摘要
针对寿命服从Weibull分布、维修时间服从指数分布的可修系统,首先根据"累积失效概率相等"原则,提出一种改进的指数确定方法,能够将Weibull分布等效成为指数分布,然后利用马尔可夫过程理论近似计算备件需求量。根据系统运行过程,设计了一种基于ExtendSim软件的仿真算法用来检验给定备件数量下的保障度,并以马尔可夫系统为例验证了该仿真算法的正确性。在此基础上,通过正交试验分析与设计,对两种指数算法的结果进行比较,结果表明改进算法能够有效降低原算法的保守性。
Aiming at the repairable system which follows the Weibull distribution of unit＇s life as well as the exponential distribution of the repair time,an improved exponential distribution determination method is proposed to convert the Weibull distribution to exponential distribution according to the principle of ＂cumulated failure probability corresponding＂.Then,the spare demand can be calculated approximately by the Markov process theory.According to the operating process,a simulation algorithm is designed to calculate the supporting probability with given spares based on extendSim software.The correctness of this algorithm is verified with the example of the Markov system.On this condition,the paper compares and verifies the results of the two Exponential methods through the orthogonal experiment design method.The result shows that the improved algorithm can depress the conservativeness of the primary method.
出处
《系统工程与电子技术》
EI
CSCD
北大核心
2012年第5期966-972,共7页
Systems Engineering and Electronics
基金
总装预研项目(51327020105
51304010206)
海军工程大学博士生创新基金(HGBSJJ2011009)资助课题
关键词
可修系统
备件需求
WEIBULL分布
指数算法
正交试验设计
repairable system
spare demand
Weibull distribution
exponential method
orthogonal experiment design
