Research On Multi-Agent System Fault Diagnosis Method Based On Belief Rule Base

Large-scale multi-agent systems are widely used in various fields,and the gradual increase in system complexity makes the probability of failure of each agent in the system also increasing.Cooperative behavior among agents improves the efficiency of the system in completing tasks,but it also causes the failure of a single agent to spread to the entire system,increasing the risk of overall system failure.In actual engineering environment,the problems of incomplete system status information and lack of control information are unavoidable,which limits the application of existing multi-agent fault diagnosis methods based on unknown observers in actual systems.Therefore,this paper carries out the research of multi-agent system fault diagnosis method based on belief rule base(BRB).The main research contents are as follows.Firstly,this paper proposes a fault diagnosis architecture for multi-agent systems based on BRB.Aiming at the second-order multi-agent system under communication interference,a consensus control strategy based on sliding mode control algorithm is proposed,and based on this,a centralized fault diagnosis model based on BRB is constructed.When the system is fault-free,the proposed consistency controller can achieve the consistency of the multi-agent system.When a fault occurs in the system,the designed fault diagnosis model can effectively diagnose the faulty agents.In order to verify the effectiveness of the proposed method,simulation verification is carried out in a multi-vehicle system.Secondly,as the scale of multi-agent system expands and the number of agents in the system increases,the centralized fault diagnosis model is difficult to deal with the high complexity and high-dimensional computation problems of large-scale multi-agent system.Therefore,this paper proposes a distributed fault diagnosis method for multiagent systems based on BRB.Using the relative information between neighboring agents,a distributed fault diagnosis model is established,which effectively solves the problems that the existing centralized fault diagnosis model is difficult to obtain the global information of the system and the model computation is overburdened.In order to verify the effectiveness of the proposed method,simulations are performed in a multi-vehicle system.Finally,on the basis of the proposed distributed fault diagnosis method,for the multiagent system with multiple faulty agents,considering the problem that some agents are affected by the failure of its neighboring agents or other factors,resulting in the low accuracy of fault diagnosis results,a topology switching strategy based on the output diagnosis accuracy is proposed.By updating the topology structure of the multi-agent system,the fault diagnosis results are more accurate and the reliability of the fault diagnosis model is improved.In order to verify the effectiveness of the proposed method,simulations are performed in the Pioneer3-AT mobile robot system.