Event-Triggered Bipartite Time-Varying Formation Control for Multiagent Systems With Unknown Inputs
针对有向符号图上的多智能体系统，在领导者与跟随者均受外部扰动且状态不可测的条件下，设计了未知输入观测器和事件触发控制协议，实现了二分时变编队跟踪并避免了Zeno行为。
This article addresses the issues of the bipartite time-varying formation (BTVF) control for multiagent systems (MASs) on signed digraphs. All the designs are performed under the assumptions that the leader and followers suffer from external disturbances, the control input signal of the leader is unreachable to any follower, and the state variables of the followers are unmeasurable. To begin with, an unknown input observer (UIO) is designed for each follower using a traditional interval observer to obtain the state estimates. To realize the BTVF tracking, a distributed consensus error dynamic system is constructed. Furthermore, a distributed unknown input reconstruction method is developed to estimate the multiple disturbances in the consensus error system. Then, an event-triggered BTVF control protocol is proposed which allows two antagonistic time-varying formations to be formed, while excluding Zeno behavior. A simulation of a group of wheeled robots is used to demonstrate the performance of the proposed methods.