Consensus Control For Nonlinear Multi-agent Systems Based On State Reconstruction

The consensus control of multi-agent systems has been increasingly favored by many scholars owing to its broad application prospects in many fields,such as the aerospace technology,intelligent transportation and sensor networks.In practice,agents usually have nonlinear dynamic characteristics,their modules including sensors,controllers,actuators and the information interaction between agents are connected or completed via networks.However,the limitation of bandwidth and channel capacity in real networks may lead to network congestion,and then result in delays in information transmission.Moreover,the complicated network environment may cause the faults of communication links between agents,and then induce the change of topology structure.Besides,partial state information of agents may be immeasurable or not shared globally,which needs to reconstruct the state information to provide feedback control signals.Therefore,by considering factors such as the topology change,delays,unmeasured or unshared states,this thesis studies the leaderfollowing distributed consensus control problem of nonlinear multi-agent systems based on state reconstruction.The main research works are as follows.(1)The leader-following state feedback consensus control problem is investigated for nonlinear multi-agent systems based on variable frequency switching topology.Based on an optimized topology switching rule,the distributed switched observers are designed for followers to reconstruct the state information of leader.Then,by using dynamic gain control method,the observers-based distributed controllers are proposed,which improve the chattering problem of existing controllers.Subsequently,it is proved that the controller achieves the leader-following state consensus.The algorithm allows that partial switching topologies do not contain the directed spanning tree with leader as the root,and realizes the variable frequency switches between topologies.(2)The leader-following state feedback consensus control problem is investigated for nonlinear multi-agent systems subject to known input delays.Based on undirected fixed communication topology,the distributed finite-time observers are designed for followers to reconstruct the state information of leader.Then,the observers and input delays-based distributed controllers are proposed by truncated prediction control method.Subsequently,based on the Lyapunov stability theorem,sufficient conditions are given for realizing the leader-following state consensus.The algorithm permits that the input of the leader is an unknown time-varying function and input delays are different time-varying functions.(3)The output feedback-based leader-following fully distributed output consensus control problem is studied for nonlinear multi-agent systems.By using the relative output information of agents,the distributed reduced-order dynamic compensators are developed for followers to reconstruct the relevant state information.Then,the compensators-based fully distributed linear-like controllers are proposed based on the dynamic gain control method.Subsequently,it is proved that the controller can realize leader-following output consensus in global sense.The algorithm effectively reduces the transmission amount of information among agents and allows that all followers have strong nonlinear dynamics.(4)The leader-following output feedback consensus control problem is investigated for nonlinear multi-agent systems with unknown input delays.Based on the relative output information of agents,the distributed observers and classical observers are designed for followers to reconstruct the state information of leader and followers,respectively.Then,by means of the truncated prediction control method,the double observers-based and input delay-independent distributed controllers are proposed.Subsequently,sufficient conditions are given for achieving the leader-following state consensus.The algorithm permits that input delays are unknown different time-varying functions.(5)The leader-following output feedback consensus control problem is studied for nonlinear multi-agent systems with unknown state delays.Based on the relative outputs of agents,the distributed dynamic compensators are developed for followers to reconstruct the relevant state information.Then,the compensators-based and state delays-independent distributed controllers are designed based on dynamic gain control method.Subsequently,it is proved that the controller renders the leader-following state consensus.The algorithm permits that agents have high-order uncertain nonlinear dynamics.Finally,it is extended to multi-agent systems with state delays and strong nonlinear dynamics.