基于多元退化数据的可靠性分析方法  

Reliability analysis approach based on multivariate degradation data

作　　者：钟强晖[1] 张志华[2] 梁胜杰[1] 

机构地区：[1]海军工程大学兵器工程系,武汉430033 [2]海军工程大学应用数学系,武汉430033

出　　处：《系统工程理论与实践》2011年第3期544-551,共8页Systems Engineering-Theory & Practice

基　　金：国家高技术研究发展计划(863计划)(2007AA11Z216);国家自然科学基金(50708094)

摘　　要：针对具有多个退化特征量的产品,提出了基于多元退化数据的一般可靠性分析方法和处理步骤.首先通过相关分析对退化特征量进行筛选,利用退化特征量分布与寿命分布的关系估计产品在各检测时刻的可靠度,最后通过产品寿命分布拟合进行可靠性评估.由于不依赖退化轨迹的假设,该方法具有较好的适用性,而对退化特征量的相关性分析使计算更加容易,并为产品检测指标的简化提供参考.最后给出仿真例子验证了该方法的可行性和有效性.This paper presents a general reliability analysis approach for products having multi-degradation characteristic variable. In this approach degradation characteristic variables are filtered through correlation analysis, then the reliability at each inspection time point is estimated based on the relationship between degradation characteristic variables＇ distribution and life distribution. Reliability assessment is performed through distribution fitting. This approach has better adaptability due to being independent of degrada- tion path assumptions. Furthermore, correlation analysis of degradation characteristic variables simplifies calculation and provides guidance for reducing inspection index. Finally, this paper gives a simulation example to demonstrate the feasibility and validity of this approach.

关 键 词：多元退化数据 主成分分析 多元正态分布 可靠性 

分 类 号：TB114.3[理学—概率论与数理统计]