基于可解释机器学习框架的列车乘车区段客流分布预测方法

doi:10.16097/j.cnki.1009-6744.2024.02.025

交通运输系统工程与信息 ›› 2024, Vol. 24 ›› Issue (2): 249-262.DOI: 10.16097/j.cnki.1009-6744.2024.02.025

基于可解释机器学习框架的列车乘车区段客流分布预测方法

孙国锋^1a，景云^*1a,1b，李和壁²，田志强^3a,3b，田小鹏^3a

1. 北京交通大学，a.交通运输学院，b.智慧高速铁路系统前沿科学中心，北京100044；2.中国铁道科学研究院集团有限公司，铁道科学技术研究发展中心，北京100081；3.兰州交通大学，a.交通运输学院， b. 高原铁路运输智慧管控铁路行业重点实验室，兰州730070

收稿日期:2023-11-14 修回日期:2024-02-16 接受日期:2024-02-23 出版日期:2024-04-25 发布日期:2024-04-25
作者简介:孙国锋(1994- )，男，甘肃通渭人，博士生。
基金资助:
国家自然科学基金 (52372300, 72161023)；中央高校基本科研业务费专项资金(2023YJS146)。

An Interpretable Machine Learning Framework-based Approach for Predicting Passenger Flow Distribution in Train Riding Sections

SUNGuofeng1a, JING Yun*1a,1b, LI Hebi2, TIAN Zhiqiang3a,3b, TIAN Xiaopeng3a

1a. School of Traffic and Transportation, 1b. Frontiers Science Center for Smart High-speed Railway Systems, Beijing Jiaotong University, Beijing 100044, China; 2. Railway Science and Technology Research and Development Center, China Academy of Railway Sciences Corporation Limited, Beijing 100081, China; 3a. School of Traffic and Transportation, 3b. Key Laboratory of Railway Industry on Plateau Railway Transportation Intelligent Management and Control, Lanzhou Jiaotong University, Lanzhou 730070, China

Received:2023-11-14 Revised:2024-02-16 Accepted:2024-02-23 Online:2024-04-25 Published:2024-04-25
Supported by:
NationalNaturalScienceFoundation of China (52372300, 72161023)；Fundamental Research Funds for the Central Universities of Ministry of Education of China (2023YJS146)。

摘要/Abstract

摘要： 为解释客运产品特征对列车乘车区段客流分布预测的影响，本文提出一种基于可解释机器学习框架的高速铁路列车乘车区段客流分布预测方法。首先，提出基于梯度提升树模型的高速铁路列车乘车区段客流分布预测框架，构建不同梯度提升树模型(GBDT、XGBoost、LightGBM及CatBoost)的高速铁路列车乘车区段客流分布预测模型；其次，计算特征贡献重要度，基于SHAP(SHapleyAdditive exPlanations)方法实现特征变量优化，揭示单一特征和交互特征与列车乘车区段客流分布预测的非线性关系。北京南—上海虹桥间列车客流分布预测结果表明：4种模型可精准预测客流分布结果，GBDT，XGBoost，LightGBM及CatBoost在测试集的决定系数分别为 0.9664，0.9601，0.9680及0.9715；特征优化后，按贡献重要度排序依次为标杆车，票价，旅行时间，日期，星期，车次及出发时间；特征优化后，CatBoost-7模型在验证集中的决定系数为0.9458；日期和标杆车对客流分布预测呈现非线性正相关，旅行时间对客流分布预测呈现非线性负相关，低旅行时间、高票价及出发时间整点的标杆车对客流分布预测产生正向影响。本文研究结果能够为高速铁路客运产品设计提供一定参考价值。

