无失效数据场合威布尔分布失效概率的Bayes估计

Bayesian Estimation of Failure Probability of Weibull Distribution with no Failure Data

在无失效数据场合下,通过对威布尔分布函数形状参数m的不同取值范围进行分析,得到对应失效概率p i的取值范围.将关于p i的减函数(1-p i)2作为先验分布的核,对p i进行Bayes估计,再结合加权最小二乘法估计可靠性参数.最后进行了算例分析,结果表明,本文所提出的方法提高了失效概率p i的估计精度.

In the case of no failure data,by analyzing the different value range of shape parameters of Weibull distribution function,a corresponding range of failure probability was obtained.Taking the decreasing function of failure probability as the kernel of prior distribution,Bayes estimation was carried out,and then the reliability parameters were estimated by distribution curve method.Finally,an example was given to illustrate the effectiveness of the proposed method.