An Modular method for System Reliability Analysis based on Binary Decision Diagram and Bayesian Network

Reliability analysis plays a critical role in operation safety of complex system, especially for rail transport system. This paper presents a modular method,using Binary Decision Diagrams and Bayesian Networks techniques, to analyze system reliability. System fault tree is divided into some independent subtrees which are solved with different solution techniques depending their characteristics, Binary Decision Diagrams for static fault trees, and Bayesian Networks for dynamic fault trees. Then system reliability can be obtained by integrating the solution of all independent subtrees. The modular analysis approach, using Binary Decision Diagrams and Bayesian Networks, can over main shortcomings of Markov Chains, and efficiently analyze reliability of complex system.