Research on Condition Monitoring Technology for Nuclear Power Plant Equipment Based on Kernel Principal Component Analysis

In order to solve the limitations of the traditional monitoring methods for nuclear power plants, this paper proposes to introduce Kernel Principal Component Analysis (KPCA) into the online monitoring field of nuclear power plant equipment, and design the monitoring method and online monitoring strategy. In order to verify the effectiveness of the algorithm, it has been applied in the real monitoring case of the motor driven main feed water pump a nuclear power plant in China. The simulation results show that the KPCA algorithm can adapt to the requirements of nuclear power plant equipment monitoring, and can provide earlier warnings of failure than the existing threshold monitoring methods. At the same time, compared with the conventional PCA algorithm, the KPCA algorithm can extract the nonlinear relationship between variables, identify different operating modes of the device, and effectively reduce false alarms.