Reliability evaluation of Wiener degradation system based on Bayesian network

Dynamic Bayesian network method is widely used for system reliability analysis. However, this method does not consider the degradation performance of system. In this paper, we present a reliability evaluation method for a complex system based on Wiener process and Dynamic Bayesian network. Firstly, we establish the fault tree based on the system reliability block diagram and establish the Dynamic Bayesian network according to the time characteristics of nodes. For the case that system degradation performance subjects to Wiener process and the sample size is small, we utilize an objective Bayes method to model Wiener process. By combining Wiener process and Dynamic Bayesian network, we get the system reliability characteristics and utilize Dynamic Bayesian network derivation ability to evaluate reliability and residual life of the complex system. Finally, a case study on steam turbine power system with reliability degradation testing is used to demonstrate accuracy and effectiveness of proposed method.