金属化膜电容器Gamma过程性能退化建模的自助法  

Bootstrap Method of Metallized Film Capacitor Gamma Process Degradation Modeling

作　　者：刘学敏[1] 厉海涛[1] 张威[2] 周经伦[1] 周忠宝[3] 

机构地区：[1]国防科技大学信息系统与管理学院,长沙410073 [2]徐州空军学院军交运输系,徐州221000 [3]湖南大学工商管理学院,长沙410082

出　　处：《高电压技术》2011年第1期198-202,共5页High Voltage Engineering

基　　金：国家自然科学基金(70901024);国家高技术研究发展计划(863计划)(2008AA04Z409);国防科技大学优秀博士生创新基金(0905001)~~

摘　　要：针对安全关键系统中长寿命高可靠性产品失效数据稀缺的问题,研究了一种部件寿命预测的自助法。首先建立了基于Gamma过程的性能退化模型。而后,考虑到各个体间的制造及运行环境等随机差异,应用自助法对个体退化数据再抽样求得了个体形状参数α和尺度参数β;在此基础上,对个体α、β再自助抽样求得总体形状参数和尺度参数;之后,运用Gamma过程寿命预测方法得到了总体任务可靠度寿命的频率分布图。最后,以金属化膜电容器为例进行了应用分析。结果表明,该方法有益于全面掌握产品的寿命特征,提高系统可靠性安全性。Because the failure data of long lifetime and high reliability item are rare in the safety-critical system,we put forward a bootstrap for lifetime forecasting.We firstly proposed a performance degradation model based on Gamma process.Then,taking account of difference between manufacture and run condition among individuals,we adopted bootstrap on degradation data to get shape parameter α and scale parameter β.Thereby,bootstrap on α,β for shape parameter and scale parameter of population was adopted.According to this model,lifetime probability distribution of population task reliability can be gained.Lifetime forecasting of metallized film capacitor is analyzed,and the results show that the model is beneficial to finding lifetime characteristic of items and improving system reliability and safety.

关 键 词：退化建模 Gamma过程 参数估计 自助法 寿命分布 金属化膜电容器 

分 类 号：TM53[电气工程—电器]