航空学报 > 2008, Vol. 30 Issue (6): 1550-1553

利用相似产品信息的电子产品可靠性Bayes综合评估

杨军,申丽娟,黄金,赵宇   

  1. 北京航空航天大学 工程系统工程系
  • 收稿日期:2007-10-09 修回日期:2008-02-19 出版日期:2008-11-25 发布日期:2008-11-25
  • 通讯作者: 赵宇

Bayes Comprehensive Assessment of Reliability for Electronic Products by Using Test Information of Similar Products

Yang Jun,Shen Lijuan,Huang Jin,Zhao Yu   

  1. Department of System Engineering of Engineering Technology, Beijing University of Aeronautics and Astronautics
  • Received:2007-10-09 Revised:2008-02-19 Online:2008-11-25 Published:2008-11-25
  • Contact: Zhao Yu

摘要: 在电子产品试验的可靠性评估中,为提高估计精度,经常利用历史样本数据来确定先验分布。但在工程实际中,历史样本和样本实际上属于不同的总体,这对可靠性评估结果有显著的影响。为此,采用相似系统分析确定历史样本和样本的相似程度,将其归纳为继承因子;然后根据历史样本信息确定产品可靠性的历史后验,基于无信息先验得到产品可靠性的更新后验;最后通过继承因子,综合历史后验和更新后验,得到产品可靠性的融合后验,并在此基础上进行可靠性推断。该方法不仅充分利用了相似产品的信息,而且突出了产品的独有特性。

关键词: 指数分布, 可靠性评估, Bayes分析, 相似系统分析, 继承因子

Abstract: In the reliability assessment of electronic products, the history sample and the current sample often belong to different populations, which may have remarkable effect on the result of the assessment. In this article, the similarity degree between the history sample and the current sample is studied by similarity system analysisand denoted by an inheritance factor. The history posterior is obtained from the history sample and the innovation posterior is given by the noninformative prior. The two posteriors reflect the history information and the current test information respectively. The complex posterior is obtained from the history posterior and the innovation posterior via the inheritance factor, and statistical inference is made on the basis of the complex posterior. This method makes full use of the information of similar products and emphasizes the particular characteristics of the evaluated product.

Key words: exponential distribution, reliability assessment, Bayes method, similarity system analysis, inheritance factor

中图分类号: