摘要
利用配分布曲线法处理威布尔分布无失效数据求可靠性参数估计时,最关键是对失效概率pi(i=1,2,...,k)的估计。失效概率p,估计的精度决定了最终可靠性参数估计的精度和可靠度估计的精度。文章利用理论观点,即后验分布是对先验分布的调整,在多层Soyes估计的基础上,提出了一种利用后一时刻失效概率的后验分布来修正前一时刻失效概率先验分布,即修正截尾时刻失效概率的取值上界的方法。通过轴承实验的具体实例,验证了方法的稳健性和可行性。
When the distribution distribution curve method is used to estimate the reliability parameters of Weibull distribution data without failure, the most important thing is to estimate the failure probability [PI(I =1,2... k)]. The accuracy of failure probability estimation determines the accuracy of the final reliability parameter estimation and the reliability estimation. Making use of the Bayes theory, that is, the posterior distribution is the adjustment of the prior distribution, and based on the Bayes multilayer estimation, this paper proposes a posterior distribution of the post failure probability to modify the prior distribution of the prior time failure probability, namely the method of correcting the value upper bounds of the failure probability at the truncation time. Through the specific example of bearing experiment, the paper verifies the robustness and feasibility of the proposed method.
作者
李爽
李云飞
高海军
Li Shuang;Li Yunfei;Gao Haijun(College of Mathematics and Information,China West Normal University,Nanchong Sichuan 637009,China)
出处
《统计与决策》
CSSCI
北大核心
2019年第24期9-12,共4页
Statistics & Decision
基金
四川省教育厅自然科学重点项目（13ZA0016）
西华师范大学英才科研基金项目（17YC381）
关键词
威布尔分布
配分布曲线
改进失效概率
取值上界
Weibull distribution
distribution distribution curve
improved failure probability
upper bound of value
