北京航空航天大学学报 ›› 2019, Vol. 45 ›› Issue (11): 2237-2246.doi: 10.13700/j.bh.1001-5965.2019.0130

• 论文 • 上一篇    下一篇

基于随机相关的电子部件二元加速退化可靠性评估

盖炳良1,2, 滕克难1, 王浩伟1, 王文双1, 陈健1, 宦婧3   

  1. 1. 海军航空大学 烟台 264001;
    2. 中国人民解放军 91115部队, 舟山 316000;
    3. 江苏科技大学 计算机学院, 镇江 212003
  • 收稿日期:2019-03-27 出版日期:2019-11-20 发布日期:2019-11-30
  • 通讯作者: 滕克难.E-mail:Tengkn@sina.com E-mail:Tengkn@sina.com
  • 作者简介:盖炳良 男,博士研究生。主要研究方向:装备可靠性评估、加速试验技术等;滕克难 男,博士,教授,博士生导师。主要研究方向:装备延寿理论与技术。
  • 基金资助:
    国家自然科学基金(51605487)

Reliability assessment for electronic components with bivariate accelerated degradation based on random correlation

GAI Bingliang1,2, TENG Kenan1, WANG Haowei1, WANG Wenshuang1, CHEN Jian1, HUAN Jing3   

  1. 1. Naval Aviation University, Yantai 264001, China;
    2. Unit 91115 of the PLA, Zhoushan 316000, China;
    3. School of Computer Science, Jiangsu University of Science and Technology, Zhenjiang 212003, China
  • Received:2019-03-27 Online:2019-11-20 Published:2019-11-30
  • Supported by:
    National Natural Science Foundation of China (51605487)

摘要: 针对加速应力下电子部件二元相关退化可靠性分析难题,提出一种基于随机相关的可靠性分析方法。采用考虑个体差异的Wiener过程模型建立边缘退化过程模型,并基于加速因子不变原则建立了模型参数与加速应力的关系;构建了基于Copula函数的随机相关模型,采用两阶段贝叶斯参数估计方法进行参数估计,综合运用散点图、偏差信息准则(DIC)值以及Kendall τ的非参数估计值等方法进行随机相关模型选择,并采用蒙特卡罗仿真方法进行可靠度计算。最后采用实例验证了所提方法有效性,为考虑个体差异的贮存可靠性评估提供了技术支撑。

关键词: 可靠性评估, 加速退化数据, 二元退化, 随机相关, 电子部件

Abstract: Targeting at the difficulty of reliability analysis for electronic components with bivariate correlation accelerated degradation data, a reliability assessment method based on random correlation is proposed. The Wiener process model with random effect is used to model the marginal degradation process considering the individual difference, and the relationship between model parameters and acceleration stress is established by using acceleration factor constant principle. Then, a bivariate degradation model with random correlation based on Copula function is established. A two-stage Bayesian method is introduced to facilitate the parameter estimation, and the scatter plots, deviance information criterion (DIC) and the non-parametric estimation of Kendall τ are used for random correlation model selection. The reliability calculation is carried out by Monte Carlo simulation method. Finally, an example is used to verify the effectiveness of the proposed method.The paper has significant meaning for the storage reliability assessment considering individual differences.

Key words: reliability assessment, accelerated degradation data, bivariate degradation, random correlation, electronic component

中图分类号: 


版权所有 © 《北京航空航天大学学报》编辑部
通讯地址:北京市海淀区学院路37号 北京航空航天大学学报编辑部 邮编:100191 E-mail:jbuaa@buaa.edu.cn
本系统由北京玛格泰克科技发展有限公司设计开发