电子产品性能退化试验与故障趋势预测研究

In order to improving reliability of electronic product, more efficient performance degradation techonology and fault trend prediction algorithm based on performance degradation analysis are researched in this project. Accelerated degradation testing (ADT) method and optimum design algorithm, including single stress and multi-stress, are discussed. By introducing the conception of performance reliability, performance degradation analysis method for electronic product with multi-degradation parameters in multi-stress testing environment is proposed. Fault Feature Extraction methods Based on wave-matching and wavelet analysis are researched, based on these, fault trend prediction model based on multi-stress performance analysis is found. A new electronic equipment fault trend prediction algorithm is put forward by introducing signal processing technology and principal component analysis (PCA). Based on these theory algorithms, a software system with the function ADT optimum design and fault trend prediction for electronic product is come true. This project will provide a new technology approaches for electronic equipment fault prediction, and will solve the key problem of performance degradation analysis and fault prediction model for multiple parameters in multiple testing environment. By these researches, fault prediction theory for electronic product will be more scientific and advisable;and will provide theory support for electronic equipment preventive maintenance. Research contents in this project have important theory value and application prospect. besides, We hope that these research findings will enrich and develop the fault prediction theory and reliability theory。

本项目以提高电子产品可靠性为目的，研究更高效率的性能退化试验技术和基于性能退化分析的故障趋势预测算法。包括：研究单应力及综合应力加速退化试验方法，探讨加速试验优化设计算法；引入性能可靠性概念，研究电子产品综合应力条件下多参数的性能退化分析方法；研究基于波形匹配、小波分析等理论的故障特征提取方法，建立基于综合应力性能退化分析的故障趋势预测模型；利用相像系数法及主成分分析法等，探讨多参数条件下的电子装备故障趋势预测新算法。在解决上述理论及算法的基础上，实现一个有实用价值的电子产品加速退化试验设计与故障趋势预测的软件系统。本研究有望为电子产品故障预测提出一个新的技术途径，解决综合应力下多参数性能退化分析和故障预测中的难题，使电子产品的故障预测更加科学合理，为电子设备预防性维修提供理论支持。具有重要的理论价值和应用前景。希望申请的研究工作能够促进电子产品故障预测理论乃至可靠性理论的完善和发展。

课题基于广义故障预测的概念，以电子产品为对象，主要研究了加速退化试验技术和基于产品性能退化的故障趋势预测、可靠性预测以及寿命预测的理论与方法。并且，在分析产品故障规律的基础上拓展了研究内容，进一步探讨了提高产品可靠性的数字电路故障自检测和自修复方法。取得的主要研究成果有：①提出了基于MC仿真的加速退化试验优化设计技术，为电子产品加速试验提供了高效的新方法；创新了电源涌浪与随机冲击的电应力加速试验方法，建立了部件级的多应力加速试验系统，为产品可靠性分析与评价提供了依据。②研究了融合试验数据和外场使用数据的使用寿命预测方法，建立了多性能退化参数模型，提出了时变应力条件下的预测方法，为提高预测的有效性和合理性建立了实用的方法体系。③通过基于波形匹配和小波分析理论的故障特征提取和故障分析方法研究，建立了数字电路自检测、自修复的理论基础。