Event-Triggered Anti-Attack Security Control of Cyber-Physical Systems With Bio-Inspired Metaheuristic Algorithm

Abstract

This work addresses the event-triggered anti-attack security control problem for cyber-physical systems (CPSs) under denial-of-service (DoS) attacks using a bio-inspired metaheuristic algorithm. First, a fuzzy switching model with the consideration of actuator failure and saturation is developed to approximate the uncertain nonlinear CPSs. Second, a valid DoS attack framework is established to model the nonperiodic DoS attacks with different types of start and end moments. Under this framework, a correlation model incorporating a reliable dynamic event-triggered (DET) mechanism and the valid DoS attack is constructed. To optimize the balance between control performance and network communication burden, a parameter optimization strategy for the DET mechanism is proposed utilizing the hippopotamus optimization algorithm. Furthermore, the Wirtinger inequality is utilized for reliable DET controller synthesis to reduce the conservatism. Finally, experimental tests are conducted to examine the effectiveness and merits of the proposed anti-attack security control method.