Consensus Control For High-Order Nonlinear Multi-agent Systems

With respect to the multi-agent systems consensus control, most of the existing works studied only the ?rst-order and second-order dynamics. However, lower-order linear model used in the existing theoretical results is too simple to describe the dynamic characteristics accurately, which ignore the higher-order and nonlinear properties of the actual physical system. These are resulting in the gap between multi-agent theory and practical application. Hence, this thesis considers consensus control of higher-order nonlinear multi-agent systems as the main topic. Several kinds of high-order nonlinear multi-agent systems consensus control problems are investigated. This thesis focus on how to design distributed controller based on the traditional nonlinear theory, the extension of nonlinear theory application and how to increase the robustness of nonlinear multi-agent systems.Under the Backstepping framework, the controller design method for nonlinear high-order multi-agent systems, including distributed adaption control, distributed robust control,distributed adaption robust control and the combination of the above control strategies.First, a class of higher order nonlinear multi-agent systems with parameter uncertainties is considered. According to the model features, coupled with the constraints of the multi-agent network topology and the traditional Backstepping technique, a distributed error system for the each order of multi-agent systems is established. Then a high-order multi-agent systems control problem is broken into multiple low-order multi-agent systems coordination. By Lyapunov stability analysis methods and Barbalat Lemma, the asymptotic stability of the multi-agent systems is proved. Finally, a simulation example including11 agents’ system is proposed to verify the result.Secondly, the consensus control problem of a class of high-order multi-agent systems with unknown dynamics and unknown bounded disturbance is considered. According to the Lipschitz condition of the nonlinear function of the unknown dynamics, fuzzy logic system method is used to estimate the unknown dynamics.And the tuning problem of the unknown parameter is dealt with by a distributed adaption controller. Then, the estimation error of the unknown dynamics is considered as a class of disturbance which will be dealt with by a distributed robust controller with external disturbance together. By graph theory, Lyapunov stability analysis methods and Barbalat Lemma, the asymptotic stability of the multi-agent systems is proved. Finally, a simulation example including 5agents’ system is proposed to verify the result.Thirdly, the consensus control of a class of nonlinear multi-agent systems with unknown disturbance(the bound function is unknown) and unknown dynamics(the bound function of the estimation error of unknown dynamics is unknown) is investigated. Based on the distributed robust controller, an adaption law is embedded to adjust the robust gain value. The proposed distributed adaption robust controller compensate the estimation error of the unknown dynamics. Finally, a Matlab simulation is proposed to verify the e?ectiveness.Finally, this paper studies the application of consensus control. Speci?cally, nonlinear multi-agent systems consensus control problem is considered, which has multi task constraints in the environment of obstacles. The priorities of the tasks is computed based on the null space behavior control methods. The unknown dynamics of the systems is dealt with by fuzzy logic system method. And the fast terminal sliding mode control makes the fast convergence of the error. Based on Lyapunov stability analysis methods, the asymptotic stability of the multi-agent systems and the tasks stability are proved. Finally,simulation examples with Matlab and Pionner 3-AT are proposed to verify the theory..