Bayesian Estimation of Residual Life for Weibull Distributed Components by Fusing Expert Knowledge

Abstract

Residual life estimation is of crucial significance in reliability engineering. Traditional methods are limited when components are characterized as high reliability, long life and small sample size. However, in practical engineering, experts can provide valuable reliability information, which could obviously improve the estimation precision of residual life when compared with conventional approaches. In this paper, a Bayesian method of residual life estimation by utilizing expert knowledge is proposed. Firstly, the prior distribution from expert information is determined. After fusing field lifetime data, posterior distribution can be obtained. Finally, residual life is estimated by the proposed sample-based method and both the Bayes estimate and credible interval are obtained. The proposed method in this paper is validated by the simulation study and the results prove the proposed method is rather satisfactory and robust.