Research on the Construction of Intelligent Fault Prediction and Diagnosis Model for RPN Intermediate Range ACCG4 Board Under Multi-Base Collaboration

In the external nuclear measurement system (RPN) of a nuclear power plant, the ACCG4 board plays a key role. It is the core component of signal conversion, and its performance is directly related to the accuracy of reactor power monitoring and the stability of the system. This study proposes a new intelligent fault prediction and diagnosis model based on a multi-base collaborative strategy to improve the fault diagnosis efficiency and accuracy of the ACCG4 board. The study first established a multi-base collaborative fault diagnosis architecture. By integrating ACCG4 panel operation data from different nuclear power plants, a shared data resource library was formed, and big data technology was used to improve the accuracy of fault prediction. GNN is responsible for parsing the network structure relationship between boards, while CNN focuses on mining the deep features of electrical signals. Through testing in simulated environments and multiple fault scenarios, the study found that the intelligent fault prediction and diagnosis model improved the fault prediction accuracy by 35% and the diagnosis accuracy by 30% under various circumstances, and the diagnosis time was compared with traditional Methods were reduced by 45%. These results fully prove the significant effect of this model in optimizing the fault diagnosis performance of ACCG4 panels, and provide solid technical support for the stable operation of the RPN system of nuclear power plants. In addition, this research further expands the applicable scope of the model and provides new ideas and methods for intelligent operation and maintenance of nuclear power plants.