系统工程与电子技术

• 可靠性 • 上一篇    下一篇

基于顺序Dirichlet分布的Bayes可靠性增长评估方法

邢云燕1, 蒋平2   

  1. 1. 国防科学技术大学军事高科技培训学院, 湖南 长沙 410073;
    2. 国防科学技术大学信息系统与管理学院, 湖南 长沙 410073
  • 出版日期:2017-04-28 发布日期:2010-01-03

Bayesian evaluation method for reliability growth test based on the ordered Dirichlet distribution

XING Yunyan1, JIANG Ping2   

  1. 1. College of Continuing Education, National University of Defense Technology, Changsha 410073, China; 2. College of
    Information System and Management, National University of Defense Technology, Changsha 410073, China
  • Online:2017-04-28 Published:2010-01-03

摘要:

针对产品可靠性增长试验过程中试验样本量少这一特点,提出一种基于顺序Dirichlet分布的Bayes可靠性增长评估方法。利用顺序Dirichlet分布的统计特性,对产品可靠性增长试验过程中的专家信息进行规范化描述,并在Bayes试验鉴定框架下作为验前信息给出其验前分布的确定方法,以及产品可靠性增长过程中各试验阶段产品可靠性的评估方法。通过综合考虑产品各试验阶段的历史信息,主观信息和客观信息等多种信息形式,解决了小样本情况下产品可靠性增长试验性能指标的评估问题。

Abstract:

Considering the characteristic that sample size of products used for the reliability growth test is small, a Bayesian evaluation method for reliability growth test based on the ordered Dirichlet distribution is proposed. Standardized description procedure of expert information during the reliability growth process and prior distribution quantified from expert information in Bayesian framework are figured out by applying statistical features of the ordered Dirichlet distribution. The evaluation method of system reliability at each reliability growth test stage is also presented. Due to combining historic information, subjective information and objective information, the evaluation problem about system reliability growth test under the condition of small sample size is solved efficiently.