动车组单部件两级非完美维护策略及优化研究

对动车组部件的维修保养是保证车辆在运营期间具有一定可靠度的重要措施,其维护周期的确定直接影响铁路运输安全.基于我国动车组的多级维护制度,提出部件的多等级非完美维护策略.选取动车组四级修时更换的部件为研究对象,建立了一种基于里程故障威布尔分布的两级非完美维护模型.以我国现行的动车组维护计划为框架,提出了动车组部件的两种维护方案,将维修时间和维修成本作为优化目标,应用多目标遗传算法对研究区间内的维修周期和所需的维修等级进行决策,并与单一非完美维护模型的优化结果进行比较.研究结果表明,两级非完美维护策略能以更低的维护成本得到满足可靠度要求的维护计划,且在不严格匹配整车维修等级的部件维护方案下得到的维护计划能够在较少的维护成本和维护次数下使部件处于较高的可靠度,更具经济性和安全性. 

Bi-level Imperfect Maintenance Strategy and Optimization for Single Component of Electric Multiple Unit

The maintenance of EMU (electric multiple unit) components is an important measure to ensure the operation reliability of vehicles. In addition, the determination of maintenance interval has direct influence on the safety of rail transport. Therefore, based on the multi-level maintenance rule of EMU in China, a multi-level imperfect maintenance policy was proposed. A component replaced in the 4-class repair of EMU was selected as an objective, and a bi-level imperfect maintenance model was established on the basis of the Weibull distribution of mileage fault. By utilizing the current maintenance rule of EMU as a template, two maintenance plans for the component were structured with maintenance cost and maintenance time as the optimal goals. The maintenance intervals and maintenance levels for each section of interest were optimized by using multi-objective genetic algorithm. Comparison with the optimization results of single maintenance mode shows that maintenance schedules based on the bi-level imperfect maintenance model can meet the reliability requirement at low cost. And maintenance schedules derived from case which does not exactly match vehicle repair levels can get higher component reliability with less maintenance costs and low requency.