数据驱动的高速公路自学习元胞传输模型

doi:10.16097/j.cnki.1009-6744.2025.05.009

交通运输系统工程与信息 ›› 2025, Vol. 25 ›› Issue (5): 103-113.DOI: 10.16097/j.cnki.1009-6744.2025.05.009

数据驱动的高速公路自学习元胞传输模型

林培群¹，黄超铄¹，周楚昊^*2，庞崇浩¹，邓锴宇¹

1. 华南理工大学，土木与交通学院，广州510641；2.广东技术师范大学，汽车与交通工程学院，广州510450

收稿日期:2025-06-24 修回日期:2025-09-02 接受日期:2025-09-04 出版日期:2025-10-25 发布日期:2025-10-25
作者简介:林培群(1980—)，男，广东潮州人，教授，博士。
基金资助:
国家自然科学基金(52072130)；广东省自然科学基金(2025A1515010046)。

A Data-driven Cellular Transmission Model for Self-learning on Expressways

LIN Peiqun¹, HUANG Chaoshuo¹, ZHOU Chuhao^*2, PANG Chonghao¹, DENG Kaiyu¹

1. School of Civil and Transportation Engineering, South China University of Technology, Guangzhou 510641, China; 2. School of Automobile and Transportation Engineering, Guangdong Polytechnic Normal University, Guangzhou 510450, China

Received:2025-06-24 Revised:2025-09-02 Accepted:2025-09-04 Online:2025-10-25 Published:2025-10-25
Supported by:
National Natural Science Foundation of China(52072130)；Natural Science Foundation of Guangdong Province, China (2025A1515010046)。

摘要/Abstract

摘要： 高效的交通仿真模型能够为交通管理部门提供实时和短期的路段流量变化情况，为主动交通管理与路网优化疏导提供科学依据。然而，在复杂交通场景下，模型参数易受环境影响发生变化，导致仿真精度下降。本文提出一种数据驱动的自学习元胞传输模型(Self-Learning Cell Transmission Model, SL-CTM)。模型采用数据驱动方式，通过对元胞输入特征、内部状态与输出流量的自适应拟合，自主学习元胞传输模型中需要人工标定的参数，有效规避复杂参数标定过程，提升仿真的准确性与运行效率。基于广东省南二环高速公路和佛开高速公路实测数据的验证结果表明：与随机森林模型相比，SL-CTM在两条道路的流量仿真加权平均绝对误差百分比(Weighted Mean Absolute Percentage Error,WMAPE)分别下降17.55%和15.83%；与长短期记忆网络相比，SL-CTM在两条道路的流量仿真WMAPE分别下降12.37%和10.50%；说明SL-CTM在使用更少初始特征的同时具备更强的流量突变响应能力；与SUMO(Simulation of Urban Mobility)仿真软件相比，SL-CTM的WMAPE下降55.90%，仿真速度提升72.57%，在高流量场景中表现出更优的仿真性能。研究表明，SL-CTM能够显著提升交通仿真的精度与计算效率，为复杂交通环境下的动态交通管理提供更为可靠的技术支持。

关键词: 智能交通, 交通流预测, 元胞传输模型, 高速公路, 数据驱动建模

Abstract: An efficient traffic simulation model can offer a scientific basis on proactive traffic management and network optimization for transportation authorities by providing the changes of real-time and short-term traffic flow. However, in complex traffic scenarios, model parameters are susceptible to environmental influences, leading to the decrease in simulation accuracy. This paper proposes a data-driven Self-Learning Cellular Transmission Model (SL-CTM), which employs a data-driven approach, adaptively fitting the input features, internal states, and output flows of cells. It can autonomously learn the parameters that typically require manual calibration in the cellular transmission model, thereby effectively avoiding the complex parameter calibration process and enhancing simulation accuracy and operational efficiency. Validation results based on the empirical data from the South Second Ring Expressway and the Fokai Expressway in Guangdong Province indicate that, compared with the Random Forest model, the SL-CTM reduces the Weighted Mean Absolute Percentage Error (WMAPE) of traffic flow simulation by 17.55% and 15.83% on the two expressways, respectively. Compared with the Long Short-Term Memory (LSTM) model, the SL-CTM achieves reductions of 12.37% and 10.50% in WMAPE, respectively. These findings demonstrate that the SL-CTM is capable of achieving a stronger responsiveness to sudden traffic flow variations while requiring fewer initial features. Compared to the SUMO simulation software, SL-CTM achieves a 55.90% reduction in WMAPE and a 72.57% improvement in simulation speed, which exhibits superior performance in high-traffic scenarios. The study demonstrates that SL-CTM significantly improves the accuracy and computational efficiency of traffic simulation, providing more reliable technical support for dynamic traffic management in complex traffic environments.

