Federated Learning (FL) has emerged as a promising framework for privacy-preserving distributed learning by enabling model training across multiple agents (clients) without exchanging raw data. However, FL remains vulnerable to inference attacks and faces challenges in scalability, robustness, and efficiency, especially in wireless and decentralized environments. In this paper, we propose a novel distributed multiagent wireless federated learning algorithm with client-level local differential privacy, where agents communicate over wireless channels and aggregate model updates using wireless channel-aware computation. By explicitly modeling the wireless channel and incorporating artificial noise, each agent performs a consensus-based gradient estimation using its own and its neighbors’ updates. We rigorously prove that under a strongly connected topology and suitable learning-rate conditions, convergence is guaranteed. Client-level differential privacy is further ensured via power control at the channel level. Our approach offers a robust and scalable solution for mission-critical applications such as healthcare and autonomous driving in adversarial environments. Comprehensive simulation results validate the theoretical convergence properties, analyze the privacy-utility tradeoff, and empirically validate privacy protection against membership inference attacks. Results confirm that the distributed algorithm achieves performance comparable to centralized federated learning while providing stronger privacy guarantees and improved fault tolerance.

The main objective of this paper is to present a new algorithm framework to deal with distributed multiagent control subject to cost function uncertainties. Multiagent control has evolved significantly, with recent advances focusing on rigorous convergence analysis and diverse dynamical models. While the design of multiagent control with performance guarantee has also been developed, it generally assumes perfect knowledge of cost functions, which is often unrealistic due to the intrinsic difficulty of translating the designer’s goal into the signals related to instantaneous costs in some complex tasks, and uncertainties from disturbances, model inaccuracies, or adversarial interactions. This paper addresses optimal cooperative control for multiagent systems with uncertain cost functions. We propose a novel distributed policy gradient method that incorporates a scenario-based approach to handle uncertainties by sampling and integrating a practical distributed estimation algorithm to reduce computational load. Our algorithm ensures convergence while minimizing the global cost function using only local information exchange. The proposed approach bridges the gap between theoretical multiagent control and practical applications.

Accurate trajectory prediction is crucial in autonomous driving to ensure safe and efficient navigation, yet effectively modeling complex interactions among multiple agents remains a significant challenge. Many existing methods still suffer from over-reliance on HD maps, high computational cost, and a lack of interpretability in interaction reasoning. In response, the proposed model innovatively incorporates counterfactual reasoning into social interaction modeling to tackle the challenges of interaction-aware multi-agent trajectory prediction, prioritizing both accuracy and efficiency. In light of the spatiotemporal interaction mechanism and the inherent human cognition governing agents’ motion, the approach simultaneously generates multi-modal trajectories for all agents in a scenario, providing a novel perspective for modeling social interactions through causal reasoning. The results demonstrate that our map-free, lightweight trajectory prediction model rivals the performance of state-of-the-art methods and shows notable improvements over various baselines on publicly available real-world datasets.

In this paper, we report a new adaptive model reference control algorithm for a class of DC-AC inverters. The proposed design is based on a simplified discrete-time domain model with completely unknown system and input parameters. The control variable is designed as the turn-on interval of switch combinations within one switching cycle. The asymptotical stability of the proposed control is rigorously proved by using the Lyapunov stability method.

Cell balancing is crucial for charging supercapacitor cells to prevent cells from over-charging. Most existing cell-balancing charging methods typically adopt an output feedback control, i.e., the terminal voltages of cells are directly utilized in the controller design. One limitation of these methods is the voltage drop effect when the charging is terminated, which degrades the system capacity and results in cell imbalance. To address this challenge, in this article, we propose an observer-driven charging method for supercapacitors. The switched resistor circuit is applied and is further modeled using the switched systems theory, where the RC model of cells is considered. The communication interactions among cells is modeled using the graph theory. A switching Luenberger observer is designed to estimate the voltage of the equivalent capacitor of each cell, and a consensus-based switching control law is designed to charge and balance supercapacitors. The closed-loop system model is derived using the block diagram. A laboratory testbed has been built to verify the effectiveness of the proposed charging method. Experimental results show that the proposed method can effectively alleviate the voltage drop effect when compared with existing charging methods.

This paper considers the networked system optimization problem by cooperatively finding an approximately optimal solution to the overall network convex cost function, which is the sum of the individual cost functions of nodes (agents) in a networked system. A new distributed gradient descent algorithm is proposed based on the distributed estimation of the sum of gradients of individual cost functions using a consensus-type coordination algorithm. The proposed algorithm can address unknown directed communication topologies and only requires limited communications and information exchanges among nodes in the networked system. Under the assumption that the communication graph is strongly connected, the convergence of the proposed algorithm is rigorously analyzed. Simulation examples are presented to demonstrate the applicability and effectiveness of the proposed algorithm.

