变点模型下Weibull分布恒加试验的Bayes分析  

BAYESIAN ANALYSIS OF CONSTANT-STRESS ACCELERATED LIFE TESTING OF WEIBULL DISTRIBUTIONS WITH A CHANGE POINT

作　　者：黄月兰[1] 汤银才[2] 

机构地区：[1]广西民族师范学院数计系,崇左532200 [2]华东师范大学金融与统计学院,上海200241

出　　处：《系统科学与数学》2013年第9期1105-1112,共8页Journal of Systems Science and Mathematical Sciences

基　　金：国家自然科学基金(11271136)资助项目

摘　　要：加速寿命试验是一种极为有效而经济的寿命试验方法,其理论日趋成熟,在实践中开始得到应用和推行.文章给出了失效机理产生突变的Weibull分布定时和定数场合恒定应力加速寿命试验的Bayes分析,并利用Gibbs抽样方法和MH算法解决了分布的形状参数取为连续先验时各参数的Bayes估计.模拟结果表明,文中所建立的模型及所用的方法是有效和优良的.Accelerated life test is a very effective and economical way of life tests. This paper gives a Bayesian statistical analysis of Weibull constant-stress ccelerate life test, when there is sudden change of failure mechanism. The Bayesian estimation is given based on the Gibbs sampling technique and MH algorithm. The simulation result shows that the model and the method proposed is efficient and well-behaved.

关 键 词：恒加寿命试验 变点 GIBBS抽样 MH算法 BAYES分析 

分 类 号：O212.8[理学—概率论与数理统计]