Research On Fault-tolerant Control For Multi-agent Systems With Actuator Faults

In recent years,multi-agent systems have received great attention from many scholars in engineering and scientific circles due to their broad application prospects in the fields of multi-unmanned system collaborative control,distributed computing,multisensor networks,intelligent transportation,and networked control.However,with the development of computer technology,the multi-agent system is gradually unable to guarantee its security and reliability while expanding its scale.The increase in the frequency of system failures has a great impact on system performance or stability,and can even cause serious damage to the system,resulting in inestimable economic and social losses.Therefore,the combination of fault-tolerant control and distributed control makes the multi-agent system tolerate faults,which can further improve the reliability and control quality of the system,and meet the increasing requirements for system security in the fields of industry and military.Further improving the reliability and security of complex and large-scale multi-agent systems provides an important research idea,and promotes the widespread application of related theories in more engineering projects.It has important theoretical research significance and social application value.In the future,it will also become a research hotspot in this field.This paper studies the fault-tolerant control problem of multi-agent systems,and the contributions are as follows:(1)A distributed robust fault-tolerant consensus control scheme is proposed to solve the problem of multiple faults and non-matching nonlinear multi-agents in undirected switching communication topology.The unknown nonlinear function is identified by neural network and the norm of weight coefficient in neural network is estimated by adaptive technique.Different from the existing methods for estimating weight coefficients,the number of adaptive parameters can be reduced by estimating its norm.By combining the topology-based average dwell time time technique with Lyapunov stability theory,sufficient conditions are obtained to ensure the stability of the closed-loop system.(2)Lipschitz nonlinear multi-agent systems with actuator faults and intermittent communication under undirected communication topology are studied.Lipschitz conditions are used to estimate unknown nonlinear functions,and an adaptive mechanism is used to design a novel distributed robust fault-tolerant controller to eliminate the adverse effects of nonlinear functions and compound faults.In addition,the closed-loop system is proved to be stable by means of topology-based average dwell time time technique and Lyapunov stability theory.Compared with existing controllers for nonlinear multiagent systems,the proposed scheme can achieve fault-tolerant consensus of nonlinear multi-agent systems even under the influence of intermittent communication,actuator faults and switching communication topology.(3)A fault-tolerant output regulation control scheme is proposed for heterogeneous multi-agent systems with input saturation,interruption and failure in directed communication topology.Assuming that all subsystem matrices are different,a fully distributed observer is designed on each follower without using any global information in order to estimate the state of the external system or virtual leader.A novel observer-based coordinated fault-tolerant controller is designed to compensate for actuator faults and input saturation.Combining Lyapunov stability theory and low gain feedback theory,it is proved that the proposed method can solve the problem of semi-global coordinated fault-tolerant output regulation.(4)A two-layer fault-tolerant control scheme consisting of network layer and physical layer is proposed for a multi-agent system with input saturation under unknown control inputs.At the network layer,under the jointly connected topology,a fully distributed observer is proposed for the deal with unknown control input,and a fully distributed virtual cooperative control scheme is designed without the information of Laplacian matrix eigenvalues.At the physical layer,a collaborative saturation fault-tolerant tracking control method based on observer is proposed for the actual system with actuator faults and input saturation.(5)For heterogeneous multi-agent systems with input saturation and actuator faults,a bipartite fault-tolerant containment control framework based on directed signed network graph is proposed.A containment distributed virtual cooperative controller is designed based on directed signed network for observer system in network layer.A saturation fault-tolerant containment controller is designed for heterogeneous agents in the physical layer by using output regulation method.The integrated action of controllers at the network layer and physical layer finally achieves output containment,that is,the leader’s output forms the desired convex hull,while the follower’s output converges to the convex hull composed of the leader’s output.