The application of integrated AI fault diagnosis method in the missile test system

Aiming at the characteristic and limitation of various generic fault diagnosis methods, a novel integrated AI fault diagnosis design method suitable for the results analysis in the missile test system is present. In this method, UUT (Unit under Test) model, knowledge base, Fuzzy theory and Neural Network model are used to express the domain knowledge in the field of missile test. Then, productive rule based expert system is used to achieve the knowledge organization and usage. Comparison, analysis and explanation of diagnostic results and machine learning are also completed with expert system. Mission assignment and data exchange between the different methods are accomplished by using SQL database and ODBC. This kind of method overcomes the limitation of single fault diagnosis method. It can effectively enhance the developing speed of the missile fault diagnosis system and the work efficiency of the whole system.