This paper investigates the consensus problem of the multi-agent systems when subjected to external and malicious data injections. In particular, the primary objective is to develop a resilient network in the presence of attacks which aim to destabilize consensus dynamics by injecting bounded signals stealthily. To this end, a competitive strategy was utilized to establish a hidden layer of virtual system interconnected with the original system, so that the leader–follower consensus of the multi-agent systems can be guaranteed under proper topological conditions, and the technique derivation provides explicit rules on how to ensure stability of the overall system. Simulation results validate the theoretical contributions of this paper.

本文研究了在遭受外部和恶意数据注入时多代理系统的共识问题。特别是，主要目标是在存在攻击的情况下开发一种弹性网络，该网络旨在通过秘密注入有界信号来破坏共识动态。为此，采用了竞争策略来建立与原始系统互连的虚拟系统的隐藏层，以便可以在适当的拓扑条件下确保多主体系统的领导者与跟随者共识，并且技术推导可以提供明确的信息。有关如何确保整个系统稳定性的规则。仿真结果验证了本文的理论贡献。