Research On Fault-Tolerant Control Of Multi-Robot System With Time-Varying Topology

Multi-robot technology is a hot spot in the field of robotics.At present,the research on faults mainly focuses on the faults of topology or the internal components of robots,without considering the influence of the two coupling on the system.In this paper,the distributed passive and active fault-tolerant control methods of multi-robot systems are studied respectively,under the jointly connected topology and continuous time-varying topology,aiming at the coupling fault of communication topology and actuator.The specific research is as follows:(1)For the multi-robot tracking system on the actuator faults with jointly connected topology,the topology switches over time and the switching rules are based on Markov process.In order to solve the coupling faults,an error system is established based on algebraic graph theory,Riccati inequality,and Lyapunov stability theory.To find the optimal solution by using LMI toolbox of MATLAB,a passive fault-tolerant controller is designed,and a consistent protocol is proposed under the topology structure.Then,the stability of the multi-robot system is proved by Lyapunov theorem of stability.Through numerical examples,the effectiveness of the consensus tracking protocol is verified by UAV and unmanned vehicle simulation.(2)Under the jointly connected topology,an active fault-tolerant controller is designed to solve the faults.Because of the unknown fault information,the system fault is difficult to be accurately detected.Although the controller can choose a larger gain to deal with the fault,it will cause serious instability of the system.For actuator failure,the partial failure observer is designed when the topology may not be connected.Based on the observer estimation,a fault-tolerant adaptive law for coupling faults in the jointly connected topology is established by combining the adaptive technology with Lyapunov stability theory.Then a distributed active fault-tolerant tracking control algorithm is proposed,that is,the adaptive fault-tolerant tracking control algorithm to achieve the tracking consistency of multi-robot system.(3)Polytopic model is used to construct continuous time-varying topology for tracking problems with faults under continuous time-varying topology.Based on the established state error variables,a kind of sliding mode variable is designed,and a fault-tolerant tracking control strategy with robustness is proposed by combining Lyapunov function.The simulation results show that successful tracking is achieved with proposed passive fault-tolerant control under time-varying topology and actuator faults.(4)Active fault tolerance can be realized under continuous time-varying topology.Based on the quantitative estimation method of fault upper limit of distributed observer,an adaptive fault-tolerant tracking control strategy with active adjustment of control parameters is established to compensate the faults such as unknown actuator,so as to obtain the consistent tracking results of active fault-tolerant control under time-varying topology faults.Compared with the passive control algorithm,active mode has better fault-tolerant tracking performance and faster tracking time,showing good robustness.This thesis will improve the fault tolerance ability of multi-robot system and enrich the theory system of cooperative control and multi-robots.