关键词: 铁路运输, 客流分布预测, 可解释机器学习, 列车乘车区段, 非线性关系

Abstract: In order to clarify the impact of railway passenger transportation services on the prediction of passenger flow distribution, we propose a method based on an interpretable machine learning framework to predict passenger flow distribution in high-speed railway sections. First, we propose a framework capable of predicting passenger flow distribution in sections by using gradient-boosted tree models. Meanwhile, we construct different gradient-boosted tree models, including GBDT, XGBoost, LightGBM, and CatBoost. Secondly, the importance of feature contributions and feature variables are calculated using the SHapley Additive exPlanations (SHAP) method. A non-linear relationship between different features and passenger flow distribution is revealed. The experiment from Beijing South to Shanghai Hongqiao shows that all four models accurately predict the distribution. The coefficients of determination for GBDT, XGBoost, LightGBM, and CatBoost in the test set are 0.9664, 0.9601, 0.9680, and 0.9715 respectively. After optimizing the features, the order of importance in the contribution is as follows: benchmark train, ticket price, travel time, date, day of the week, and train code departure time. The coefficient of determination for the CatBoost-7 model in the validation set after feature optimization is 0.9458. Both the date and the benchmark train show a non-linear positive correlation with the passenger flow distribution prediction, while the travel time shows a non- linear negative correlation. In addition, low travel time, high ticket price and the benchmark train departing exactly at the scheduled departure time positively influence the passenger flow distribution prediction. This study provides valuable insights into the design of high-speed rail passenger transportation services.

Key words: railway transportation, passenger flow distribution forecast, interpretable machine learning, train-riding segments, non-linear relationship

中图分类号:

U293. 1

孙国锋, 景云, 李和壁, 田志强, 田小鹏. 基于可解释机器学习框架的列车乘车区段客流分布预测方法[J]. 交通运输系统工程与信息, 2024, 24(2): 249-262.

SUNGuofeng, JING Yun, LI Hebi, TIAN Zhiqiang, TIAN Xiaopeng. An Interpretable Machine Learning Framework-based Approach for Predicting Passenger Flow Distribution in Train Riding Sections[J]. Journal of Transportation Systems Engineering and Information Technology, 2024, 24(2): 249-262.

导出引用管理器 EndNote|Ris|BibTeX

链接本文: http://www.tseit.org.cn/CN/10.16097/j.cnki.1009-6744.2024.02.025

http://www.tseit.org.cn/CN/Y2024/V24/I2/249

参考文献

[1]周浪雅,王亦乐,谢余晨,等.站城融合背景下高速铁路综合枢纽短时客流预测研究[J].铁道学报,2023,45 (4): 1-7. [ZHOULY, WANGYL, XIEYC, et al. Prediction of short-termpassenger flow of high-speed railway integrated passenger hub under station-city integration[J]. Journal of the China Railway Society, 2023,45(4):1-7.]

[2]李洁,彭其渊,文超.基于LSTM深度神经网络的高速铁路短期客流预测研究[J].系统工程理论与实践, 2021, 41(10): 2669-2682. [LI J,PENGQY,WENC. Short term passenger fow prediction of high speed railwaybasedonLSTMdeepneuralnetwork[J].Systems Engineering-Theory & Practice, 2021,41(10):2669 2682.]

[3]何必胜,朱永俊,陈路锋,等.基于时空图神经网络的高速铁路车站短期客流预测方法[J].铁道学报,2022, 44(9): 1-8. [HEBS, ZHUYJ, CHENLF, et al.A spatial-temporal graph neural network forpredictionof short-termpassenger flow athigh-speedrailwaystation [J]. Journal of the ChinaRailwaySociety,2022,44(9): 1-8.]

[4] FENGX,ZHANGH,GANT, et al.Randomcoefficient modeling research on short-term forecast ofpassenger flowintoanurbanrail transitstation[J].Transport,2016, 31(1):94-99.

[5]李夏苗,曹尧谦,王慧晶,等.基于模糊预测的高速铁路客流OD表的推算方法[J].交通运输系统工程与信息,2014,14(5):93-99. [LIXM,CAOYQ,WANGHJ, et al. An estimationmethod of passengerOD flows of high-speed railway network based onfuzzypredication [J]. Journal of TransportationSystemsEngineering and InformationTechnology,2014,14(5):93-99.]

[6] YANG F, SHUAIC, QIANQ, et al. Predictabilityof short-termpassengers' originanddestinationdemandsin urbanrail transit[J]. Transportation, 2023, 50(6): 2375 2401.

[7] LVS,WANGK, YANGH,etal.Anorigin-destination passenger flowpredictionsystembasedonconvolutional neural network and passenger source-based attention mechanism[J]. Expert SystemswithApplications, 2024, 238:121989.

