This research project focuses on the basic scientific problem of failure prognostics for high-power white LEDs (light emitting diodes) with the characteristics of “multiple failure modes” and “non-stationary degradation process”. A novel multiple failure prognostics method is proposed by combining physics-of-failure models and performance-degradation data. Firstly, a unified and generalized physics-of-failure model is established that can be used to describe the interaction of multiple failure modes and mechanisms. The LEDs’ photometric, colorimetric and electricity properties with “sporadic jumping” and “non-stationary depreciation” characteristics are studied. The non-stationary stochastic process-based degradation models are respectively established for photoelectric parameters, chromaticity coordinates and spectral power distribution (SPD). The Copula function theory is utilized to construct the joint failure probability distribution function (PDF) for multiple failure modes. Then, more accurate life prediction and reliability estimation can be realized through the comprehensive application of the above models and methods. Finally, some series of typical LED lighting products are selected as test samples. A combination of the multi-physics simulation and accelerated degradation test is applied for verification purposes. Through this research project, the necessary and important basis of theoretical and scientific is laid and provided for complex electronics (LED system as the representative) prognostics. On the other hand, the necessary guidance and technical support are also provided for fast, accurate, and low cost life (including RUL) prediction and reliability qualification of new type of microelectronic devices (LED as the representative).
本项目研究围绕“具有多模式、非平稳退化特征的大功率白光LED寿命预测”基础问题展开,提出多失效模式条件下失效物理模型与性能退化数据相融合的寿命预测新方法。首先构建一个可以表征LED光输出退化、颜色偏移等多失效模式及机理交互作用的失效物理统一模型。针对LED光电色性能参数带跳跃、非平稳退化特征,分别构建描述光电特性、色度以及光谱功率分布等特征量退化的非平稳随机过程(如Lévy过程)模型。利用Copula函数理论构建多失效模式联合概率分布模型,进而实现准确的LED寿命预测和可靠性评估等。选取典型LED光源产品为案例,采取仿真分析与实物试验相结合的方式进行应用验证。通过本项目研究,一方面为LED系统为代表的复杂电子产品寿命预测理论和方法的进一步完善奠定理论基础和提供科学依据;另一方面也为LED为代表的新型微电子器件进行快速、准确、低成本的寿命预测、可靠性鉴定及合格认证等提供方法指导和技术支撑。
LEDs具有“高能效”、“长寿命”、“绿色环保”等特点,已开始在包括国防军事、航空航天、城市建设、通讯网络、汽车工业、医疗设备、消费电子等诸多工程系统和产品中得到越来越广泛的应用。本项目研究围绕“具有多模式、非平稳退化特征的大功率白光LED寿命预测”基础问题展开,提出了可以同时考虑LED多失效模式条件下失效物理及性能非平稳随机退化过程进行寿命预测的新方法,开展了LED光输出退化及颜色偏移失效物理模型、LED光电色性能以及光谱功率退化预测方法、多失效模式融合的寿命预测方法等方面的研究。研究成果弥补了当前LED 寿命预测及可靠性评估的理论认知不足和技术方法缺陷。本项目的研究工作,一方面为LED系统为代表的复杂电子产品寿命预测理论和方法的进一步完善奠定了必要的理论基础并提供了重要的科学依据;另一方面,研究提出的方法为以LED为代表的新型微电子器件进行快速、准确、低成本的寿命预测、可靠性鉴定及合格认证等提供了必要的方法指导和技术支撑。
{{i.achievement_title}}
数据更新时间:2023-05-31
监管的非对称性、盈余管理模式选择与证监会执法效率?
基于LASSO-SVMR模型城市生活需水量的预测
低轨卫星通信信道分配策略
宁南山区植被恢复模式对土壤主要酶活性、微生物多样性及土壤养分的影响
基于多模态信息特征融合的犯罪预测算法研究
LncRNA-Dleu2协同miR15/16调控CD4+T细胞亚群分化参与多发性硬化发病机制的研究
非平稳非高斯随机过程的模拟与预测研究
基于失效物理的小子样长寿命产品寿命预测理论
非平稳非高斯随机载荷作用下结构振动疲劳分析与寿命预测研究
大功率白光LED封装的非牛顿流体微涂覆过程建模与控制