address this in the inverse optimal framework, thereby avoiding the need to solve the Hamilton–Jacobi–Bellman (HJB) equations, and establish a general result that is optimal with respect to a meaningful cost. We present several design examples that impose varying levels of penalty on the control effort, including arbitrarily bounded control. Furthermore, we show that the inverse optimal controller possesses an infinite gain margin thanks to the system being driftless, and by leveraging this property, we extend the design to an adaptive controller that handles model uncertainty. Finally, we compare the performance of the non-adaptive inverse optimal controller with its adaptive counterpart.

Keywords: Nonholonomic systems, Lyapunov methods, Constrained control
Abstract: This paper presents a framework for stabilizing the Dubins vehicle model to zero in finite time (deadbeat parking) by interpreting distance as a time-like variable. We develop control laws that bring the system to a desired position and orientation even when the forward velocity cannot be directly actuated. While the controllers employ inverse-distance gains, we show that the control input remains bounded for all time. In addition to basic deadbeat parking, we incorporate safety considerations by proposing algorithms that prevent the vehicle from crossing in front of the target, enforce deceleration as it approaches the target, and guarantee parking without curb violations. The resulting methods are well-suited for missile guidance and fixed-wing pursuit, however they are broadly applicable to physical systems that are represented by the Dubins vehicle model.

Keywords: Nonholonomic systems, Lyapunov methods
Abstract: Since the mid-1990s, it has been known that, unlike in Cartesian form where Brockett’s condition rules out static feedback stabilization, the unicycle is globally asymptotically stabilizable by smooth feedback in polar coordinates. In this note, we introduce a modular framework for designing smooth feedback laws that achieve global asymptotic stabilization in polar coordinates. These laws are bidirectional, enabling efficient parking maneuvers, and are paired with families of strict control Lyapunov functions (CLFs) constructed in a modular fashion. The resulting CLFs guarantee global asymptotic stability with explicit convergence rates and include barrier variants that yield “almost global” stabilization, excluding only zero-measure subsets of the rotation manifolds. The strictness of the CLFs is further leveraged in our companion paper, where we develop inverse-optimal redesigns with meaningful cost functions and infinite gain margins.

Keywords: Nonholonomic systems, Lyapunov methods
Abstract: In a companion paper, we present a modular framework for unicycle stabilization in polar coordinates that provides smooth steering laws through backstepping. Surprisingly, the same problem also allows application of integrator forwarding. In this work, we leverage this feature and construct new smooth steering laws together with control Lyapunov functions (CLFs), expanding the set of CLFs available for inverse optimal control design. In the case of constant forward velocity (Dubins car), backstepping produces finite-time (deadbeat) parking, and we show that integrator forwarding yields the very same class of solutions. This reveals a fundamental connection between backstepping and forwarding in addressing both the unicycle and, the Dubins car parking problems.

Keywords: Optimal control, Adaptive control, Nonholonomic systems
Abstract: This paper presents a three-dimensional extremum seeking (ES) control framework for autonomous source localization in environments containing obstacles. Unlike conventional path planning methods that rely on prior maps or GPS, our approach enables a nonholonomic vehicle to seek an unknown source using only local measurements, without explicit position information. We adopt a Rimon–Koditschek navigation function to simultaneously encode the objective of source attraction and obstacle repulsion, allowing the ES controller to operate safely in complex 3-D environments. Theoretical analysis based on averaging theory establishes exponential convergence to an ellipsoidal region around the extremum point. Two simulation scenarios are conducted to validate the effectiveness of the proposed method: one with a single ellipsoidal obstacle and another involving multiple convex obstacles. The results demonstrate that the vehicle can successfully reach the source while avoiding collisions. This framework offers a feasible solution for GPS-denied applications such as rescue robotics and environmental monitoring.

Keywords: Reinforcement learning, Optimal control, Identification for control
Abstract: This paper develops a convex-optimization-based inverse reinforcement learning (CO-IRL) framework that infers all cost-functional weights and the optimal policy for linear-quadratic regulator (LQR) control systems in both discrete and continuous time. The framework incorporates adaptive dynamic programming (ADP) to eliminate the need for explicit knowledge of the system dynamics. It solves the inference problem as an iterative CO problem under appropriate convexity constraints, enabling low-cost weight recovery subject to constraints and jointly learning cost weights (for both states and controls) and value functions for policy improvement, resulting in faster convergence. We first present a model-based variant, then extend it to a model-free, data-driven version by constructing a Q-function. Convergence is proven, and simulations show substantially faster convergence than non-optimization-based IRL.

Keywords: Reinforcement learning, Optimal control, Optimization
Abstract: In this paper, we introduce ``unlearning''-based, data-driven defenses against poisoning attacks on the learning of linear quadratic controllers. We first establish conditions for detecting whether a learned controller has been compromised. We then develop defense strategies that either reconstruct the system matrix to recover the unattacked Riccati solution or solve optimization problems that adjust output and cost matrices to neutralize the attack’s influence. Simulation results confirm the effectiveness of the proposed methods in restoring optimal control performance.