[8]陈方遒,景云,郭思冶.基于旅客出行选择特征的高速铁路浮动定价策略研究[J].铁道学报,2022,44(6):11 17. [CHENFQ, JINGY,GUOSY.Researchonhigh speed railway floating pricing strategy based on passenger travel behavior characteristics[J]. Journal of theChinaRailwaySociety,2022,44(6):11-17.]

[9]李和壁,梁家健,高扬.基于MLPs-dynFWA模型的高速铁路短时客流预测方法研究[J].铁道运输与经济, 2021, 43(6): 28-36. [LI H B, LIANG J J, GAO Y. Short term passenger flow prediction of high speed railway based on MLPs-dynFWA model[J]. Railway Transport and Economy, 2021, 43(6): 28-36.]

[10] BAPAUME T, COME E, AMELI M, et al. Forecasting passenger flows and headway at train level for a public transport line: Focus on atypical situations[J]. Transportation Research Part C: Emerging Technologies, 2023, 153: 104195.

[11] XUE Q, ZHANG W, DING M, et al. Passenger flow forecasting approaches for urban rail transit: A survey[J]. International Journal of General Systems, 2023, 52(8): 919-947.

[12] CHEN Y, SHA Y, ZHU X, et al. Prediction of Shanghai metro line 16 passenger flow based on time series analysis[J]. Operations Research and Fuzziology, 2016, 6 (1): 15-26.

[13] CAO L, LIU S G, ZENG X H, et al. Passenger flow prediction based on particle filter optimization[J]. Applied Mechanics and Materials, 2013, 373: 1256 1260.

[14] JIAO P, LI R, SUN T, et al. Three revised Kalman filtering models for short-term rail transit passenger flow prediction[J]. Mathematical Problems in Engineering, 2016, 2016: 9717582.

[15] ZHANG J, CEN F, CUI Z, et al. Deep learning architecture for short-term passenger flow forecasting in urban rail transit[J]. IEEE Transactions on Intelligent Transportation Systems, 2020, 22(11): 7004-7014.

[16] KE J, ZHENG H, YANG H, et al. Short-term forecasting of passenger demand under on-demand ride services: A spatio-temporal deep learning approach[J]. Transportation Research Part C: Emerging Technologies, 2017, 85: 591-608.

[17] WEN K, ZHAO G, HE B, et al. A decomposition-based forecasting method with transfer learning for railway short-term passenger flow in holidays[J]. Expert Systems with Applications, 2022, 189: 116102.

[18] SUN Y, JIANG Z, GU J, et al. Analyzing high speed rail passengers' train choices based on new online booking data in China[J]. Transportation Research Part C: Emerging Technologies, 2018, 97: 96-113.

[19] 程谦,杨光,胡启洲.基于混合Logit模型的旅客对短途高速铁路列车选择行为[J].中国铁道科学,2021,42 (2): 183-192. [CHENG Q, YANG G, HU Q Z. Passenger choice behaviors for short-haul HSR trains based on mixed Logit model [J]. China Railway Science, 2021, 42 (2): 183-192.]

[20] 刘佳伟,杨信丰,马艺轩.突发事件下考虑旅客时间价值的列车开行方案研究[J]. 铁道科学与工程学报, 2023, 20(5): 1598-1610. [LIU J W, YANG X F, MA Y X. Train operation plan considering the time value of passengers under emergencies[J]. Journal of Railway Science and Engineering, 2023, 20(5): 1598-1610.]

[21] 夏铭泽, 李博,张博,等.基于均衡性和能力利用的高速铁路标杆车运行图编制方法研究[J/OL].铁道科学与工程学报, (2023-12-21) [2024-02-02]. https://doi. org/10.19713/j.cnki.43-1423/u.T20231426. [XIA M Z, LI B, ZHANG B. Research on benchmark train working diagram based on equilibrium and capacity utilization [J/OL]. Journal of Railway Science and Engineering, (2023-12-21) [2024-02-02]. https://doi.org/10.19713/ j.cnki.43-1423/u.T20231426.]

[22] DIETTERICH T G. Ensemble learning[J]. The Handbook of Brain Theory and Neural Networks, 2002, 2(1): 110 125.

