Event-triggered Coordinate Control Of Multi-agent Systems

With the rapid development of science and technology,multi-agent systems have attracted considerable attention from various fields,which have been applied to mathematics,control,computer and so on.A variety of collective behaviors are generated by local interactions between multiple agents,such as consensus,flocking,formation,etc.The event-triggered control updates only when the event is triggered.It can reduce the frequency of updates and avoid the waste of resources to some extents.In this paper,we consider multi-agent models mainly based on event-triggered control.The main results are as follows.Firstly,we study the consensus tracking problem based on observer for a class of leader-follower multi-agent systems.Assume that the leader’s control input is unknown.In order to track such a leader,a distributed observer is constructed for each follower to estimate the control input of the leader,based on which a dis-tributed feedback controller is designed.Different from the existing results,both the observer and the control arc based on event.Both the centralized and de-centralized event-triggering time sequences are considered.If the adjacent gragh is connected,the tracking problem can be solved by the method of Lyapunov function and LaSalle invariance principle.It is further shown that the Zeno phe-nomena can be avoided.One can prove that two adjacent time inter-execution instants have a positive lower bound.It changes traditional continuous control to discrete communication.Then,a numerical example is given to illustrate the effectiveness of the controller and observer.Secondly,we consider the observer-based consensus problem of a general lin-ear multi-agent system.Assume that only the relative output feedback between the neighboring agents can be measured.To solve the consensus problem,an observer-based relative output feedback controller is designed and both the ob-server and controller are based on event.The consensus problem can be solved by the Lyapunov method and linear matrix inequalities.It is further shown that the Zeno phenomenon of triggering time sequences is avoided.A numerical example is presented to illustrate the effectiveness of the obtained theoretical results.Finally,we summarize the two types of multi-agent systems and look forward to the future researches.