Fault Tolerant Consensus Control For Multi-Agent Systems With Actuator Faults

Curre.ntly,with the rapid development of network technology,control theory and artificial intelligence technologyhow to establish coordinated control among multiple agents has become a hot research topic in the field of control field.This proposed method is used to complete a complex task by multiple agents with simple structure to form a large-scale networked system through information interaction.If adopting the traditional centralized control method,the control center of multi-agent systems needs to be established,which is difficult to accomplish the coordination and stability control of the system.As a result,distributed control method arises at the historic moment,which has the characteristics of good robustness,high flexibility,strong expansibility and so on.So this method has been widely applied to the coordinated control of consensus of multi-agent systems,such as satellite attitude conrol,mobile robot formation,multi-unmanned aerial vehicle search and rescue,power dispatching and other fields.In addition,this method uses only the local information of each subsystem,which can reduce the communication resources,wireless range and broadband.However,due to the increasing scale,the complexity of communication transmission network,long-term and high-load operation of the multi-agent systems,the subsystem failure or communication fault may occur,inevitably.If the faults can not be found and handled timely,it will lead to the failure of the control function and even bring great economic losses.Therefore,in order to ensure the security and reliahility of multi-agent systems,it is an important research direction to estahlish the corresponding fault-tolerant control.On the basis of the previous work of the consensus of multi-agent systems,the fault tolerant control of the multi-agent systems is studied under the actuator fault.The main achievements are outlined as the following aspects:1)The fault tolerant control method based on unknown input observer is proposed to solve the actuator fault,problem of multi-agent system with external disturbance.According to the pole-placement constraints,the distributed compensator of the leader has better compensation effect than the constant gain or given matrix gain of the existing compensator.For the constructed augmented multi-agent system,the observahility is proved and the unknown input observer is designed.Then,the distributed fault tolerant controller is designed by using the states and faults information.Finally,it is proved that the proposed method can achieve the consensus tracking of multi-agent systems while compensating the effect of faults and external disturbances,and the effectiveness of the proposed method is verified by simulation.2)The actuator fault problem for a class of nonlinear multi-agent systems is invwstigated by proposing a fixed-time observer.This method is described by the relative states of neighbors and auxiliary variable.Then,to achieve the fault detection task of multi-agent systems,the residual vector is constructed by using the estimation of states and auxiliary variable.After the fault is detected,based on the fault estimation algorithm under the fixed-time observer,the fault tolerant controller is proposed to achieve the consensus tracking task within the settling time.3)The time-varying coupling weight-valued adaptive fault-tolerant controller is proposed to solve the actuator fault problem of multi-agent systems with unknown input.This method not only overcomes the asymmetric Laplace matrix caused by the directed communication topology and the limitation of the rank of control matrix,but also eliminates the chattering effect caused by the discontinuous control signal.In addition.the modification term is added in the adaptive fault estimation,which can filter out the high-frequency oscillations.Then,the stability of multiagent.system is proved by the Lyapunov stability theory,and all the signals in the resulting closed-loop systems are uniformly ultimately bounded.4)The adaptive fixed-time fault tolerant controller and parameters updating law are proposed for the second-order nonlinear multi-agent systems with the loss of efdectiveness and bias faults.This method is based on the relative states of neighbors.the adding a power integrator method and adaptive control technology.Since the occurrence time and magnitude of the actuator faults are completely unknown and unpredictable,the proposed method overcomes the initial condition problem of the convergence time in finite time fault-tolerant control results,and solves the asymmetric Laplace matrix caused by the directed topology,which widens the application scope of this method.In addition,to facilitate the stability analysis of the fixed-time fault-tolerant controller,a new mathematical inequality is proposed.The consensus tracking target of the multi-agent systems is realized,where the convergence time of tracking error is not affected by the unknown faults and initial condition.Finally,it is proved that the adaptive laws,tracking errors are uniformly ultimately bounded even in the presence of the faults.5)The problem of time-varying formation fault-tolerant control for multi-agent systems with multiple leaders is investigated by proposing a class of adaptive faulttolerant formation time-varying controller and parameter updating laws,which contains the relative states of neighbors,adaptive control technology and distributed control strategy.The proposed control protocol can compensate the influences of the fault on agents and formation.Then,the stability of multi-agent systems is proved by using the Lyapunov stability theory,and all the signals of the closed-loop system are uniformly ultimately bounded.