基于自适应遗传算法的LED器件多应力条件下寿命快速评估系统模型  

Modeling of the rapid lifetime assessment system with multiple stresses for LED devices using adaptive genetic algorithm

作　　者：黄苏丹[1,2] 刘淮源 曹广忠[2] 周霖[1] 吴海林[3] 胡益民[1] 敬刚[1] 曹鸣皋 肖文鹏[1] 刘岩[1] 

机构地区：[1]深圳清华大学研究院广东省光机电一体化重点实验室,深圳清华大学研究院深圳市光机电一体化重点实验室,深圳518060 [2]深圳大学深圳电磁控制重点实验室,深圳518060 [3]深圳大学材料学院,深圳518060

出　　处：《中国科学：技术科学》2016年第9期940-949,共10页Scientia Sinica(Technologica)

基　　金：国家重点基础研究发展计划(编号:2011CB013106);国家自然科学基金(批准号:51275312);国家科技支撑计划(编号:2014BAH23F04);深圳市科技计划(编号:JSGG20141015153303491;KC2014JSJS025A)资助项目

摘　　要：发光二极管(light-emittingdiode,LED)作为正在兴起的固态冷光源具有效率高、寿命长、体积小、响应快、抗振、驱动电压低、绿色环保等特点,在照明光源、仪表指示等领域显示出较大的发展潜力.针对LED器件寿命在实际使用环境中应力影响因素复杂、加速寿命试验模型考虑的应力因素不全面以及LED器件寿命预测模型准确性低等问题,本文提出了一种LED器件多应力条件下寿命快速评估系统模型.该模型适用于各种工艺制造的不同结构的LED器件.以光效衰减30%为寿命依据,以环境温度、湿度和驱动电流为输入量、LED器件寿命为输出量,采用设计的自适应遗传算法确定被测器件的辨识参数.利用构建的加速寿命试验获得较高温度湿度电流应力下的少量样本数据,基于得到的样本数据辨识该模型的未知参数,获得了模型.在常温常湿常电流应力下,通过该模型进行了LED器件寿命预测.研究结果表明:在任意温度湿度电流应力下,提出的模型可以快速较精确地实现LED器件寿命预测;在常温常湿常电流应力下,基于提出的模型得到的LED器件寿命接近于LED行业平均技术水平的LED预期寿命;提出的模型应用于LED器件寿命可靠性测评具有实用性,并可继续拓展成为纳入更多应力的模型.The light-emitting diode（LED） is a promising candidate for the lighting,instrument reading,etc.,due to its features of high efficiency,long lifetime,small size,fast response,anti vibration,low-voltage operation,and environmental protection,which is an emerging cold light source.This paper proposes a novel model of the rapid lifetime assessment system with multiple stresses for LED devices,to solve the problems of the complex environmental stresses imposed on the lifetime of LED devices,the model of accelerated life testing without consideration of multiple stresses,and the low-precision lifetime assessment model.This model is suitable for all types of LED devices.Taking the operating time to 30%of lumen depreciation as the lifetime of LED devices,the environmental temperature and humidity,and input current are the inputs of this model,and the lifetime of the LED device is the output of this model.The parameters of this model are determinated by using the designed adaptive genetic algorithm.A several number of sample data is first obtained through accelerated life testing under the condition of high temperature,humidity,and current.The parameters of this model are then identified with the obtained sample data.This model is further built by applying the identified parameters.Additionally,the lifetime evaluation is carried out via this model under the condition of normal temperature,humidity,and current.Research results demonstrate that the proposed model can evaluate the lifetime of LED devices fast with a comparatively high accuracy;the evaluated lifetime of the proposed model is close to the realizable life expectancy of the LED industry under the condition of normal temperature,humidity,and current;the practicability of the proposed model for lifetime reliability evaluation of LED devices is verified;the proposed model can be further developed with much more stresses.

关 键 词：发光二极管 寿命快速评估 加速寿命试验 自适应遗传算法 参数辨识 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程] TN312.8[自动化与计算机技术—控制科学与工程]