摘要: |
航天产品现场使用数据具有小子样、零失效的特点,要实现其工程化可靠性评估,需要充分利用各种环境下的试验信息和相似型号产品的试验信息及使用信息,因此,提出了一种适用于Weibull分布航天产品可靠性综合评估的MMLE—Bayes方法。首先通过MMLE方法得到额定任务时间可靠度的置信下限估计,再通过构造合理的验前分布实现Bayes可靠性评估。在此过程中,为合理融合各来源的验前信息,提出了环境因子和相似因子的概念及其估计方法。最后通过一种航天产品的可靠性评估实例,说明了MMLE—Bayes—Weibull方法的的实用性和有效性。 |
关键词: 航天产品 可靠性评估 Weibull分布 零失效 修正极大似然估计方法 相似因子 环境因子 |
DOI:10.3969/j.issn.1001-893X. |
投稿时间:2006-09-18修订日期:2006-11-10 |
基金项目:国家高技术研究发展计划(863计划) |
|
MMLE-Bayes Reliability Assessment Method for Aerospace Products with Weibull Distribution |
|
() |
Abstract: |
The senery data of aerospace products is of such characteristics as small sample and zero failure. In order to realize the reliability assessment, prior information from multiple sources such as test data in different environment and test data of similar product should be well used by environmental factor and similar factor. The method named MMLE - Bayes is proposed, in which the confidence low level of task reliability is obtained through modified maximum likehood method. Finally, an example is given to illustrate the efficiency of the MMLE - Bayes - Weibull method. |
Key words: aerospace product,reliability assessment,Weibull distribution,zero failure,modified maximum likelihood estimation method(MMLE),similar factor,environmental factor |