考虑失效相关性的长寿命系统可靠性置信限综合评估模型  

Comprehensive Reliability Confidence Limit Assessment of Long-lived System Involving Failures Correlation

作　　者：唐家银[1] 陈崇双[1] 锁斌[2] 何平[1] 梁红琴[3] 

机构地区：[1]西南交通大学数学学院统计系,四川成都611756 [2]中国工程物理研究院电子工程研究所,四川绵阳621900 [3]西南交通大学机械工程学院,四川成都611756

出　　处：《数理统计与管理》2017年第5期866-878,共13页Journal of Applied Statistics and Management

基　　金：教育部人文社会科学研究青年基金(11YJCZH154);国家自然科学基金(51205326);中央高校基本科研业务费专题项目(2682014ZT29);中国铁路总公司科技研究开发计划项目(2013J006-B);四川省统计科学研究计划项目(2016sc50)

摘　　要：根据高长寿命产品的试验数据特征,采用退化轨迹拟合、两点分布和区间估计思想,分位数置信限-失效阈值干涉、可靠度预估响应模拟,分别给出了小样本和大样本试样下的零部件可靠性置信限评估方法。运用Copula理论拟合零部件失效间的动态相关性,从而对应性地建立了系统级可靠性置信限综合评估模型。该模型改善了独立性系统可靠性置信限评估方法预估结果过于保守的情形。算例验证了该类模型的可行性和有效性。According to the characteristics of test data of long-lived products, fitting degradation path, two-point distribution, interval estimation were combined, and based on the interference between quantile confidence limits and failure threshold, reliability estimated response simulation, the methods to assess the reliabilities of components were presented under the small sample and large samples respectively. Copulas theories were used to fit the dynamic correlation among failures of components, then the reliability comprehensive evaluation models were established on the level of systems, which can improve the excessive conservatism resulted from independence. Finally, an example is given to demonstrate the feasibility and effectiveness of the proposed models.

关 键 词：失效相关性 长寿命系统 可靠性 置信限 退化轨迹 COPULA 

分 类 号：O212[理学—概率论与数理统计] TB114.3[理学—数学]