Resilient Consensus for Discrete-Time Multiagent Systems With a Dynamic Leader and Time Delay: Theory and Experiment
研究了存在恶意智能体、动态领导者和通信延迟时，一阶和二阶离散时间多智能体系统的弹性一致性问题，提出了控制器并进行了仿真和实验验证。
The ever-present cyber-attacks have posed a significant challenge to consensus of multiagent systems (MASs), due to their capability of compromising agents. These threats underscore the critical need to design control strategies that can endow MASs with resilience. In this article, we address resilient consensus problems for first- and second-order discrete-time MASs, considering the presence of malicious agents, a dynamic leader, and communication delay. First, a resilient controller is designed for first-order MASs, and sufficient conditions are derived based on existing robust graph concepts to achieve consensus with ultimately bounded error. Next, since the derived error bound grows factorially with the system scale, a novel graph structure and a modified controller are proposed to limit the growth to a linear rate. Building upon the obtained results, an estimator-based control framework is introduced to solve resilient consensus problems for second-order MASs. Finally, comparative simulations and practical experiments based on unmanned ground vehicles and unmanned-aerial-vehicles are conducted to validate the effectiveness and practicability of the proposed control methods.