Keywords: Reinforcement learning, Robotics, Lyapunov methods
Abstract: In this work, we consider the problem of executing multiple tasks encoded by value functions, each learned through Reinforcement Learning, using an optimization-based framework. Prior works develop this framework but did not address when learned value functions can be concurrently executed. This work’s main contributions consist of theorems which provide necessary and sufficient conditions to concurrently execute sets of learned tasks within subsets of the state space using the previously proposed min-norm controller. These theorems provide insight into when learned control tasks can be made concurrently executable, when they may already be so, and when concurrent execution is not possible under the proposed framework. We also extend the proposed framework to account for value functions trained with a discount factor, making it more compatible with standard RL practices.

Keywords: Reinforcement learning, Robust control, Autonomous systems
Abstract: This paper presents a reinforcement learning-based path following controller for a fixed-wing small uncrewed aircraft system (sUAS) that is robust to uncertainties in the aerodynamic model of the sUAS. The controller is trained using the Robust Adversarial Reinforcement Learning framework, where an adversary perturbs the environment (aerodynamic model) to expose the agent (sUAS) to demanding scenarios. In our formulation, the adversary introduces rate-bounded perturbations to the aerodynamic model coefficients. We demonstrate that adversarial training improves robustness compared to controllers trained using stochastic model uncertainty. The learned controller is also benchmarked against a switched uncertain initial condition controller. The effectiveness of the approach is validated through high-fidelity simulations using a realistic six-degree-of-freedom fixed-wing aircraft model, showing accurate and robust path following performance under a variety of uncertain aerodynamic conditions.

Keywords: Reinforcement learning, Time-varying systems, Control applications
Abstract: In this paper, we study the use of robust model-independent bounded extremum seeking (ES) feedback control to improve the robustness of deep reinforcement learning (DRL) controllers for a class of nonlinear time-varying systems. DRL has the potential to learn from large datasets to quickly control or optimize the outputs of many-parameter systems, but its performance degrades catastrophically when the system model changes rapidly over time. Bounded ES can handle time-varying systems with unknown control directions, but its convergence speed slows down as the number of tuned parameters increases and, like all local adaptive methods, it can get stuck in local minima. We demonstrate that together, DRL and bounded ES result in a hybrid controller whose performance exceeds the sum of its parts with DRL taking advantage of historical data to learn how to quickly control a many-parameter system to a desired setpoint while bounded ES ensures its robustness to time variations. We demonstrate the generality of our combined ES-DRL controller approach with numerical studies of three very different dynamic systems: 1) a general time-varying system, 2) automatic tuning of the Low Energy Beam Transport section at the Los Alamos Neutron Science Center linear particle accelerator, and 3) an intermittent-contact robotic block pushing task with a time-varying goal.

Keywords: Neural networks, Reinforcement learning, Optimal control
Abstract: This paper presents a resilient reinforcement learning (RL)-based safety-aware optimal adaptive tracking framework for nonlinear discrete-time affine systems under state constraints and adversarial sensor, actuator, and reward attacks. Sensor and actuator attacks are modeled through networked communication, while reward attacks affect the reward function or temporal difference error (TDE). A multilayer neural network (MNN)-based actor-critic architecture estimates the cost and optimal policy, using a clipped TDE and an adaptive Gaussian-based forgetting factor to balance resilience and optimality. Sensor and actuator attack resilience is achieved via a fault-aware cost-to-go function and a control barrier function (CBF) constraint embedded in a quadratic program (QP). The effectiveness is validated on a rear-wheel-drive vehicle.

Keywords: Systems biology, Cellular dynamics, Biomedical
Abstract: Mathematical models of chronic inflammatory diseases such as psoriasis often capture biological detail but overlook the regulatory feedback structures that sustain pathological persistence and shape therapeutic response. We develop a predictive multi-scale framework that links systemic immune–skin interactions with gut microbiome feedback, extending beyond conventional psoriasis models that focus narrowly on local cytokine pathways. This formulation reveals that disease trajectories may be governed by a small set of nonlinear motifs that generate multistability-consistent with observed remission, relapse, and residual inflammation. Sensitivity and bifurcation analyses identify keratinocyte decay and probiotic-associated microbial growth as critical parameters for steering the system away from "diseased" equilibria. Leveraging these insights, we design a hybrid intervention strategy that combines impulsive phototherapy with continuous probiotic supplementation. Numerical solutions show that this approach may reduce treatment burden, accelerate remission, and be adaptable to heterogeneous initial conditions, highlighting opportunities for tailoring therapy to patient-specific conditions. Beyond psoriasis, the framework establishes a scalable foundation for analyzing coupled microbial–immune systems and for designing therapeutic strategies that exploit underlying feedback structures.

Keywords: Biomedical, Modeling, Robotics
Abstract: Maintaining balance during unexpected foot slip is a challenging task for bipedal locomotion. Simplified motion models from linear inverted pendulum (LIP) have demonstrated as an effective tool for recovery gait planning. However, the existing used LIP models often neglect whole-body angular momentum for slip-and-fall recovery design. We propose to use an angular-momentum LIP with flywheel (ALIP-F) model to incorporate two control inputs, center of pressure (CoP) and whole-body angular momentum, into balance recovery design. It is shown that the angular momentum provides a comparable balance recoverable actuation as the CoP control. The model analysis and comparison explain and quantify the use of angular momentum in bipedal balance recovery under foot slip. We further conduct human slip-and-fall experiments to validate the model analysis and properties. This work provides modeling and analysis methods for understanding slip recovery and designing robust biped balance recovery control.

