中国安全科学学报 ›› 2019, Vol. 29 ›› Issue (S2): 181-186.doi: 10.16265/j.cnki.issn1003-3033.2019.S2.030

• 公共安全 • 上一篇    下一篇

广深高速铁路列车分类型晚点预测*

胡瑞1,2, 徐传玲1,2, 冯永泰1,2, 文超*1,2,3 副教授, 王全泉4   

  1. 1 西南交通大学 综合交通运输国家地方联合工程实验室,四川 成都 610031;
    2 西南交通大学 综合交通大数据应用技术国家工程实验室,四川 成都 610031;
    3 滑铁卢大学 铁路研究中心,加拿大 滑铁卢 N2L3G1;
    4 中国铁路广州局集团公司 运输部,广东 广州 510088
  • 收稿日期:2019-08-04 修回日期:2019-10-12 出版日期:2019-12-30 发布日期:2020-10-28
  • 作者简介:胡 瑞 (1995—),男,四川成都人,硕士研究生,主要研究方向为铁路运输组织优化。E-mail: 791257593@qq.com。
  • 基金资助:
    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)。

Prediction of different types of train delay of Guangzhou-Shenzhen high-speed railway

HU Rui1,2, XU Chuanling1,2, FENG Yongtai1,2, WEN Chao1,2,3, WANG Quanquan4   

  1. 1 National United Engineering Laboratory of Integrated and Intelligent Transportation, Southwest Jiaotong University, Chengdu Sichuan 610031, China;
    2 National Engineering Laboratory of Integrated Transportation Big Data Application Technology, Southwest Jiaotong University, Chengdu Sichuan 610031, China;
    3 Railway Research Center,University of Waterloo, Waterloo N2L3G1, Canada;
    4 Transport Department, China Railway Guangzhou Group Co.,Ltd.,Guangzhou Guangdong 510088, China
  • Received:2019-08-04 Revised:2019-10-12 Online:2019-12-30 Published:2020-10-28

摘要: 为保障高速铁路列车按图行车,降低潜在的行车冲突风险,利用列车运行实绩研究高速铁路列车晚点及其传播特性。首先,通过描述性统计分析广深高铁列车运行实绩数据,得到广深高铁列车运行的初始晚点分布;然后,运用层次聚类算法聚类分析晚点列车,得到4类晚点列车序列,并提取重要的晚点特征变量;最后,基于随机森林模型预测各类晚点列车序列的晚点时间。研究结果表明:结合随机森林模型预测4类晚点列车晚点时间的准确度达到了84%以上。

关键词: 高速铁路, 列车晚点, 数据驱动, 层次聚类, 随机森林模型

Abstract: In order to ensure that high-speed railway trains operate according to schedule and reduce the risk of potential traffic conflicts, the delay of high-speed trains and their propagation rules were studied by using the train operation performance. Firstly, through the descriptive statistical analysis of the actual train operation data of the Guangzhou-Shenzhen high-speed railway, the primary delay distribution of the train operation was obtained. Secondly, the hierarchical clustering algorithm was used to analyze the delayed trains, and four kinds of delayed train sequences were obtained. Based on this, the delay feature variables were extracted. Finally, the random forest model was used to predict delay time of all kinds of delay train sequences. The results show that combined with the random forest model, the accuracy of predicting delay time of four kinds of delayed trains is more than 84%.

Key words: high-speed railway, train delays, data driven, hierarchical clustering, random forest model

中图分类号: