Secure set-membership formation of multi-agent systems via an anti-attack control strategy

Abstract

Secure leader-following formation for multi-agent systems subject to parameter perturbations and saturation constraints is investigated in this article. To capture the variation of attack behaviors across different stages, a more general DoS attack model is developed using a Markov process with time-varying transition probabilities. Then, a memory-based anti-attack scheme is presented for designing the secure observer-based controller, supplying an effective method to weaken the impact of DoS attacks on formation control. Subsequently, the set-membership concept is introduced to derive some sufficient conditions such that the one-step ahead formation errors are always confined to the tracking ellipsoidal set. Combining the sufficient conditions with saturation constraints, an optimization problem is proposed to obtain the optimal ellipsoidal sets. For reaching the leader-following formation, an online optimization algorithm is provided to recursively solve the observer-based controller. To reduce the execution frequency of the optimization problem, a novel event-triggered mechanism with attack-dependent adaptive threshold is embedded in the optimization algorithm. Finally, the effectiveness and advantages of the proposed methods are certified via a simulation example.