Adaptive neural resilient control for networked switched systems with event-triggered communication and multiple cyber attacks

Abstract

This paper is concerned with adaptive neural resilient control problem for networked switched systems under event-triggered communication. The multiple cyber attacks including denial-of-service (DoS) attacks and deception attacks are introduced. The closed-loop system with double switching signals is constructed to describe the effect of DoS attacks. A positive lower bound is fixed in event-triggering scheme (ETS) to avoid Zeno behavior resulted by deception attacks. In order to mitigate the influence of deception attacks on system performance, neural networks are used to approximate injected false information and the corresponding adaptive control strategy is developed. Based on neural networks and the resulting closed-loop system, multiple Lyapunov functions method and average dwell time technique are adopted, and sufficient conditions are provided to ensure that all states are semi-globally uniformly ultimately bounded (SGUUB). Subsequently, some criterions are presented to deign the parameters of controller gain and ETS. Finally, an example is given to validate the effectiveness of the proposed method.