Resilient Consensus of Second-Order Multi-agent Systems Subject to Malicious Attacks

This paper addresses the resilient consensus problem of second-order multi-agent systems attacked by malicious agents. To solve the problem, an event-triggered mean subsequence reduced (E-MSR) algorithm is proposed. It means some extreme values of its neighbor nodes, which are most likely to be malicious nodes, will be ignored in the protocol generation of an agent. Under the event-triggered strategy, only when the triggering condition is satisfied, the agent sends its current state to its neighbor nodes to update their control algorithms, which effectively reduces the communication between agents. A numerical example is provided to demonstrate the effectiveness of the proposed approach.