广义逐步混合截尾下Marshall-Olkin 扩展指数分布的可靠性分析  

Reliability analysis under Marshall-Olkin extended exponential distribution based on generalized progressively hybrid censored scheme

作　　者：肖金安 贺兴时[1] 王燕[1] XIAO Jinan;HE Xingshi;WANG Yan(School of Science,Xi′an Polytechnic University,Xi′an 710048,China)

机构地区：[1]西安工程大学理学院

出　　处：《纺织高校基础科学学报》2019年第3期317-323,共7页Basic Sciences Journal of Textile Universities

基　　金：陕西省重点研发项目(2018W-021);陕西省教育厅自然科学专项基金(18JK0333);西安工程大学研究生创新基金(CHX2019058)

摘　　要：研究Marshall-Olkin扩展指数分布的可靠性指标。基于该分布的广义逐步混合截尾模型,通过经典估计和贝叶斯估计给出该分布的未知参数估计。利用数值迭代方法和渐近正态理论,给出未知参数的最大似然估计值及渐进置信区间。在先验分布为伽马分布的条件下,利用Metropolis-Hastings抽样算法得到了未知参数的贝叶斯估计值和最大后验密度可信区间。数值模拟结果表明,贝叶斯估计的均方误差和区间长度均优于经典估计方法。In order to study the reliability index of Marshall-Olkin extended exponential distribution, based on the generalized progressively hybrid censored model, the unknown parameters of the distribution are estimated by classical estimation and Bayesian estimation methods.By numerical iteration method and asymptotic normal theory,the maximum likelihood estimates of the unknown parameters are given.Under the condition that the prior distribution is gamma distribution, the Bayesian estimator and the maximum posterior density confidence interval of unknown parameters are obtained by using Metropolis-Hastings sampling algorithm.The numerical simulation results show that the mean square error and interval length of Bayesian estimator are better than those of classical estimator.

关 键 词：Marshall-Olkin扩展指数分布 广义逐步混合截尾 Metropolis-Hastings抽样算法 参数估计 

分 类 号：O213.2[理学—概率论与数理统计]