长寿命卫星活动部件Bayes-Weibull可靠性评估方法  

Bayes-Weibull reliability assessment method for long life satellite moving components

作　　者：金光[1] 冯静[1] 

机构地区：[1]国防科技大学信息系统与管理学院,湖南长沙410073

出　　处：《系统工程与电子技术》2009年第8期2020-2023,共4页Systems Engineering and Electronics

摘　　要：长寿命卫星活动部件可靠性评估具有小样本、无失效数据的特点,采用大样本寿命数据统计方法或通常的无失效数据分析方法难以解决寿命分布建模与可靠性评估问题。针对Weibull寿命型卫星活动部件,提出了一种有效的Bayes可靠性评估方法。综合利用相似型号活动部件数据估计Weibull分布形状参数,利用相似因子和参数Bootstrap方法构造Bayes验前分布。实例分析表明,该方法具有较高的精度,可以解决小样本无失效条件下,长寿命卫星活动部件可靠性评估的困难问题。The reliability assessment of long life satellite moving components has small sample size and zerofailure, so classical large sample life data based statistical methods and current zero-failure data analysis methods are no longer valid for the life distribution modeling and reliability assessment of the components. This paper studies Weibull life type satellite moving components, and provides an effective Bayes reliability assess- ment method, which synthesizes various data of similar models and constructs Bayes prior distribution using similarity factor and Bootstrap method. Example analysis shows that the provided method has higher precision and provides a valid way of solving reliability assessment difficulties of long life satellite moving components under small sample and zero-failure.

关 键 词：可靠性评估 小样本 无失效 活动部件 BAYES方法 

分 类 号：TP114[自动化与计算机技术—控制理论与控制工程]