Ticket allocation and train stopping plans are important parts of railway transportation organization.At present,most of the ticket allocation plans are based on fixed train stopping plans,which limit the scope of tic...Ticket allocation and train stopping plans are important parts of railway transportation organization.At present,most of the ticket allocation plans are based on fixed train stopping plans,which limit the scope of ticket allocation.Trains can only serve passenger demands between stopping stations,leading to a loss of passenger demands at non-stopping stations,resulting in low seat occupancy rates and low revenue for railway enterprise.In order to better meet passenger demands,improve seat occupancy rates and increase the revenue of railway enterprises,this paper constructs a collaborative optimization model of ticket allocation and stopping plans based on stochastic demand and passenger choice behaviours in different time periods.Combined with CPLEX solver,the simulated annealing algorithm is designed to solve the problem.At the same time,new neighbourhood solution generation strategies of train stopping plans and ticket allocation plans under given stopping plans are designed.The experimental results show that in small-scale and large-scale experiments,the proposed method increases revenue by 0.14%and 9.09%,respectively,and effectively improves seat occupancy rates.展开更多
针对传统停留点识别方法对非标准化分类的POI(Points of Interest)数据和停驻时间依赖程度较高、可迁移性较差等弊端,提出了一种基于轨迹几何特征的语义停留点识别方法。通过充分利用轨迹自身所包含的有效信息,深入挖掘停留点附近子轨...针对传统停留点识别方法对非标准化分类的POI(Points of Interest)数据和停驻时间依赖程度较高、可迁移性较差等弊端,提出了一种基于轨迹几何特征的语义停留点识别方法。通过充分利用轨迹自身所包含的有效信息,深入挖掘停留点附近子轨迹的重叠规律并进行量化,同时引入标准化分类的临时停留相关的POI数据,对语义停留点和临时停留点进行进一步的区分。在此基础上,利用企业级的订单数据对语义停留点识别结果进行出行目的标注并进一步构建识别效果评价指标,有效地解决了停留点识别结果验证过程中的主观性和复杂性问题。最后,以重型货车的分支——集卡拖车为研究对象,并基于某集装箱运输公司TMS(Transportation Management System)系统提供的集卡拖车轨迹数据和订单数据,对长三角业务地区开展实证研究。研究结果表明:商家地址匹配准确率为96.5%,语义停留点识别结果具有真实语义的比率为85.5%。与现有研究相比,所提方法的准确性得到了显著提升,为多源场景的大规模推广应用提供了可行性基础。研究结果反映了港口城市集疏运货运活动的空间分布特征及区域联系,并揭示了不同类型重型货车出行活动的异质性,可为城市管理者制定更加全面合理的货运政策提供支撑,并为货运企业管理车队、拓展增值业务提供依据。展开更多
基金supported by the National Natural Science Foundation of China(Grant No.72471247)the General Project of the Hunan Provincial Natural Science Foundation of China(Grant No.2022JJ30057)the Systematic Major Research Project of the China Railway(Grant No.P2022×012).
文摘Ticket allocation and train stopping plans are important parts of railway transportation organization.At present,most of the ticket allocation plans are based on fixed train stopping plans,which limit the scope of ticket allocation.Trains can only serve passenger demands between stopping stations,leading to a loss of passenger demands at non-stopping stations,resulting in low seat occupancy rates and low revenue for railway enterprise.In order to better meet passenger demands,improve seat occupancy rates and increase the revenue of railway enterprises,this paper constructs a collaborative optimization model of ticket allocation and stopping plans based on stochastic demand and passenger choice behaviours in different time periods.Combined with CPLEX solver,the simulated annealing algorithm is designed to solve the problem.At the same time,new neighbourhood solution generation strategies of train stopping plans and ticket allocation plans under given stopping plans are designed.The experimental results show that in small-scale and large-scale experiments,the proposed method increases revenue by 0.14%and 9.09%,respectively,and effectively improves seat occupancy rates.
文摘针对传统停留点识别方法对非标准化分类的POI(Points of Interest)数据和停驻时间依赖程度较高、可迁移性较差等弊端,提出了一种基于轨迹几何特征的语义停留点识别方法。通过充分利用轨迹自身所包含的有效信息,深入挖掘停留点附近子轨迹的重叠规律并进行量化,同时引入标准化分类的临时停留相关的POI数据,对语义停留点和临时停留点进行进一步的区分。在此基础上,利用企业级的订单数据对语义停留点识别结果进行出行目的标注并进一步构建识别效果评价指标,有效地解决了停留点识别结果验证过程中的主观性和复杂性问题。最后,以重型货车的分支——集卡拖车为研究对象,并基于某集装箱运输公司TMS(Transportation Management System)系统提供的集卡拖车轨迹数据和订单数据,对长三角业务地区开展实证研究。研究结果表明:商家地址匹配准确率为96.5%,语义停留点识别结果具有真实语义的比率为85.5%。与现有研究相比,所提方法的准确性得到了显著提升,为多源场景的大规模推广应用提供了可行性基础。研究结果反映了港口城市集疏运货运活动的空间分布特征及区域联系,并揭示了不同类型重型货车出行活动的异质性,可为城市管理者制定更加全面合理的货运政策提供支撑,并为货运企业管理车队、拓展增值业务提供依据。
