机械工程学报
機械工程學報
궤계공정학보
CHINESE JOURNAL OF MECHANICAL ENGINEERING
2014年
16期
33-40
,共8页
唐圣金%郭晓松%周召发%司小胜%周志杰
唐聖金%郭曉鬆%週召髮%司小勝%週誌傑
당골금%곽효송%주소발%사소성%주지걸
加速退化模型%维纳过程%剩余寿命估计%步进应力%极大似然估计
加速退化模型%維納過程%剩餘壽命估計%步進應力%極大似然估計
가속퇴화모형%유납과정%잉여수명고계%보진응력%겁대사연고계
accelerated degradation model%wiener process%remaining useful life estimation%step stress%maximum likelihood estimation
步进应力加速退化试验的建模与剩余寿命估计是产品视情维护和健康管理中的一个关键问题。针对传统加速退化建模方法没有考虑同类产品之间个体差异的问题,提出一种基于维纳过程的加速退化建模和剩余寿命估计方法。利用维纳过程对单个应力下的退化过程建模,并将维纳过程中的漂移系数看成随机变量以描述个体差异。依据该模型,可以得到产品的剩余寿命信息和可靠性信息。提出一种随机变量加速模型,该模型将传统加速模型中的参数看成随机变量以描述个体差异,传统加速模型是其特例。针对步进应力加速退化试验的测量数据,提出一种加速退化模型未知参数的两步极大似然估计方法。以激光器为例的仿真试验结果表明,所提方法的模型拟合性和剩余寿命估计精度都优于现有建模方法,具有一定的工程应用价值。
步進應力加速退化試驗的建模與剩餘壽命估計是產品視情維護和健康管理中的一箇關鍵問題。針對傳統加速退化建模方法沒有攷慮同類產品之間箇體差異的問題,提齣一種基于維納過程的加速退化建模和剩餘壽命估計方法。利用維納過程對單箇應力下的退化過程建模,併將維納過程中的漂移繫數看成隨機變量以描述箇體差異。依據該模型,可以得到產品的剩餘壽命信息和可靠性信息。提齣一種隨機變量加速模型,該模型將傳統加速模型中的參數看成隨機變量以描述箇體差異,傳統加速模型是其特例。針對步進應力加速退化試驗的測量數據,提齣一種加速退化模型未知參數的兩步極大似然估計方法。以激光器為例的倣真試驗結果錶明,所提方法的模型擬閤性和剩餘壽命估計精度都優于現有建模方法,具有一定的工程應用價值。
보진응력가속퇴화시험적건모여잉여수명고계시산품시정유호화건강관리중적일개관건문제。침대전통가속퇴화건모방법몰유고필동류산품지간개체차이적문제,제출일충기우유납과정적가속퇴화건모화잉여수명고계방법。이용유납과정대단개응력하적퇴화과정건모,병장유납과정중적표이계수간성수궤변량이묘술개체차이。의거해모형,가이득도산품적잉여수명신식화가고성신식。제출일충수궤변량가속모형,해모형장전통가속모형중적삼수간성수궤변량이묘술개체차이,전통가속모형시기특례。침대보진응력가속퇴화시험적측량수거,제출일충가속퇴화모형미지삼수적량보겁대사연고계방법。이격광기위례적방진시험결과표명,소제방법적모형의합성화잉여수명고계정도도우우현유건모방법,구유일정적공정응용개치。
Accelerated degradation process modeling and remaining useful life(RUL) estimation based on Accelerated degradation testing(ADT) are the essential parts of condition based maintenance and health management. Among the literature about the accelerated degradation process modeling, the current methods didn’t consider the unit-to-unit variation of the equipment with the same type. Therefore, we present an accelerated degradation process modeling and RUL estimation method based on Wiener process. The Wiener process is used to model the degradation process of a single stress, and the drift coefficient is considered as a random variable to describe the unit-to-unit variation of different equipment. Based on the presented model, the information of reliability and RUL can be obtained. The parameter of the traditional acceleration model is considered as a random variable to describe the unit-to-unit variation of different equipment, and the traditional method turns out to be its special case. Through the step stress ADT data, a two-step maximum likelihood estimation method is presented to estimate the unknown parameters of the presented model. The results of the simulation about the laser degradation show that the presented method performs better than the traditional method and thus can be potentially applied in practice.