A Bayesian Reliability Evaluation Method Based on Dempster-Shafer Evidence Theory

Bayesian method is an important and suitable approach for system reliability evaluation in small sample cir-cumstance. In order to reduce the testing data, all kinds of a priori information must be used, so the information fusion of multiple sources becomes a key problem in Bayesian analysis. A new method based on the Dempster-Shafer evidence theory is proposed to realize the information fusion of multiple sources, which can integrate more a priori distributions obtained through different fusion criteria. Simulation shows that the proposed method is effective.