Research On Fault Diagnosis Technology Based On Multi-signal Flow Model

With the rapid development of industrial manufacuture,more and more complex systems have appeared.The complexity of systems leading to more complicated faults,which would cause the operation failure and result in huge time loss,property damage and even casualties.In order to improve the reliability and safety of system equipment,it is necessary to study the fault diagnosis technology of large and complex systems.In recent years,with the continuous development of diagnostic testing technology,industry experts have proposed the concept of testability design,considering the testability of faults during the period of system design.Carrying out fault diagnosis research and providing feedback on testability design is helpful for researchers to optimize system design;Fault diagnosis strategy optimzaion and fault propagation path identififcation are two crucial problems in testablilty analysis,and the efficiency of the related algorithms would influence the effects of diagnosis.This paper focus on these two problems,providing new ideas on improving the efficiency of related algorithms and making better applications of them to the testability analysis of large and complex systems.First of all,based on the theory of multi-signal flow graph,the modeling steps of multi-signal flow model are determined,and the radar transmitter system is used as an example for subsequent fault diagnosis and analysis.Secondly,several common algorithms for diagnosis strategy optimization,especially the AO* algorithm in sequential testing,are analyzed.On this basis,the AO* algorithm based on the information entropy heuristic function is proposed,which is used to he diagnosis strategy optimization in order to improve the efficiency of the algorithm.The algorithm stability is superior when applied to large and complex systems.The idea of this method is to reduce the process of constructing the Huffman tree in the AO* algorithm and reduce the running time of the algorithm to improve the efficiency of the algorithm.Then,the graph theory information of the multi-signal flow model and the basic information of the modules in the model are used to hierarchically reconstruct the multi-signal flow model,reflecting the distributed diffusion structure of the model to reduce the fault set and locate the fault source more effectively.After that,focus on the problem that historical data is insufficient in testability analysis,the structural index parameters of the fault propagation path are analyzed,and the fault propagation intensity is calculated to identify the fault propagation path in the system.Finally,the methods proposed in this paper are integrated into software to realize visual operation,and the ideas,principles and implementation of key functional modules related to software design are introduced.