Fast Consensus Analysis And Output Regulation Of Multi-agent Systems

As been an important branch for science subject of artificial intelligence(AI),the application of Multi-agent systems(MASs)has become more and more extensive with the rapid development of communication technique recently.MASs have the merits of easy extension,information sharing and high reliability.The consensus analysis problem is the fundamental issue of MASs theory,which aims at designing a reasonable and effective distributed protocol for each agent such that a group of agents can achieve the same state and work cooperatively.The output regulation problem is the basis for the cooperative control of MASs.The main content is summarized as follows:Firstly,the fast consensus control problem of second-order MASs with undirected topological networks is analyzed based on matrix theory,stability theory and graph theory.New algorithm is obtained on the basis of conventional consensus protocol.Using the frequency domain analysis method of linear systems,the sufficient conditions for the convergence and the fast convergence speed of the MASs are presented based on Nyquist stability criterion.And the new algorithm overcomes the shortcomings of the conventional Lyapunov stability methods which are difficult to construct Lyapunov function.The results are verified by the two Matlab simulation experiments with different topological networks.Secondly,the robust output regulation problem of linear heterogeneous multi-agent systems with undirected and switching network topology is addressed.Due to uncertain parameters in every agent or subsystem,an internal model method is adopted in the sense that the controller allows the system uncertain parameters to vary in a neighborhood of the nominal system.By designing two kinds of appropriate distributed control scheme,output regulation problem is well solved.A numerical example is presented to illustrate the effectiveness of the proposed method.Finally,wheeled robots system experiment platform is designed based on the MASs with leader and follower.The system is mainly composed of a leader monitoring robot,a follower robot and a host computer.The user controls the movement of the leader robot through the host computer.The follower robot acquires the position of the leader robot with the carried camera,and implements following based on the PID control algorithm.