基于ML-Ⅱ的指数分布可靠性多层Bayes估计

Multilayer Bayes Assessment for Reliability of Exponential Distribution Based on ML-Ⅱ Method

Bayes估计法是可靠性评估中应用最为广泛的方法之一,指数分布的Bayes验前概率密度函数中的重要参数主要依靠Reformulation法和Box-Tiao法确定,具有较强的主观经验性。基于Beyes估计的基本思想,以试验数据为依据,利用第二类极大似然估计法(ML-Ⅱ估计法)确定Bayes方法中的相关参数,避免了参数确定的主观性。实例表明结果合理,方法客观、可行。

：Bayes is one of the widely used reliability estimation methods. The key parameters of for- met probability are ascertained with Reformation and Box-Tiao methods, which depends on subjective ex- perience. Based on muhilayer Bayesian theory and test data, these parameters can be confirmed by cal- culation in the second maximum likelihood estimation（ ML-Ⅱ）. This method avoids subjective character- istic of expert experience method. Examples illustrate that the result is reliable and the method is objec- tive and feasible.