Host: The Japan Society of Mechanical Engineers
Name : The 30th International Conference on Nuclear Engineering (ICONE30)
Date : May 21, 2023 - May 26, 2023
Offshore small natural circulation lead-cooled fast reactors have harsh working environment and few people on duty. The operation is heavily relied on physical quantities measured by sensors to control the reactor. If the sensor fails and the physical quantity measurement of the reactor is inaccurate, the real operating state of the reactor cannot be reflected, thus affecting the operation safety. Therefore, a fault diagnosis method is necessary to diagnose faults and deal with them in time.
In this paper, the fault diagnosis method based on the combination of Principal Component Analysis (PCA) and Support Vector Machine (SVM) is proposed for sensor fault diagnosis. This data-driven method can effectively reduce the data dimension and does not require the specific state of the system and the precise mathematical model of the processing object. It can play a good role in the diagnosis object such as the reactor with complex internal structure and difficult to detect characteristics. The fault diagnosis model is established using the sensor data obtained from the dynamic model of the small lead-cooled fast reactor constructed in MATLAB/Simulink. The sensor fault data is introduced for testing. The results show that this fault diagnosis method has higher diagnosis accuracy and faster diagnosis time than other methods.
