Expert knowledge modelling software design based on Signed Directed Graph with the application for PWR fault diagnosis

Online monitoring and Fault Diagnosis and Detection (FDD) can help the operator to enhance the situation awareness. The Signed Directed Graph (SDG) has the significant advantages that SDG-based FDD not only infers the incipient faults and reveals fault propagation paths but also comprehensively explains causes of failure. In this study, an expert knowledge modelling software is designed and developed based on SDG method. Firstly, the SDG-based process monitoring and FDD algorithms are derived by coupling the threshold method and the quality trend analysis method. Secondly, the expert knowledge modelling software is designed and developed not only considering the convenience of the engineer but also the effectiveness of the SDG-based algorithms. Finally, a Pressurized Water Reactor (PWR) is applied to demonstrate the SDG modelling and verify the SDG models. The case study demonstrates that the developed expert knowledge modelling software can effectively diagnose the incipient faults and infer variable impact as the fault propagation paths

Automated summary

Online monitoring and Fault Diagnosis and Detection (FDD) can improve operator's situation awareness. The study proposes an expert knowledge modelling software based on Signed Directed Graph (SDG) to infer faults and explain failure causes. The case study demonstrates the software's effectiveness in diagnosing faults and inferring variable impact along the fault propagation paths.