Reliability Evaluation for Complex System Based on Bayesian Theory and Multi-Source Information Fusion

Abstract

For large complex systems, it is of great importance to study effective approaches for reliability evaluation. In this paper, a system reliability evaluation method based on Bayesian theory and multi-source information fusion is proposed. Firstly, model the reliability of system network, including the establishment of system reliability block diagram and the determination of the system and components’ life distributions. Then, the posterior distributions and posterior moments of component reliability are determined. With the posterior moments, the failure information of components can be converted to the prior moment of the system reliability according to the system structure, and the system prior distribution can be obtained. Finally, combined with the system-level field data, the posterior distribution of system reliability can be obtained on the basis of Bayesian theory, and be utilized to evaluate the reliability of system. To fill out the shortage of field data, the proposed method in this paper can make full use of various quantitative and qualitative prior information with high accuracy.