突发和退化竞争失效产品的老练时间优化  

Optimal burn-in time for competing failure products involving catastrophic and degradation failures

作　　者：黄秀平[1] 汤衍真[1] 孙权[1] 周经伦[1] 

机构地区：[1]国防科学技术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》2013年第8期1803-1808,共6页Systems Engineering and Electronics

基　　金：国家自然科学基金(60804054)资助课题

摘　　要：由于设计与材料的缺陷以及制造过程中的不稳定性,通常在一批产品的生产过程中会产生一定比例具有较高突发失效率的缺陷产品。为了尽量剔除缺陷产品,提高整批产品的实际使用可靠度并减少寿命周期费用,需要进行老练筛选并确定最优的老练时间。在假设突发失效时间服从指数分布、性能退化轨道以Wiener过程描述的基础上,针对同时包含突发失效和退化失效的产品,本文首先建立了包含退化失效的混合竞争失效模型;然后以平均剩余寿命(mean residual life,MRL)为目标函数建立了老练试验优化模型的有效性,并给出了数值求解方法;最后通过数值实例验证了所提出的老练试验优化模型的有效性,并对模型参数进行了灵敏度分析。Owing to the defect in design and material or the instability of productive process, there are always a certain proportion of defect products in which catastrophic failure occurs with a higher probability in a whole lot of products. In order to eliminate the catastrophic failure products as fully as possible, it is needed to conduct a burn-in screening test and determine the optimal burn-in time. Besides, the achieved reliability of the whole product can be improved and the life cycle cost can be reduced. Based on the assumption that catastrophic failure time follows an exponential distribution and the performance degradation path can be characterized by a Wiener process, this paper firstly establishes the mixed competing failure model for the products involving catastrophic and degradation failures. Then, the burn-in test optimization model is presented, where the objective function is the mean residual life （MRL）, and the numerical solution is derived. Finally, the proposed model is illustrated with a numerical example, and the sensitivity analysis is conducted for the model parameters.

关 键 词：老练试验 优化 竞争失效 平均剩余寿命 性能退化 

分 类 号：TB114.3[理学—概率论与数理统计]