Research On Fault Diagnosis Expert System Of Electro-optical Tracking Equipment Based On CLIPS

The electro-optical tracking equipment which consist of many small, complex systems is an optical, mechanical and electric integration. From the system level, traditional modeling methods and the signal processing methods both cannot achieve the requirements of fault diagnosis, while in the field of artificial intelligence, expert system can realize fault diagnosis only based on the field experience and theoretical knowledge. In recent years, the expert system has made breakthrough progress in the field of fault diagnosis of large equipment in aerospace, thus laying a solid technical foundation for the study of fault diagnosis for the electro-optical tracking equipment.The article details the composition and working process of the equipment, classifies the equipment faults and summarizes the final reasons for every fault category, then establishes fault tree for the equipment taking the events which mainly prevent operation of the equipment as top event. The fault tree which contains all the fault events of the equipment is an inverted tree where lower events are the causes of the upper event thus can expressing the fault information and the relationship among the faults. According to the characteristics of equipment fault, the knowledge representation combining production rules and frame representation is used to transform the information in fault tree into the knowledge can be stored in computer to establish the knowledge base of expert system. Forward reasoning and the control strategy based on confidence determined by fault probability and difficulty to find fault are adopted to design reasoning machine, and expert rules are designed to solve mutually exclusive knowledge.Expert system programming language CLIPS which is good at writing semantic analysis program and dealing with predicate logic is taken to design the knowledge base and the reasoning machine which are the core parts of expert system to realize fault diagnosis for the tracking equipment. MFC application framework in VC is chosen to build expert system interface for friendly human-computer interaction. The article designs the fault knowledge table, the repair and maintenance table and the diagnosis case table in Microsoft SQL Server databases to storage fault knowledge of the equipment in order to maintain and manage knowledge better. The direct-embedding methods and the ADO interface technology are picked out to embedding CLIPS and SQLSEVER into MFC, and using the technology of automatic code generation to update database in SQL Server and the knowledge base in CLIPS simultaneously. The expert system combines the three at together to achieve fault diagnosis for the electro-optical tracking equipment. In last part, the article introduces major function of the expert system through the pictures and gets tests by specific examples. Test results show that the expert system can accurately judge the faults and effectively manage the knowledge.