基于状态检测的关键设备系统预防性维修与备件管理策略联合优化

Improving reliability and availability, and reducing life cycle costs are the objectives of critical equipment management. The inspection-based preventive maintenance (PM) policy to identify the system’s condition is an important activity to meet the above objectives. There is a closely linking between maintenance activities and spare parts management, which leads to the inevitable trend of a joint optimization of PM and spare parts management based on the condition detection. Many studies have been presented on this topic; however, there are some problems need to be further explored. Thus, this project shall study the following aspects step by step: 1) a multi-level and aperiodic inspection policy with imperfect maintenance is widely used to single-unit systems, so the threshold of emergency order is merged to establish the integrated optimization model of PM and the spare ordering policy; 2) Considering economic dependence and repair sequence, the joint optimization of block-based inspection and spare parts inventory policies considering opportunistic maintenance and repair priority is presented for multi-unit systems. Depending on the state of unit on the system’s state, the single-type unit system or the two-type unit system are studied, respectively; 3) the controllable variable is introduced to estimate the remaining lifetime, so the dynamic joint optimization model is developed to update the static decisions and controllable variable simultaneously based on the static joint decisions obtained by the above two studies. The research will provide new theory for the integrated optimization of preventive maintenance and spare parts policies, and is beneficial for enterprise to make scientific decision, improving the equipment reliability and economic benefits.

提高可靠性、可用度、降低全生命周期费用是关键设备系统管理的目标，通过检测识别系统状态进行预防性维修已成为实现这些目标的重要方式。维修活动与备件管理关系紧密，研究基于状态检测的预防性维修与备件管理策略的联合优化是必然趋势。目前该领域研究已取得一定成果，但仍有一些问题需要解决，因此，本项目从三方面进行深入研究：1)单部件系统常采用多水平、非周期检测并进行不完美维修，对此引入紧急订货阈值进行预防性维修与备件订购策略整合优化；2)考虑到经济相关性、部件的维修顺序，提出基于机会维修和维修优先权的多部件系统成组检测与备件库存策略联合优化，并根据部件状态对系统状态的影响研究同类部件、两类部件系统；3)引入可控变量预测系统剩余寿命，结合1)和2)策略联合优化的结果，建立动态优化模型更新原策略并优化可控变量。本项目将为预防性维修与备件策略整合优化提供新理论，帮助企业进行科学决策以提高设备可靠性、经济效益。

关键设备系统广泛存在于日常生活和生产企业中，如运输系统、先进制造系统、高端医疗设备、航空航天系统等，这类系统具有高价值、运行周期长等特征且通常是企业的核心设备，直接影响关键任务的执行，然而由于受到外界负荷、磨损、腐蚀等因素影响，其性能会逐渐退化直至发生故障。预防性维修策略的应用将有助于提高设备可靠性，然而维修势必导致产生备件需求，反之备件制约维修活动的顺利开展。所以，本项目研究了关键设备系统的预防性维修策略与备件管理策略的联合决策问题，主要研究内容包括以下三部分：. 第一部分研究考虑不完美维修的预防性维修与备件订购联合策略优化问题。将所提出的基于两阶段延迟时间的视情维修与备件订购联合策略与未考虑备件订购的维修策略、役龄更换的维修与备件订购联合策略进行对比，所提出的策略能够降低成本，降低设备发生故障的风险。进一步，扩展到三阶段退化过程的定期检测/非周期检测与备件订购联合策略优化问题上，并在算例分析中验证了所提出的联合策略的有效性。. 第二部分是考虑紧急订货阈值的预防性维修与备件订购联合策略优化问题。通过对紧急订购备件情形添加紧急订货阈值约束，对提出的非检测周期的预防性维修与备件订购联合策略模型进行优化。选取炉衬系统作为研究对象进行案例分析，结合所提出的联合策略和收集到的炉衬寿命数据，确定最优决策，结果表明可适当延长检测周期避免频繁检测。. 第三部分是基于剩余寿命预测的维修与备件订购联合策略优化问题。考虑到设备健康状态能够实时获取进而采用维纳过程描述设备退化过程，利用实时的状态信息进行剩余寿命预测，进而将其作为备件订购的条件构建联合决策模型，优化预防性维修阈值和订货阈值。最后，以飞机发动机的案例展开研究，对所提出的策略进行验证分析。. 本项目为预防性维修与备件管理策略联合决策提供新的理论支持，并进一步用于指导企业的设备运营管理和备件的采购管理，具有重要的科学意义和价值。