Event-triggered resilient consensus control of multiple unmanned systems against periodic DoS attacks based on state predictor

Abstract

This paper develops an event-triggered resilient consensus control method for the nonlinear multiple unmanned systems with a data-based autoregressive integrated moving average (ARIMA) agent state prediction mechanism against periodic denial-of-service (DoS) attacks. The state predictor is used to predict the state of neighbor agents during periodic DoS attacks and maintain consistent control of multiple unmanned systems under DoS attacks. Considering the existing prediction error between the actual state and the predicted state, the estimated error is regarded as the uncertainty system disturbance, which is dealt with by the designed disturbance observer. The estimated result is used in the design of the consistent controller to compensate for the system uncertainty error term. Furthermore, this paper investigates dynamic event-triggered consensus controllers to improve resilience and consensus under periodic DoS attacks and reduce the frequency of actuator output changes. It is proved that the Zeno behavior can be excluded. Finally, the resilience and consensus capability of the proposed controller and the superiority of introducing a state predictor are demonstrated through numerical simulations.