Research On The Consensus Of Multi-agent Systems

In this thesis,we study the cosensus of multi-agent systems in the cases of fractional order and integer order.In the case of fractional multi-agent systems,by designing a distributed event-triggered strategy,the general linear multi-agent systems realize the adaptive consensus.Furthermore,the adaptive self-triggering algorithm is proposed for the first time in the fractional system;For chained-form integer-order multi-agent systems,we design a new controller to make the systems achieve cosensus in a fixed-time.In the environment of fractional multi-agent systems,based on adaptive control,the general linear multi-agent systems achieve consensus.The communication structure between multi agents is undirected.Based on the distributed event-triggered strategy,we designed a new adaptive controller to solve the consensus of the fractional multi-agent systems.Furthermore,we also proved that Zeno behavior will not occur in the system,which means that the next launch time is strictly greater than the previous trigger time,which shows that the controller we designed is feasible.Then,an adaptive self-triggered algorithm is also proposed to relax the requirement for continuously checking the measurement errors,in which the next update time for each agent is determined according to its local history state information.Finally,numerical simulations are used to verify the effectiveness of our proposed control strategy.Compared with the existing controllers,the controller we designed allows the system to converge faster.In the integer-order multi-agent systems,the control strategy of fixed-time leader-following consensus is investigated for multiple chained-form systems with matched disturbances.A new decentralized observer is first designed to estimate the leader state in a fixed time for each follower.Then,a novel control protocol is developed to track the estimated leader state in prescribed finite-time.Better than the existing results,the system we studied has matching disturbances,and the upper bound of the stable time function does not depend on the initial conditions.In addition,as the signum function term of the proposed controller becomes smaller,the chattering phenomenon of the control output is reduced.Numerical simulation results verify that our proposed controller can solve the problem of system consensus in a fixed time.