【摘要】 由于缺乏有效的分析手段 ,非马尔可夫可修系统的优化配置非常困难 ,而仿真的方法往往工作量巨大。文中用排队网络模型给出n :k(m)交叉储备系统的通用仿真模型 ,在此基础上设计了求解系统在可靠度约束下最优配置问题的高效率的仿真算法 ,并给出若干优化配置问题的算例。文末分别用解析的方法和仿真的方法计算了若干马尔可夫型n :k(m)系统的备件最优储备量 ,二者符合得很好。更多还原

【Abstract】 Configuration optimization (determining storage quantity, number of repairers etc) of non Markovain repairable system is a hard work due to the lack of effective analytical means. On the other hand, simulation based approach, though feasible, often has the disadvantage of large computation work. In this paper, efficient simulation based method for evaluation optimum configuration of the general non Markovian n:k(m) cross strapping cold standby system under reliability restriction is presented. Queueing network model is applied to describe the dynamic lifetime process of n:k(m) system. Based on such model, algorithm which can greatly reduce simulation work is designed for finding the optimal configuration of n:k(m) system. Several examples on optimum configuration of non Markovian n:k(m) system are presented. Finally two Markovian n:k(m) systems are used to validate the simulation method. The optimum storage quantity of such system is evaluated with both simulation approach and analytical approach. The results show the power and applicability of our simulation method.更多还原