摘要
随着对城市轨道交通日常客流出行规律的不断挖掘,运输组织创新是解决客流与运力有效匹配问题、实现系统能耗节约及社会经济效益最大化的关键手段,而重联编组运营模式可有效提高客流与运力的匹配度。通过分析重联编组与固定编组条件下车底运用问题的差异性,构建基于“影子列车”的重联编组车次接续法,以车底与车次接续、车底一致性和重联编组作业等为约束条件,以车次接续总成本最小和车底使用时间标准差最小为目标函数,构建重联编组条件下城市轨道交通车底运用方案优化模型。通过引入非线性惯性权重更新方法和动态学习因子,设计多目标混沌粒子群优化(Multi-objective Chaos Particle Swarm Optimization,MOCPSO)算法。以某城市轨道交通线路的102个车次为例验证模型的有效性,并对车次接续时间上限、车底重联解编作业和车底存放情况进行讨论分析。研究结果表明:MOCPSO算法通过引入Logistic混沌优化策略可有效跳出局部最优;车次接续时间上限越大,需要投入的车底数量越多,不宜使车次接续时间过长;在车底运用过程中应尽可能地减少联挂解编作业的次数。该方法可为决策者提供一系列不同运营投入和车底运用均衡性下的车底运用Pareto非劣方案,有助于协调线路运能利用,同时降低了轨道交通的能耗。
With the continuous exploration of the daily passenger flow travel rules of urban rail transit,innovation in transport organization is the key approach to solve the problem of effective matching between passenger flow and transport capacity,and to achieve the system energy saving and maximize the social and economic benefits.The operation mode of reconnection marshalling can effectively improve the matching degree between passenger flow and transport capacity.By analyzing the differences in rolling stock scheduling problems between reconnection marshalling and fixed marshalling,the succession method of the reconnection train based on the“shadow train”was constructed.Taking the rolling stock and train number succession,rolling stock consistency and reconnection marshalling operation as the constraints,and taking the minimum total cost of train number succession and the minimum standard deviation of rolling stock use time as the objective functions,the optimization model of rolling stock scheduling scheme of urban rail transit under reconnection marshalling was constructed.By introducing a nonlinear inertia weight update method and dynamic learning factors,a multi-objective chaos particle swarm optimization(MOCPSO)algorithm was designed.An example of 102 train numbers on an urban rail transit line was taken to verify the effectiveness of the model.The upper limit of train number succession time,the reconnection and unmarshalling operation of rolling stock and the rolling stock storage were discussed and analyzed.The results show that the MOCPSO algorithm can effectively jump out of the local optima by introducing the Logistic chaos optimization strategy.The larger the upper limit of the train number succession time,the more the rolling stock input quantity.It is not appropriate to make the train number succession time too long.In the process of rolling stock scheduling,the number of reconnection and unmarshalling operations should be reduced as much as possible.This method can provide decision-makers with a ser
作者
朱昌锋
贾锦秀
马斌
孙元广
王傑
成琳娜
ZHU Changfeng;JIA Jinxiu;MA Bin;SUN Yuanguang;WANG Jie;CHENG Linna(School of Traffic and Transportation,Lanzhou Jiaotong University,Lanzhou 730070,China)
出处
《铁道科学与工程学报》
EI
CAS
CSCD
北大核心
2024年第7期2626-2636,共11页
Journal of Railway Science and Engineering
基金
国家自然科学基金资助项目(72161024)
甘肃省教育厅“双一流”科研重点项目(GSSYLXM-04)。
关键词
城市交通
重联编组
车底运用
多目标优化
混沌粒子群算法
urban traffic
reconnection marshalling
rolling stock scheduling
multi-objective optimization
chaos particle swarm algorithm
