Research On Data Fusion Diagnostic Method Of Marine Nuclear Power Plants

The structure of marine nuclear power plants is very complicated, and the requiring to safety of marine nuclear power plants is very high. In order to ensure it run safely, it is necessary to research the diagnostic system that can diagnose the faults of marine nuclear power plants in time. Nowadays the development of diagnostic technology is very fast, but the research to fault diagnosis of marine nuclear power plants is not too much. In this paper the data fusion method is imported to the fault diagnosis of marine nuclear power plants. The fusion diagnosis system is built up using the combining method of neural networks and D-S evidence theory and using the tool of VB6.0 in WINDOWS98. The several main faults of gas turbine rotor and the primary loop of marine nuclear power plants are diagnosed by the system on the tester and simulator respectively based on the research of the faults mechanism and phenomena of marine nuclear power plants. The analyses of the results indicate that the diagnosis system can diagnose the faults exactly in time, and the precision is very high. It has the character of extended in the practice application because it can study. So the fault diagnosis system can satisfy the diagnostic require of marine nuclear power plants, it has the important application value.