This paper presents an application of the formal method to model and verifies the communication and control for a fleet of collaborative autonomous underwater vehicles (AUV) named as Eco-Dolphin. The fleet includes YellowDolphin, BlueDolphin, and RedDolphin, which are designed to collect environmental data and provide surveillance services in littoral water. The system architecture of the fleet is specified by OpCat, a formal modeling tool based on Object-Processing Methodology. The collaborative missions of the fleet are coordinated by a Ground Station (GS) through radio and acoustic dual communication channels. The real-time behaviors and control logic for such a system is complex and error-prone due to the heterogeneous nature of the system and possible faulty communication. A timed state model is built to specify the essential communication and control behaviors. UPPAAL - an open source model-checking tool is used to verify the model against its desired & undesired properties specified in temporal logic queries. Because most of autonomous multi-agent systems depend on asynchronous and distributive controllers with large sets of state variables, the practice of model-checking for such controllers is severely hindered by the state explosion problem. Therefore, innovative heuristics need to be explored to reduce the state spaces of the state model. The major contributions of this paper are: (1) to present a practical software-hardware co-design of a fleet of mission critical AUVs, (2) to demonstrate how software design patterns and architectural styles can be utilized to reduce state spaces, (3) to provide a rare case study for verifying the safety and progress properties of a platform specific model.

In this paper, we present two near-optimal methods to determine the real-time collision-free path for a mobile vehicle moving in a dynamically changing environment. The proposed designs are based on the polynomial parameterization of feasible trajectories by explicitly taking into account boundary conditions, kinematic constraints, and collision-avoidance criteria. The problems of finding optimal solutions to the parameterized feasible trajectories are then formulated with respect to a near-minimal control-energy performance index and a near-shortest distance performance index, respectively. The obtained optimal solutions are analytical and suitable for practical applications which may require real-time trajectory planning and replanning. Computer simulations are provided to validate the effectiveness of the proposed near-optimal trajectory-planning methods.

In this paper, the cooperative control problem in the presence of communication time delays is addressed for a broad class of multi-input-multi-output dynamical systems in the canonical form with arbitrary relative degrees. A new sampled-data predictive cooperative control is proposed by only using the delayed output feedback information and a set of less restrictive conditions on the connectivity of time varying sensor/communication networks are established. No fixed leader or time invariant sensor/communication networks are assumed. In particular, through the elaborated switching between the decoupled control and the proposed predictive cooperative control, the convergence of the overall system is rigorously proved under a mild condition on the choice of sampling period. Simulation results on the cooperative consensus for a group of mobile robots are provided to illustrate the effectiveness of the proposed approach

In this paper, the problem of determining an optimal collision-free path is studied for a mobile robot moving in a dynamically changing environment. A polynomial parametrization is used to represent the family of mobile robot’s paths and to solve for feasible trajectories. Then a suitable performance index is set up and the trajectory generation problem is formulated as a constrained optimization problem subject to motion constraints and the collision avoidance criterion. The optimal solution is analytical and simple. It is good for realtime trajectory planning. Finally, the framework is applied to a car-like mobile robot. The validity of our approach is supported by computer simulations.

This paper addresses the problem of determining a collision-free path for a mobile robot moving in a dynamically changing environment. By explicitly considering kinematic, model of the robot, the family of feasible trajectories and their corresponding steering controls are derived in a closed form. Then, a new collision avoidance condition is developed for the dynamically changing environment, it consists of a time criterion and a geometrical criterion, and it has explicit physical meanings in both the transformed space and the original working space. By imposing the avoidance condition, one can determine the corresponding steering angle for collision avoidance in a closed form. Such a path meets all boundary conditions, is continuous, and can be updated in real time once a change in the environment is detected. Simulations show that the proposed method is effective.

The problem of determining a collision-free path for a mobile robot moving in a dynamically changing environment is addressed in this paper. By explicitly considering a kinematic model of the robot, the family of feasible trajectories and their corresponding steering controls are derived in a closed form and are expressed in terms of one adjustable parameter for the purpose of collision avoidance. Then, a new collision-avoidance condition is developed for the dynamically changing environment, which consists of a time criterion and a geometrical criterion, and it has explicit physical meanings in both the transformed space and the original working space. By imposing the avoidance condition, one can determine one (or a class of) collision-free path(s) in a closed form. Such a path meets all boundary conditions, is twice differentiable, and can be updated in real time once a change in the environment is detected. The solvability condition of the problem is explicitly found, and simulations show that the proposed method is effective.