Pull-based Event-Triggered Cooperative Control For Multi-Agent Systems Via Sampled Data Transmission

The distributed cooperative control of multi-agent system has many applications in lots of fields,such as formation control of unmanned aerial vehicle,cooperative control of sensor network,etc.In recent years,with the in-depth study of researchers,the con-trol conditions of multi-agent systems have been considered more comprehensive.For example,while considering the control goal of the system,we also hope to reduce com-munication times and save energy consumption.Thus,some researchers give the event-triggered control strategies and the time-triggered control(sampling control)strategies.In these methods,trigger instants appear in the controller,making the controller change at the trigger instants rather than update continuously,which can reduce the number of controller updates.Based on the above control methods,a new control protocol combining sampled methods and pull-based event-triggered control is proposed.Meanwhile,a series of dis-tributed triggering functions are designed for agents to determine the next trigger point,which means continuous updating of controllers and communication among agents can be avoided.Under this framework,this paper does the following work.1.In this paper,the containment control problem of second-order multi-agent sys-tem is firstly studied.A control protocol combining aperiodic sampled and pull-based event-triggered control mechanism is proposed.Moreover,aperiodic sampling method is adopted in the trigger function of each agent,so that the trigger functions only need to calculate at the sampling point to reduce the redundant computation.Later,it is proved that the containment control error of the multi-agent system converges to zero under the drive of the controller by solving the system equation.In other words,leaders in the sys-tem can reach the ideal positions specified in advance,and the followers enter the convex hull formed by the leaders.In addition,the Zeno phenomenon was naturally ruled out by the sampling control methods.2.Secondly,the consensus tracking control problem of nonlinear multi-agent sys-tem with time delay based on Markov switching topology is studied.This paper gives a class period sampling protocols combined with pull-based event-triggered control proto-cols.Also,the trigger functions we design,only consider using agent own sampling time and the information of neighbor at trigger instants.Thus,there is no information trans-mission in two adjacent triggering instants,which means reducing the burden of commu-nication.In addition,due to the adoption Markov switching mechanism,the parameters of the controller will change with time.So the corresponding design theorem of the pa-rameter matrix of the controller is presented.Finally,by virtue of algebraic graph theory and Lyapunov stability theory,this paper proves that the proposed sufficient conditions can make the tracking error of multi-agent system go to zero,that is,multi-agent system can achieve consensus tracking control.Finally,for each problem,the corresponding numerical simulations are given,which verifies that the corresponding control protocol can make the multi-agent system achieve the control goal.Furthermore,for the first one,this paper gives the comparison of the triggering times of the pull-based event-triggered mechanism and the traditional mecha-nism,which shows the superiority of the pull-based one in reducing the updating times of the controller.