An Approach for Dynamic Reliability Assessment of Reactor Trip System of HPR1000

The structure of reactor trip system (RTS) is complex, which leads to its dynamic interaction, time dependence and probability uncertainty. However, the traditional static reliability assessment methods are difficult to characterize these three characteristics. To solve this problem, a novel approach for dynamic reliability assessment of the RTS of Hualong Pressurized Reactor 1000 (HPR1000) was proposed. Firstly, the dynamic fault tree (DFT) was used to establish the DFT model for the characterization of RTS dynamic interaction. Then, dynamic Bayesian network (DBN) and fuzzy set theory (FST) were used to establish fuzzy DBN model based on the established DFT model for the characterization of dynamic interaction, time dependence and probability uncertainty of the RTS. Finally, Latin hypercube sampling (LHS) was used to define a new fuzzy Bayesian inference algorithm. The defined algorithm was used to carry out the fuzzy Bayesian forward inference and backward inference, the dynamic reliability of RTS was calculated, and the weak points of RTS were identified. Comparing the defined fuzzy Bayesian inference algorithm with the traditional fuzzy Bayesian inference algorithm, the accuracy and precision of the algorithm defined in this paper were verified. The obtained research results provide a scientific basis for further improving the reliability of the RTS of HPR1000.