摘要
针对多状态可修系统提出了一种预防性维修策略。假定系统存在多种状态,当系统每次的工作状态处于较差工作状态时,对系统进行预防维修,该预防维修视为"非完好维修",当预防维修次数达到N时,无论在下一个工作周期系统是否仍处于较差工作状态,系统将不再进行此预防维修。运用极大似然估计来估算系统元件的状态概率,借用通用生成函数的方法来获取系统各状态的概率值,将系统的工作时间描述为随机递减的几何过程,预防维修的时间描述为随机递增的几何过程,建立最优预防维修策略的数学模型,在确保系统单位时间内期望效益最大的条件下,给出最佳的预防维修策略值N*。
In this paper, a kind of preventive maintenance strategy is proposed to study the multi-state repairable system. Assuming that the system has a variety of states and the system stay in a poor working state each time, the preventive mainte-nance which is regarded as “imperfect maintenance” is carried out; when the number of the preventive maintenance reachesthe time, no matter whether the system is still in the poor working state in the next cycle, the preventive maintenance will notbe carried out. The state probability of system components is estimated by using the maximum likelihood estimation, then, theprobability value of the system for each state is obtained from method of general generating function. The operating time of thesystem will be described as stochastically decreasing geometric process, while the time of the preventive maintenance is described as stochastically increasing geometric process, thus a mathematical model of the optimal preventive maintenance strategy is established. Under the condition of maximum expected benefits for the system, a value of the optimal preventive maintenance strategy is obtained.
出处
《机械研究与应用》
2016年第5期56-60,共5页
Mechanical Research & Application
基金
中央高校基本科研业务费专题项目:高速铁路装备安全监测和运营维护中的大数据统计分析研究(编号:2682014Z29)
四川省统计科学研究计划项目:相关性状态退化数据的统计分析及其高速机车可靠性工程化运用(编号:2016SC50)
关键词
多状态系统
极大似然估计
通用生成函数
几何过程
multi-state system
maximum likelihood estimation
general generating function
geometric process
