基于二元相关的器件级可靠性建模与剩余寿命预测研究  

Research on Device Level Reliability Modeling and Residual Life Prediction Based on Binary Correlation

作　　者：景博[1] 焦晓璇 胡家兴 黄以锋 孟向臻 JING Bo;JIAO Xiaoxuan;HU Jiaxing;HUANG Yifeng;MENG Xiangzhen(Aviation Engineering School,Air Force Engineering University,Xi'an710038,China;Aviation University of AirForce,Changchun 130000,China)

机构地区：[1]空军工程大学航空工程学院,西安710038 [2]空军航空大学,长春130000

出　　处：《强度与环境》2023年第6期70-78,共9页Structure & Environment Engineering

基　　金：陕西省自然科学基金(2022JQ-586);预先研究项目(50902060401)。

摘　　要：器件作为部件及设备的主要组成部分,其可靠性及健康状态对设备功能完好性及任务可靠性具有重大影响。本文针对复杂应力条件下单一性能退化的可靠性建模不能全面地评估器件的状态和退化,提出基于Copula函数的二元性能退化联合建模的流程和方法,将联合建模分解为边缘退化模型选择与确定和性能参数之间关联性的构建两部分。同时,基于卡尔曼滤波方法,提出二元性能退化的剩余寿命预测方法,并利用芯片试验数据进行验证,应用于基于两焊点退化的芯片可靠性建模和剩余寿命预测。As the main component of components and equipment,the reliability and health status of devices have a significant impact on the functional integrity and task reliability of the equipment.This article proposes a process and method for binary performance degradation joint modeling based on Copula function,which aims to address the issue that reliability modeling with single performance degradation under complex stress conditions cannot comprehensively evaluate the state and degradation of devices.The joint modeling is decomposed into two parts:the selection and determination of edge degradation models and the construction of correlations between performance parameters.At the same time,based on the Kalman filtering method,a residual life prediction method for binary performance degradation is proposed,and verified using chip test data,which is applied to chip reliability modeling and residual life prediction based on two solder joint degradation.

关 键 词：COPULA函数 二元性能退化联合建模 可靠性建模 剩余寿命预测 

分 类 号：V416.5[航空宇航科学与技术—航空宇航推进理论与工程]