Data-Based Event-Triggered Cooperative Optimal Output Regulation of Nonlinear Multiagent Systems
研究了非线性多智能体系统在事件触发机制下的协同最优输出调节问题，提出基于数据的前馈-反馈控制方法，利用神经网络学习求解调节器和HJB方程，并通过离策略积分强化学习实现最优控制。
This article investigates the data-based cooperative optimal output regulation problem (COORP) for nonlinear strict-feedback multiagent systems (MASs) under an event-triggered mechanism (ETM). By constructing an adaptive distributed observer, each follower can estimate the leader’s dynamic and state. In the control design, a feedforward-feedback control input is proposed based on system data. By utilizing the neural networks (NNs) to learn the solutions of the nonlinear regulator equation and the Hamilton–Jacobi–Bellman (HJB) equation, the feedforward control problem and the optimal feedback control problem can be addressed. Then, an off-policy integral reinforcement learning (IRL)-based optimal cooperative control method is proposed with actor-critic NNs (A-C NNs), and the influence caused by unknown nonlinear dynamics can be handled. <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">F</i> Through the stability analysis, it is proved that all signals in closed-loop system are uniformly ultimately bounded (UUB), and the system can achieve Nash equilibrium. To demonstrate the effectiveness of the developed optimal control method, a simulation example is provided.