Adaptive Prescribed Performance Event-Triggered Optimal Control for MIMO Systems via Command Filtering and Reinforcement Learning

This article investigates the optimal tracking control problem for uncertain multiple-input multiple-output (MIMO) nonlinear systems under constrained communication resources and prescribed performance. First, prescribed performance control is employed to confine the tracking error within given asymmetric bounds and to convert the constrained optimization problem into an equivalent unconstrained formulation. Second, a reinforcement learning algorithm with an identifier–critic–actor structure is, for the first time, integrated with command filtering in a backstepping framework to achieve optimal tracking control while mitigating the “explosion of complexity” problem. Furthermore, a novel relative-threshold-based event-triggered control mechanism is proposed, which extends the applicability of conventional event-triggered methods to MIMO systems and significantly reduces the communication load. On the theoretical side, Lyapunov analysis is conducted to prove the uniform ultimate boundedness of all closed-loop signals. The effectiveness of the proposed approach is validated through numerical simulation and experimental results on a robotic manipulator.