摘要
以单台航空发动机系统在定时维修和失效更换相结合的维修策略下的备件需求规律为研究对象,假设航空发动机中的部件寿命服从一般分布,利用Phase-Type(PH)分布的稠密性将一般分布表示为PH分布形式;然后通过仿真获取整个发动机系统的备件更换时间序列;在仿真结果基础上,利用EM算法将单台发动机的备件需求点过程拟合为马尔可夫到达过程(Markovian arrival proccss,MAP),并根据MAP性质确定了多台相同发动机组成的系统的备件需求规律;最后通过一个示例说明了本文方法的有效性.文中通过仿真获取样本数据,再采用拟合算法确定需求规律的研究思路具有很好的通用性,对他人的研究工作有一定借鉴作用.
This research investigates the spare parts demand law of an aviation engine under certain maintenance policy,which is combined with period maintenance policy and failure replacement policy. The life time of unit is assumed to follow general distribution that is represented by Phase-Type（PH） distribution according to the dense properties of PH distribution;and a simulation algorithm is designed to obtain the unit replacement time series.Then using the EM algorithm,the spare parts demand point process of single engine is fitted by Markovian arrival process（MAP）.Based on this result and properties of MAP,the spare parts demand law of the system that consists of several same engines is obtained.Finally, a practical example is put forward to illustrate the implementation of the method.In this paper,the idea that sample data come from simulation and demand law is ascertained by fitting algorithm is more general so that it can be referred to other field researches.
出处
《系统工程理论与实践》
EI
CSSCI
CSCD
北大核心
2010年第5期769-777,共9页
Systems Engineering-Theory & Practice
基金
国家自然科学基金(70501031)