Key words: intelligent transportation, traffic flow prediction, cellular transmission model, highway, data-driven modeling

中图分类号:

U491

林培群, 黄超铄, 周楚昊, 庞崇浩, 邓锴宇. 数据驱动的高速公路自学习元胞传输模型[J]. 交通运输系统工程与信息, 2025, 25(5): 103-113.

LIN Peiqun, HUANG Chaoshuo, ZHOU Chuhao, PANG Chonghao, DENG Kaiyu. A Data-driven Cellular Transmission Model for Self-learning on Expressways[J]. Journal of Transportation Systems Engineering and Information Technology, 2025, 25(5): 103-113.

导出引用管理器 EndNote|Ris|BibTeX

链接本文: http://www.tseit.org.cn/CN/10.16097/j.cnki.1009-6744.2025.05.009

http://www.tseit.org.cn/CN/Y2025/V25/I5/103

参考文献

[1] 蔡浩,李林峰,李涵,等.基于极限学习机的短期交通流预测混合优化模型[J]. 交通运输系统工程与信息, 2023, 23(5): 75–82. [CAI H, LI L F, LI H, et al. Short-term traffic flow prediction based on ASO-ELM hybrid optimization model[J]. Journal of Transportation Systems Engineering and Information Technology, 2023, 23(5): 75–82.]

[2] LIU X, JIANG Y, HUANG X, et al. MLSTAM: A multidimensional long-term spatio-temporal attention model for traffic flow forecasting by capturing time series correlations[J]. Intelligent Data Analysis, 2024: 1088467X251336926.

[3] DONG C, FENG X, WANG Y, et al. Spatiotemporal exogenous variables enhanced model for traffic flow prediction[J]. IEEE Access, 2023, 11: 95958–95973.

[4] XU C, SHAO Y, MA C, et al. A geometric deep learning approach to traffic flow shockwave prediction on freeways using vehicle trajectory data and HD map[J]. IEEE Transactions on Intelligent Transportation Systems, 2025, 26(7): 9907–9917

[5] ZHANG Y, KOUVELAS A, MAKRIDIS M A. A time-varying shockwave speed model for reconstructing trajectories on freeways using Lagrangian and Eulerian observations[J]. Expert Systems with Applications, 2024, 253: 124298.

[6] MUSTAFA A, KHATTAK K S, KHAN Z H. Predicting short-term traffic with random forest: A granular data approach[C]//Gujrat: International Conference on Energy, Power, Environment, Control and Computing (ICEPECC 2025). IEEE, 2025: 1–7.

[7] GUO Q, TAN Q, PENG Y, et al. Model-enhanced spatial-temporal attention networks for traffic density prediction[J]. Complex & Intelligent Systems, 2025, 11 (1): 19.

[8] CHEN J, XU M, XU W, et al. A flow feedback traffic prediction based on visual quantified features[J]. IEEE Transactions on Intelligent Transportation Systems, 2023, 24(9): 10067–10075.

[9] 邓明君,李书行,李响,等.基于主线密度预测的快速路入口匝道控制方法[J]. 交通运输系统工程与信息, 2024, 24(4): 94–104. [DENG M J, LI S X, LI X, et al. Expressway entrance ramp control method based on mainline traffic density prediction[J]. Journal of Transportation Systems Engineering and Information Technology, 2024, 24(4): 94–104.]

[10] 周浩,胡坚明,张毅,等. 快速路可变限速与匝道控制协同优化策略[J]. 交通运输系统工程与信息, 2017, 17 (2): 68–75. [ZHOU H, HU J M, ZHANG Y, et al. A coordinated optimization strategy of variable speed limit and ramp metering for expressway[J]. Journal of Transportation Systems Engineering and Information Technology, 2017, 17(2): 68–75.]

[11] CENEDESE C, LUCCHINI M, CUCUZZELLA M, et al. On the effect of capacity drops in highways with service stations[C]/Limassol: 2023 31st Mediterranean Conference on Control and Automation (MED), IEEE, 2023: 478–483.

[12] 易术, 黄丹阳. 融合多源数据与元胞传输模型的高速公路交通状态估计方法[J]. 交通运输工程与信息学报, 2023, 21(4): 103–114. [YI S, HUANG D Y. Freeway traffic state estimation based on multi-source data and cell transmission model[J]. Journal of Transportation Engineering and Information, 2023, 21(4): 103–114.]

