Contrastive Research on Two Kinds of Real-Time Fault Diagnosis Systems of Nuclear Power Plants

In order to guarantee the safe operation of nuclear power plants,the real-time fault diagnosis systems are developed using neural network technology and data fusion technology and adopting VB6.0 pro-gramming languages,and then the intelligence diagnosis arithmetic is expatiated in detail.The fault diagnosis systems interchange the data with the simulator timely utilizing communication procedure interface of the data,and four faults are inserted on the simulator to test the two diagnosis systems on line.The test result indicates that the fuzzy neural network technology and the data fusion technology could carry out the recog-nition of the faults,but each has its merit and the insufficiency respectively.The off-line analysis shows that,for different fault types,when there is few characteristic parameters,the fuzzy neural network diagnosis technology is better;when there are many characteristic parameter,the data fusion diagnosis technology is better.