Research On The Consensus Of Multi-agent Systems Based On Event-triggered Control Strategies

With the wide application of multi-agent in society,the consensus of multi-agent system has become a hot issue for many scholars.However,most of the control systems are updated continuously,and the operation of each agent is limited.Therefore,it is worth to study how to reduce the frequency of agent's update and operation.To solve this problem,this thesis wields event-triggered control to study the consensus of multi-agent.(1)A method to solve the fixed time consensus problem is proposed for leaderfollowing multi-agent system with nonlinear and perturbation in the case of directed topology event-triggered.Combined with the differential equations of the system,a nonlinear controlling protocol based on event-triggered is proposed.By using Lyapunov stability theory and algebraic graph theory,it is proved that the multi-agent system can achieve the leader-follower consensus in a fixed time.The condition of event-triggered is given,and Zeno behavior is eliminated by using the minimum triggering interval.(2)In multi-agent system,there is competition among agents of different groups,while there is cooperation among agents of the same group.In response to this phenomenon,a new event-triggered control protocol is proposed to reduce the communication burden of multi-agent system with cooperative competition group.By using matrix theory,barbarat lemma and Lyapunov theory,the sufficient conditions for group consensus of multi-agent systems triggered by events are given.In particular,Zeno behavior is common in the event-triggered problem,but this behavior is avoided in the end,and the minimum trigger interval is given.(3)In the directed topology,the problem of time delay consensus between follower and leader is solved by event triggering in the second-order leader-follower nonlinear multi-agent.An event-triggered mechanism and a new controlling strategy for timedelay consistency are proposed.Then,it is proved that the second-order multi-agent system can achieve time-delay consistency with leaders by using Lyapunov stability theory,algebra knowledge and graph theory knowledge.In addition,the matrix theory and inequality method are used to effectively eliminate the Zeno behavior in the above proposed controlling strategy.In this thesis,the consensus of several event-triggered control protocols and the exclusion of Zeno behavior are verified by numerical simulation,which increases the feasibility and correctness of the theoretical part.