[13] DEEPIKA, PANDOVE G. Advanced modeling and interpretation for accurate intersection traffic time prediction[J]. International Journal of Intelligent Transportation Systems Research, 2024, 22(3): 735–760.

[14] QIU A, SATHISH P A, WANG D, et al. Advanced traffic demand generation in SUMO: ML-based prediction of flow rate based on real-world measured datasets[C]/Singapore: 2024 IEEE 99th Vehicular Technology Conference (VTC2024–Spring), 2024: 1–7.

[15] 姚凯斌, 林培群. 一种考虑交叉口因素的改进元胞传输模型[J]. 交通运输系统工程与信息, 2017, 17(3): 105–111. [YAO K B, LIN P Q. An improved cell transmission model considering intersection factor[J]. Journal of Transportation Systems Engineering and Information Technology, 2017, 17(3): 105–111.]

数据驱动的高速公路自学习元胞传输模型

A Data-driven Cellular Transmission Model for Self-learning on Expressways

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]	田大新, 肖啸, 周建山. AI驱动的自动驾驶汽车轨迹预测方法综述[J]. 交通运输系统工程与信息, 2025, 25(5): 1-24.
[2]	程国柱, 王文志, 杨子涵, 汪国鹏, 陈永胜, 顾爽. 网联自动驾驶车辆专用车道设置研究综述[J]. 交通运输系统工程与信息, 2025, 25(5): 25-39.
[3]	吴付威, 杨慧, 马勇, 崔同川, 闫晟煜, 张智. L3级自动驾驶车辆两阶段预警对接管绩效影响研究[J]. 交通运输系统工程与信息, 2025, 25(5): 40-49.
[4]	窦雪萍, 杨明慧, 熊杰. 考虑自适应控制点的自动驾驶公交运行优化方法[J]. 交通运输系统工程与信息, 2025, 25(5): 50-58.
[5]	宋浪, 胡晓伟, 俞山川, 安实. 网联自动驾驶环境下交叉口车道分配与车辆轨迹协同优化[J]. 交通运输系统工程与信息, 2025, 25(5): 59-71.
[6]	王殿海, 王奕飞, 黄宇浪, 刘泳, 曾佳棋. 车牌识别设备稀疏布局下的路段行程时间估计方法[J]. 交通运输系统工程与信息, 2025, 25(5): 83-90.
[7]	李传耀, 章孟丹, 唐铁桥. 考虑模块化公交供应限制的双模式早通勤均衡分析[J]. 交通运输系统工程与信息, 2025, 25(5): 205-214.
[8]	沈鹏举, 宋丽英, 林川倩. 高铁列车晚点数据的异常检测与增量聚类方法[J]. 交通运输系统工程与信息, 2025, 25(5): 236-247.
[9]	杨静, 侯宇晴, 谢余晨, 杨安安. 城市轨道交通网络短时OD客流多任务协同预测[J]. 交通运输系统工程与信息, 2025, 25(5): 261-270.
[10]	马飞虎, 陈晓燕, 孙翠羽, 田星彤. 高速公路收费站动态车道配置的在线优化方法[J]. 交通运输系统工程与信息, 2025, 25(5): 333-342.
[11]	姚振兴, 刘贤, 赵一飞, 王亮, 王彦琛. 手机信令不均匀定位下出行端点自适应识别方法[J]. 交通运输系统工程与信息, 2025, 25(4): 44-52.
[12]	张鹏, 李兴旺, 姬炳豪, 孙超, 李文权. 路口重复放行的公交与社会车辆协同绿波优化模型[J]. 交通运输系统工程与信息, 2025, 25(4): 53-62.
[13]	代亮, 杜鹏飞, 黄自彬, 杨朋博. 基于深度强化学习的城市交通信号分层协同控制方法[J]. 交通运输系统工程与信息, 2025, 25(4): 63-72.
[14]	王庞伟, 王思淼, 雷方舒, 徐京辉, 王子鹏, 王力. 混合动作表示强化学习下的城市交叉口智慧信控方法[J]. 交通运输系统工程与信息, 2025, 25(4): 73-83.
[15]	王连震, 沈超文, 王宇萍, 薛淑祺. 网联高速公路合流区基于间隙优化的车辆协同控制方法[J]. 交通运输系统工程与信息, 2025, 25(4): 84-95.