Research On The Collaborative Control Method Of A Kind Of Distributed Multi-agent System

In recent years,due to the wide application of multi-agent systems control in such fields as distributed sensing network,unmanned aerial vehicle cooperative control and formation control,the cooperative control of multi-agent network systems has received extensive attention from many experts and scholars domestic and abroad,and it has become one of the important research topics in the control field.The actual systems in nature are almost non-linear.The research on the consistency control of nonlinear multi-agent systems is very valuable from both theoretical and practical perspectives.This paper systematically analyzes and summarizes the research background and current situation of the coordination control of nonlinear multi-agent systems from theoretical aspects,and provides relevant preliminaries.Further research is carried out on the coordination control of nonlinear multi-agent systems,especially the coordinated tracking and average tracking problems.The main contents are as follows.Based on the previous work,the consensus problem of nonlinear multi-agent systems is studied.For a class of nonlinear multi-agent systems,a consensus control method based on neural network and adaptive technology is proposed.Under the condition that the network topology is switched topology,the unknown nonlinear function is estimated by neural network.The state observer is given to estimate the state of the system.An observer-based adaptive neural network coordination control law is proposed.The parameter design of the control law does not depend on any global information,only uses the relative state information of itself and neighbor agents,which can ensure that the state of the closed-loop system is ultimate uniformly bounded.Then,the average tracking problem of nonlinear multi-agent systems is studied under switched topology,that is,the state of the agent can track the average value of the multiple time-varying reference signal.Assuming that the time-varying reference signal is generated by an unknown nonlinear dynamic system,the average tracking filter is first defined,and an adaptive controller based on neural network is designed.Lyapunov analysis method is used to prove that the state of the closed-loop system is ultimate uniformly bounded,and the average tracking problem can be solved by choosing adequate parameters.A numerical simulation system is established to verify the control algorithm given in this paper.The simulation results verify the effectiveness and feasibility of the algorithms.