Keywords: Biological systems, Stochastic systems, Modeling
Abstract: Regulation of cell size requires coordination between growth and division to maintain stable size distributions despite intrinsic stochastic fluctuations. Experiments have established the adder principle as a robust mechanism of size homeostasis, where cells add a nearly constant amount of size each cycle, independent of birth size. Theoretical studies suggest that this behavior emerges when cell cycle progression is coupled with cell growth. However, human cells may exhibit growth dynamics that deviate from the classical exponential assumption. Motivated by this, a stochastic framework for size regulation in human cells has been developed by incorporating a Hill-type growth law into the adder-based model to capture growth saturation. To represent the sequential nature of division, a stochastic multi-step model is implemented where cells progress through internal regulatory stages before dividing. This framework yields exact analytical expressions for statistical moments of the steady-state cell size distribution. Results show that stronger growth saturation increases the mean cell size while slightly reducing fluctuations compared to exponential growth. Importantly, the adder property is preserved, indicating that the reduced variability is a consequence of the growth law rather than simple scaling with mean size. Finally, we show how growth saturation affects variability at the population level during stochastic clonal proliferation.

Keywords: Biomedical, Fluid power control, Variable-structure/sliding-mode control
Abstract: Static friction, also known as stiction, is a primary source of tracking error in pneumatic cylinders, which significantly limits their use in high-precision force control tasks. This paper presents a detailed experimental study of the static friction behavior of a small pneumatic cylinder, examining how the friction force depends on piston position and supply pressure. Based on these findings, a new static friction model that considers pressure and position is developed to accurately represent the stiction phenomenon. Additionally, a Sliding Mode Controller (SMC) is designed and implemented to produce a desired output force. The SMC's performance is evaluated experimentally both with and without the proposed friction compensation model. Results show that the SMC with the integrated compensation reduces steady-state force tracking errors compared to the uncompensated controller.

Keywords: Reinforcement learning, Optimal control, Biological systems
Abstract: Designing effective chemotherapy regimens is hindered by tumor heterogeneity and drug resistance, which complicate the deployment of patient-specific model-based optimal control across diverse populations. We develop and compare closed-loop deep reinforcement learning (DRL) dosing policies with continuous (TD3) and discrete (DQN) action spaces trained on a high-dimensional heterogeneous tumor model. The DRL policies are benchmarked against a Pontryagin's Maximum Principle (PMP)-derived open-loop benchmark. We assess generalization under parametric heterogeneity using a 100-patient virtual cohort with 10% uniform perturbations in growth and drug-sensitivity parameters. Across this cohort, TD3 achieves higher average tumor reduction, while DQN yields tighter inter-patient dosing consistency, revealing a clear efficacy-consistency trade-off in this study. Our simulations assume full observation of all tumor subpopulations; translation to sparse and noisy clinical measurements will require partial-observability formulations and/or state estimation. Overall, the results show that simulation-trained DRL can learn state-dependent feedback dosing policies that complement open-loop optimal control benchmarks.

Keywords: Biomedical, Identification for control, Neural networks
Abstract: Accurate prediction of blood glucose trajectories is essential for safe automated insulin delivery and clinical decision support in type 1 diabetes. Yet existing data-driven methods based on historical patient data perform poorly in predicting long-term outcomes. This work presents a fractional-order neural network (FONN) framework for the identification and control of blood glucose dynamics in patients affected by type~1 diabetes. The proposed approach integrates Grünwald–Letnikov operators into the neural network architecture, enabling the inclusion of long-term memory effects into the learning process. In open-loop system identification, the FONN is benchmarked against its integer-order ablation (IONN), and a Long Short-Term Memory (LSTM) model. On a widely used in-silico simulator, LSTM attains the best short-horizon forecasts, whereas FONN yields higher accuracy as the prediction horizon increases. Closed-loop evaluations in the context of data-driven model predictive control confirm the advantages of the FONN compared to a linear model-based formulation and the LSTM approach itself, with improved performance in terms of time in normoglycemia and euglycemia.

Keywords: Constrained control, Lyapunov methods, Autonomous systems
Abstract: Verifying the safety of controllers is critical for many applications, but is especially challenging for systems with bounded inputs. Backup control barrier functions (bCBFs) offer a structured approach to synthesizing safe controllers that are guaranteed to satisfy input bounds by leveraging the knowledge of a backup controller. While powerful, bCBFs require solving a high-dimensional quadratic program at run time, which may be too costly for computationally-constrained systems such as aerospace vehicles. We propose an approach that optimally interpolates between a nominal controller and the backup controller, and we derive the solution to this optimization problem in closed form. We prove that this closed-form controller is guaranteed to be safe while obeying input bounds. We demonstrate the effectiveness of the approach on a double integrator and a nonlinear fixed-wing aircraft example.