基于车流径路选择偏好的铁路车流运行径路动态预测方法研究  

Research on dynamic prediction method of railway wagon flow routing based on wagon flow route selection preference

作　　者：张红斌 李军 金福才 ZHANG Hongbin;LI Jun;JIN Fucai(Beijing Jingwei Information Technology Co.Ltd.,Beijing 100081,China;Dispatching Center,CHINA RAILWAY,Beijing 100844,China;Institute of Computing Technologies,China Academy of Railway Sciences Corporation Limited,Beijing 100081,China)

机构地区：[1]北京经纬信息技术有限公司,北京100081 [2]中国国家铁路集团有限公司调度指挥中心,北京100844 [3]中国铁道科学研究院集团有限公司电子计算技术研究所,北京100081

出　　处：《铁路计算机应用》2021年第4期31-35,共5页Railway Computer Application

基　　金：国家自然科学基金项目(U1934216);中国国家铁路集团有限公司科技研究开发计划课题(K2019X021);中国铁道科学研究院集团有限公司科研项目(2019YJ109,2020YJ012)。

摘　　要：准确的车流径路预测是编制高质量日班计划的前提条件。利用运输信息集成平台提供的海量车流运行轨迹历史数据,提出一种简单易行的铁路车流运行径路动态预测方法;通过对长期和近短期车流运行轨迹历史数据的统计,生成车流径路选择偏好参数表,可反映路网中多种复杂因素对于车流径路选择的影响;基于车流径路选择偏好参数,结合车辆实时位置跟踪信息,动态预测车流运行径路;搭建大数据应用环境,完成车流径路预测程序的开发与部署,采用命中率和准确率2项指标评价该方法的应用效果,结果表明:预测结果准确率较高,具有较为满意的实用价值。Accurate prediction of wagon flow routing is a prerequisite for making a high quality daily and shift freight traffic plan.By utilizing massive historical data of wagon flow from the transportation information integration platform,a simple dynamic prediction method of railway wagon routing is proposed,which is based on long-term and short-term wagon flow route selection preference parameters that are derived from historical data statistics of freight wagon path and can reflect the effects of many factors in the railway on wagon route selection.By combining wagon route selection preference parameters and real-time position tracking data of wagons,wagon flow routing can be predicted.Furthermore,an environment of big date application is set up and the program of wagon flow routing prediction is developed and deployed.In the end,two index of prediction hit rate and prediction accuracy are defined and calculated to evaluate the effects of the prediction and the calculations indicate that the prediction results have high accuracy and satisfactory practical value.

关 键 词：车流径路预测 车流径路选择偏好 运输集成平台 大数据应用 预测评价 

分 类 号：U292[交通运输工程—交通运输规划与管理] TP39[交通运输工程—道路与铁道工程]