A novel distributed neural FTC strategy for interconnected nonlinear systems with unknown higher powers and its applications to CIPs

Abstract

This article focuses on the distributed fault-tolerant control (FTC) problem for a class of interconnected nonlinear systems with unknown higher powers. In order to reduce the excessive resource consumption and address the denial-of-service (DoS) attacks, an event-triggered communication mechanism (ECM) with multiple DoS detectors is developed for interconnected nonlinear systems for the first time, especially unknown higher powers are included, instead of strictly equal to one as in the relevant results. Besides, unlike the previous works where only actuator faults are considered, the FTC strategy proposed in this article can handle communication component failures as well. Then, an adaptive neural controller is constructed by utilizing back-stepping technique and the tracking error of each subsystem is proven to converge to a small neighborhood of the origin even in the present of DoS attacks based on Lyapunov stability theory. Finally, the developed strategy is applied to a class of coupled inverted pendulum (CIP) systems and the simulation result demonstrates the validity.