In this brief, a novel distributed leader-following consensus controller for linear multi-agent systems (MASs) with one leader and $N$ followers is designed for optimal leader-following consensus control. A user-friendly global objective function is well designed in a quadratic form. The proper choice of the weighting matrices and necessary and sufficient conditions for optimizing the objective function are given. The proposed controller can guarantee the optimization of the objective function and the achievement of the leader-following consensus of the MASs. To deal with the coupling in the controllers, we propose a dynamic average consensus-based algorithm to estimate the controller distributedly, and the estimate errors converge to zero in finite time. A simulation example is provided to demonstrate the results.

In this article, the optimal leader-following consensus control problem of multi-agent systems is solved using a novel neural network-based (NN-based) integrated heuristic dynamic programming (IHDP) algorithm under a distributed learning framework. The optimal controller is designed to find a Nash equilibrium of the multi-agent differential graphical game. According to the principle of optimality, this problem can be transformed into solving a set of coupled Hamilton-Jacobi-Bellman (HJB) equations, which are difficult to solve directly because of the nonlinearity and strong coupling of the equations. The proposed algorithm combines policy iteration (PI) and value iteration (VI), and the Nash equilibrium of performance index among agents is achieved distributedly, i.e., all agents update their control inputs only use local information. The convergence of the algorithm is rigorously proved. Neural network (NN) approximation implementation is given for the unknown structures of local value functions of agents. The effectiveness of the algorithm is verified by numerical simulation.

This paper focuses on simultaneous estimation of states and faults for a linear time-invariant (LTI) system observed by sensor networks. Each sensor node is equipped with an observer, which uses only local measurements and local interaction with neighbors for monitoring. The observability of said observer is analyzed where non-local observability of a sensor node is required in terms of the system state and faults. The distributed observers present features of $H_{\infty }$ performance to constrain the influence of disturbances on the estimation errors, for which the global design condition is transformed into a linear matrix inequality (LMI). The LMI is proven to be solvable given collective observability of the system and a suitable $H_{\infty }$ performance index. Moreover, in the case that no disturbances exist, fully distributed observers with adaptive gains are designed to asymptotically estimate the states and faults without using any global information from the network. Finally, the effectiveness of the proposed methods is verified through case studies on a spacecraft’s attitude control system.

This paper investigates the distributed optimal cooperative tracking control problem for multi-input linear time-invariant (LTI) systems. In this context, the system inputs are generated by a group of agents that communicate with each other over a network, i.e., each control input channel is considered as an agent, which can communicate over a network to transmit information and compute control input. Unlike centralized optimal tracking control, where inputs are designed using global information, each agent in the distributed framework has access only to its own input matrix and communicates solely with its neighbors within the network. This limitation introduces significant challenges in designing the optimal controller. To address this issue, an information fusion method is first proposed, enabling each agent to derive its optimal controller in a distributed manner. For scenarios where the system model is unknown, a fusion-based learning algorithm is further developed. The convergence and optimality of this algorithm are rigorously proved. A simulation example is provided to illustrate the effectiveness of the proposed approach.

In this Letter, a physics-based data-driven strategy is developed to predict vortex-induced drag on a circular cylinder under non-uniform inflow conditions—a prevalent issue for engineering applications at moderate Reynolds numbers. Traditional pressure-signal-based models exhibit limitations due to complex vortex dynamics coupled with non-uniform inflow. To address this issue, a modified fully connected neural network architecture is established that integrates upstream velocity measurements (serving as an inflow calibration) with pressure-signal-based inputs to enhance predictive capability (R2∼0→0.75). Direct numerical simulations at Reynolds number Re=4000 are implemented for model training and validation. Iterative optimizations are conducted to derive optimized input configurations of pressure sensor placements and velocity components at upstream locations. The optimized model achieves an R2 score of 0.75 in forecasting high-amplitude drag coefficient fluctuations (Cd=0.2–1.2) within a future time window of one time unit. An exponential scaling between model performance and optimized pressure signal inputs is observed, and the predictive capability of sparsely distributed but optimized sensors is interpreted by the scaling. The optimized sensor placements correspond to the physical mechanism wherein flow separation dynamics play a governing role in vortex-induced drag generation. This work advances machine learning applications in fluid–structure interaction systems, offering a scalable strategy for forecasting statistics in turbulent flows under real-world engineering conditions.