Reliability Analysis of Multi-State Systems Based on Bayesian Network

The traditional fault tree analysis is very difficult to solve the reliability problem of multi-state system with various failure modes. In this paper, a reliability analysis method for multi-state systems based on Bayesian network is proposed. The general steps of building a Bayesian network of multi-state systems are provided, and the calculation method of the consequence probability and the importance degree of RRW, RAW and FV are also given in this paper. The proposed Bayesian network model can be used to describe the multi-state system well, and it can take full advantage of historical information for system fault diagnosis and prediction. Finally, an application example is illustrated to verify the effectiveness and feasibility of the proposed method.