Reinforcement Learning-Based Adaptive Event-Triggered Fuzzy Control for Cyclic Switched Stochastic Nonlinear Systems With Actuator Faults

In this article, the adaptive fuzzy optimized tracking control problem of uncertain cyclic switched stochastic nonlinear systems with actuator faults and prescribed performance is studied under average cycle dwell time. For achieving a better optimal control strategy, the information of the switching signal is taken into account by the reinforcement learning algorithm of the identifier-critic-actor architecture. Meanwhile, a novel mode-dependent event-triggered optimized mechanism for subsystem is proposed to resolve the impact of asynchronous switching on system performance, which not only does not have strict assumptions but also reduces the communication burden. Moreover, to eliminate the impact caused by actuator failures, fault-tolerant compensation scheme depending on the switching signal is designed. By presenting Lemma 3 and using the coordinate transformation technique, it is proven that the choice of the performance index function still conforms to the control protocol under the prescribed performance control framework. In the optimized backstepping control design, the proposed controller is able to ensure that all signals in the closed-loop system are bounded in probability and tracking error meets performance. Finally, both numerical and practical simulations are given to illustrate the effectiveness of the proposed optimization scheme.