无失效数据场合指数分布可靠度Bayes估计的改进

Improvement of Bayesian Estimation of Exponential Distribution Reliability for Zero-Failure Data

在无失效数据场合下,根据指数分布的凸性和无记忆性,修正了可靠度R_(i)的取值上界.在平方损失下,得到了R_(i)的Bayes估计,利用加权最小二乘法得到平均寿命的估计.最后进行了算例分析,结果表明,将可靠度估计精度提高了98.68%.

In the case of no failure data,according to the convexity and non-memory of exponential distribution,the upper bound of reliability value of R_(i)(i=2,3,…,m)is modified.Under the square loss,the Bayes estimate of R_(i) is obtained.By use of the least squares,an estimate of the average life span is worked out.Finally,an example is given to show that the accuracy of reliability estimation is increased by 98.68%.