Abstract: The geometric maintenance process and its extended model are main tools to study the standby system that can better describe the distribution of continuous faults and its maintenance time.To study the optimization problem of quasi-period maintenance window that was dynamically and randomly distributed after periodic Preventive Maintenance (PM) affected by some factors such as production cycle and maintenance capacity,aiming at the two-unit diverse cold-standby system,a quasi-periodic PM policy optimization model with a maintenance window was constructed by assuming that the failure process of main unit and spare unit were subject to the extended geometric process with recovery factor and exponential distribution respectively.The optimal quasi-periodic PM policy was obtained by maximizing the revenue rate of system.The influence of system parameters on the optimized results was discussed through sensitivity analysis.

Key words: extended geometric process, cold-standby system, quasi-periodic, preventive maintenance, two-unit system, maintenance window

摘要： 由于几何维修过程及其拓展模型能够较好地描述故障及其维修时间的随机分布,成为贮备系统可靠性建模的主要工具之一。同时,周期预防性维修活动受生产周期、维修力量等因素影响后呈动态随机事件,表现为拟周期形式分布于其维修窗口。为研究服从拓展几何维修过程的贮备系统的拟周期预防性维修策略优化问题,以互异的两单元冷贮备系统为研究对象,在假设主单元失效过程服从具有修复因子的拓展几何维修过程,贮备单元失效过程服从指数分布的前提下,构建了带维修窗口的拟周期预防性维修策略优化模型。通过最大化系统收益率,获得了最优的拟周期预防性维修策略,通过敏感性分析讨论了系统参数对优化结果的影响。

关键词: 拓展几何过程, 冷贮备, 拟周期, 预防性维修, 两单元系统, 维修窗口