【作者】 高萍；

【导师】 吴甦； 贾希胜；

【作者基本信息】 清华大学 ， 管理科学与工程， 2008， 博士

【摘要】 设备的日益复杂与快速更新给维修决策带来了新的挑战。本文针对复杂设备维修决策过程中确定维修方式、定时维修周期和基于状态的维修(Condition based Maintenance, CBM)时机三个环节,建立和完善了相关数学模型,改进或提出了有关决策方法,有效解决了维修方式不当、计划性维修中费用高而可用度低以及CBM中的维修时机不准确等问题。在总结分析复杂设备特点及其维修方式的基础上,从可靠性、维修性、监测性和经济性角度分析了影响维修决策的因素,建立了复杂设备各重要功能单元重要度评估指标体系,为维修方式决策提供了可靠依据。对于组成部件具有故障相关性的复杂设备,建立了确定其可靠性寿命分布的威布尔非致命冲击模型,并给出了该模型参数的评估方法。针对复杂设备故障及维修类型多样的特点,建立了可用度约束下维修费用最小的计划维修周期最优模型,该模型不仅适用范围广泛,而且在确保可用度满意的同时使维修费用最小。针对复杂设备运行状态因环境的不同而不同的特点,提出了基于威布尔比例风险模型(Weibull Proportional Hazards Models, WPHM)的CBM维修时机决策模型,给出了设备在线监测和离线周期监测可用度最大时维修时机决策过程,该模型可以利用设备的状态监测历史数据确定CBM维修时机。基于本文提出的重要度评估指标体系,确定了铁路机车柴油机的重要功能单元(Functionally Significant Item, FSI)的维修方式。在分析了柴油机FSI相关故障的基础上,应用本文提出的威布尔非致命冲击模型确定了柴油机及相关故障单元的可靠性寿命分布,检验结果验证了该模型的准确性。应用本文提出的可用度约束下维修费用最小的计划维修周期最优模型,得到了铁路机车增压器和变速箱的最优维修周期,分析结果表明可以在保证可用度满意的同时产生较好的经济效益。应用基于WPHM的CBM维修时机决策模型和方法得到了油田输油泵和注水泵的维修时机决策图,给出了CBM维修时机决策策略。更多还原

【Abstract】 More and more new subjects about the maintenance decision research come forth because of the complexity of equipment. By improving the decision model and method of maintenance mode, scheduled maintenance cycle and condition based maintenance (CBM) threshold, the goal of this thesis is to solve the maintenance problems of complex equipment.The characters and the categories of maintanence mode of complex equipment have been analyzed firstly. And the criticality evaluation index system has been constructed including reliability, maintainability, inspection and economy factors. Then a bivariate Weibull distribution model, non-fatal shock mixture model, has been established to determine the reliability life distribution of the complex system with dependent components. And a synthetical approach has been suggested to estimate the parameters of the distribution model. Furthermore, a multi-objective maintenance cycle optimization model has been presented to find the sechduled maintenance cycle to achieve the satisfying availability with minimal maintenance cost. The applicability of this model is better than the existing research because the multiformity of the maintence and failure type has been taken into account. Finally, a kind of CBM maintenance threshold decision model has been described which is based on the Weibull proportional hazards models (WPHM). The model may reflect the relation between the operation state and the inspection index according to the exsting inspection data.The functionally significant item (FSI) and the maintenance mode of the railway locomotive engine have been obtained firstly. Then the reliability life distribution of the engine has been determined by employing the Weibull non-fatal shock mixture model. And the feasibility and veracity of the model have been proved by the analysis of the goodness-of-fit. Finally, the scheduled maintenance cycle of the supercharger and transmission of the locomotive have been calculated by multitarget sechduled maintenance cycle optimization model. And the CBM decision policy of the injection pump and water injecting pump by employing the decision model and process presented in this paper. All application study of these two preventive maintenance modes has shown the validity and applicability of the decision mode and method.更多还原