摘要
针对高速铁路网络化运营背景下带维修约束的多车种、多动车所的动车组运用优化问题,提出一个新颖的2阶段优化分解算法。第1阶段基于多商品网络流理论(MCFP),建立不考虑维修约束的动车组交路计划混合整数线性规划模型(MILP),以获得多种动车组备选交路计划方案;第2阶段将获得的备选交路计划方案中的交路转换为单个动车组单元的可行路径集合,建立带维修约束的动车组单元指派模型来检验其维修可行性。结果表明:所提出的算法相比于现场人工方法节省4个8编组的动车组、减少38.4%的总空驶距离并降低3%的运营成本;通过更加灵活的编组方式进一步提高了动车组的运用效率。
The train-set scheduling optimization problem with multiple train-sets and multiple depots considering maintenance requirementswas investigated,and a novel two-stage decomposition algorithm was proposed.Firstly,a multi-commodity flow-based MILP optimization model for train-set schedulingwithout consideringmaintenance requirementswas established to obtain multiple candidates train-set schedules.Thenbased on the multiple candidate train-set schedules generated in the first stage,the second stage of the algorithm generated a set of feasible paths for each single train-set unit,where a path allocation subproblem with maintenance constraints was developed to check the maintenance feasibility.The results show that the proposed optimization approach can save foureight-car size train sets,reduce the total deadhead mileage by 38.4%and save 3%operating cost compared to the manual schedule,and that the operation efficiency of train-setcan be further improvedthrough a more flexible way of composition.
作者
钟庆伟
张永祥
文超
彭其渊
王锰锰
ZHONG Qingwei;ZHANG Yongxiang;WEN Chao;PENG Qiyuan;WANG Mengmeng(School of Transportation&Logistics,Southwest Jiaotong University,Chengdu Sichuan 610031,China;National United Engineering Laboratory of Integrated and Intelligent Transportation,Chengdu Sichuan 610031,China;Railway Research Center,University of Waterloo,Waterloo,N2L3G1,Canada;Zhengzhou Depot,China Railway Zhengzhou Group Co.,Ltd.,Zhengzhou Henan 450000,China)
出处
《中国安全科学学报》
CAS
CSCD
北大核心
2019年第S02期10-17,共8页
China Safety Science Journal
基金
国家自然科学基金资助(71871188,61503311)
四川省科技厅应用基础研究项目(2018JY0567).
作者简介
钟庆伟(1991—),男,四川什邡人,博士研究生,主要研究方向为高速铁路动车组运用管理。E-mail:qingweizhong@my.swjtu.edu.cn;文超,副教授;彭其渊,教授
