存在工作历程产品的加速寿命试验统计分析  

Statistical Analysis of Accelerated Life Testing for Products with Operational Period

作　　者：汪亚顺[1] 张春华 陈循[1] 谭源源[1] 

机构地区：[1]国防科技大学机电工程与自动化学院,湖南长沙410073

出　　处：《国防科技大学学报》2008年第1期94-98,共5页Journal of National University of Defense Technology

基　　金：国家部委资助项目(203020102)

摘　　要：加速寿命试验(ALT)可在较短时间内获得产品的寿命及可靠性信息。利用ALT对产品的剩余寿命进行评估时,常常将已工作过的产品进行抽样并投入试验,在这一类样本的ALT数据统计分析时如何处理初始工作时间,成为ALT应用中的一个重要问题。工程实际中评估此类样本的剩余寿命时常常忽略初始工作时间,将其视为"用后如新"或"无记忆性"产品。但此假设必须以产品寿命服从指数分布为前提,而大部分机电产品的寿命服从Weibull分布,因而该方法在应用时必然会产生较大误差。针对这一问题提出了一种新的基于时间折算的ALT数据统计分析方法,并利用Monte Carlo仿真对其估计特性进行对比研究,结果表明此方法能有效评估存在初始工作历程产品的剩余寿命,估计精度优于原方法。Accelerated life testing （ALT） of a product or material can obtain information quickly on its life or reliability information. Produets or materials with operational period are often sampled and tested to evaluate their residual life. However, it is a main challenge how to deal with the operational period in statistical analysis for ALT application. The products are often treated as ＂new after operation＂ or ＂memoryless＂ and their operational period is ignored in residual-life evaluation on the assumption that their failure times follow exponential distribution. Since failure times of many mechatronic products follow Weibull distribution, the estimated precision of residual life of the products is low in the way mentioned above. This paper presents a novel approach based on equivalent time to statistically analyze ALT data of products or material with operational period. Estimated precision of the approach was studied by Monte Carlo simulation method. Results show that the approach can effectively estimate the residual life of the products and is more precise than that adopted previously.

关 键 词：加速寿命试验 MONTE Carlo仿真 工作历程 可靠性 

分 类 号：TB114.3[理学—概率论与数理统计]