Study on fault diagnosis of nuclear power plant based on grey cluster theory

Because of the grey features of symptom parameters when nuclear power plant is in failure state, This paper brings forward grey cluster models for fault diagnosis in nuclear power plant. Two methods are introduced in this paper to structure the grey cluster models. One cluster model based on AB0 type grey correlation can analysis the fault type by ranking the AB0 type grey correlations between the collected sequences and standard fault pattern sequences, the another cluster model based on grey whitening weight function can confirm the fault type by calculating the grey cluster values between the collected sequences and standard fault pattern sequences. An example of the steam generator typical failures is introduced to verify the feasibility of grey cluster analysis method for fault diagnosis in nuclear power plant. The results show that the grey cluster method has the advantages of simple model and reliable fault diagnosis results. (authors)