Research On Key Technology Of Fault Diagnosis Of Virtual Maintenance Platform

Virtual maintenance platform is a software platform developed based on virtual simulation technology to simulate the disassembly and assembly operation of equipment components.Maintenance personnel can carry out no physical maintenance operation practice on the platform,quickly get familiar with the equipment structure,and improve the maintenance proficiency.In order to enrich the training form of virtual maintenance platform and improve the integrity and simulation of maintenance process,fault diagnosis module is added to the virtual maintenance platform,and the function of demonstrating the principle of fault diagnosis and visualizing the whole process of diagnosis is realized.Referring to the basic structure of expert system,the key technologies of fault diagnosis based on fault tree structure are studied.To simplify the diagnostic operation process,reduce the manual diagnosis workload and improve the efficiency,a new visual tracking reasoning method based on the minimum cut sets and structure importance degree is proposed.Firstly,based on the structure importance degrees of basic events,the recommended diagnostic sequence is initialized to guide users’ diagnostic steps.Then,by tracking users’ diagnosis results of basic events and combining with the minimum cut sets,undiagnosed nodes and cut sets are excluded or locked,the recommended diagnosis sequence is updated in real time,and the diagnosis process is dynamically displayed until users find all fault points according to the guidance.The case application results show that,compared with the traditional fault tree reasoning method,this method can dynamically modify the diagnosis direction and significantly reduce the number of diagnoses.To realize the visual display of operation steps and reasoning principles in the process of fault diagnosis and improve the interactive performance,the process of visual diagnosis is designed based on the traditional expert system.Focusing on knowledge representation and visual display method of fault tree structure,the whole process of fault diagnosis is divided into five stages: knowledge acquisition,fault setting,operation monitoring,diagnosis reasoning and detection troubleshooting.The business process of each stage is designed and implemented.Finally,the availability and visibility of the visual diagnosis process and the fault diagnosis module of the virtual maintenance platform are proved by the application analysis of the fault tree of pitch motion complete failure of the flight control system.