[23] CHEN T, GUESTRIN C. Xgboost: A scalable tree boosting system[C]//New York: In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 2016.

[24] KE G L, MENG Q, FINLEY T, et al. LightGBM: A highly efficient gradient boosting decision tree[C]//San Diego: Neural Information Processing Systems Foundation, Inc. Advances in Neural Information Processing Systems 30 (NIPS 2017). 2017.

[25] PROKHORENKOVA L, GUSEV G, VOROBEV A, et al. CatBoost: Unbiased boosting with categorical features [C]//Montréal: 32nd Conference on Neural Information Processing Systems, 2017.

[26] LUNDLBERG S M, LEE S I. A unified approach to interpreting model predictions[C]// New York: Proceedings of the 31st International Conference on Neural Information Processing Systems, 2017.

基于可解释机器学习框架的列车乘车区段客流分布预测方法

An Interpretable Machine Learning Framework-based Approach for Predicting Passenger Flow Distribution in Train Riding Sections

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

[1]	帅斌, 罗佳楠, 冯心妍, 黄文成. 虚拟编组下基于跟驰模型的列车群运行控制方法研究[J]. 交通运输系统工程与信息, 2024, 24(3): 1-11.
[2]	余琼霞, 候怡腾, 孙俊杰, 侯忠生. 高速列车受限自适应有限次迭代学习容错控制[J]. 交通运输系统工程与信息, 2024, 24(3): 140-150.
[3]	周文梁, 蒋志刚, 柴乃杰, 徐光明. 高速铁路列车开行方案与票价票额综合优化[J]. 交通运输系统工程与信息, 2024, 24(3): 151-163.
[4]	罗钦, 黄杉, 宋剑伟, 曾翠峰, 陈菁菁, 李伟. 基于资源约束的地铁运营施工任务调度研究[J]. 交通运输系统工程与信息, 2024, 24(2): 188-198.
[5]	刘柯良, 陈坚, 邱智宣, 张迪, 唐祯, 彭倩. 城市路内违章停车时空特征及影响因素作用模型[J]. 交通运输系统工程与信息, 2024, 24(2): 234-248.
[6]	苏鹏飞, 王果, 闵永智, 陶金豆, 李旭东, 刘昕玥. 高速铁路信号系统碳排放边界划分及分析[J]. 交通运输系统工程与信息, 2024, 24(2): 293-303.
[7]	常祎妹, 汪洋, 朱晓宁. 基于运载能力的铁路集装箱中心站轨道吊与集卡协同优化[J]. 交通运输系统工程与信息, 2024, 24(1): 188-198.
[8]	吴艺迪, 何世伟, 周汉. 重载铁路智慧港口站作业计划综合优化模型及算法研究[J]. 交通运输系统工程与信息, 2023, 23(6): 215-226.
[9]	张正坤, 朱昌锋, 景云, 邢进. 基于候车时间和效益损失的城轨列车运行计划优化[J]. 交通运输系统工程与信息, 2023, 23(5): 215-226.
[10]	庞子帅, 王丽雯, 彭其渊. 基于强化学习的干扰条件下高速铁路时刻表调整研究[J]. 交通运输系统工程与信息, 2023, 23(5): 279-289.
[11]	林俊亭, 倪铭君. 基于扩张状态观测器的虚拟编组触发模型预测控制[J]. 交通运输系统工程与信息, 2023, 23(4): 134-146.
[12]	洪治潮, 张锦, 孙文杰, 申皓, 赵钢, 梁宏斌. 区块链背景下的中欧班列网络配流研究[J]. 交通运输系统工程与信息, 2023, 23(4): 270-281.
[13]	彭宏勤, 张国伍. “双碳”目标下高铁与民航高质量发展策略——“交通7+1论坛”第五十九次会议[J]. 交通运输系统工程与信息, 2023, 23(3): 1-5.
[14]	侯涛, 唐丽, 牛宏侠. 基于数据驱动的高速列车速度复合控制研究[J]. 交通运输系统工程与信息, 2023, 23(3): 145-152.
[15]	邰国璇, 黄友能, 李春驰, 王悉. 基于灵活编组的市域快轨时刻表优化方法研究[J]. 交通运输系统工程与信息, 2023, 23(3): 195-203.