Burr-XII分布在具有随机移走逐步增加截尾模型下可靠性指标的Bayes估计  

Bayesian estimation of reliability performances for the Burr-XII distribution based on progressive censoring with random removals

作　　者：孙天宇[1] 师义民[1] 王亮[1,2] 

机构地区：[1]西北工业大学应用数学系,西安710072 [2]西安电子科技大学理学院,西安710071

出　　处：《系统工程理论与实践》2012年第7期1476-1484,共9页Systems Engineering-Theory & Practice

基　　金：国家自然科学基金(70471057,71171164)

摘　　要：在具有随机移走的逐步增加Type-Ⅱ截尾模型下,假定每一个失效时刻移走的未失效样本数都服从二项分布,当Burr-XⅡ分布中两个形状参数均未知时,分别在q-对称损失函数和LINEX损失函数下给出了两个未知形状参数、移走概率以及可靠性指标的Baves估计和Bayes后验区间估计,并给出了参数估计的容许性证明.最后利用Monte Carlo方法对各种估计结果进行了模拟分析.This study considers the Bayesian estimation for the Burr-XII distribution based on progressive Type-II censoring with random removals under the assumption that the number of units removed at each failure time follows a binomial distribution. The two unknown shape parameters, the probability of random removals, reliability performances and their corresponding Bayesian posterior interval estimation are derived under q-symmetric entropy loss function and LINEX loss function, when both of the shape parameters are unknown, and the permissibility of estimation is also demonstrated. At last, some analyses of numerical results are performed by Monter Carlo simulations for illustrative purposes.

关 键 词：随机移走 Burr-X11分布 q-对称熵损失 LINEX损失 BAYES估计 

分 类 号：O231.2[理学—运筹学与控制论]