基于自然驾驶数据的驾驶人跟驰行为内在异质性预测与建模

doi:10.16097/j.cnki.1009-6744.2023.05.004

交通运输系统工程与信息 ›› 2023, Vol. 23 ›› Issue (5): 33-44.DOI: 10.16097/j.cnki.1009-6744.2023.05.004

• 智能交通系统与信息技术 • 上一篇下一篇

基于自然驾驶数据的驾驶人跟驰行为内在异质性预测与建模

张铎¹，饶红玉^{1, 2}，刘佳琦¹，王俊骅¹，孙剑^*1

1. 同济大学，道路与交通工程教育部重点实验室，上海 201804；2. 杭州海康威视数字技术股份有限公司，杭州 310051

收稿日期:2023-04-26 修回日期:2023-06-25 接受日期:2023-07-10 出版日期:2023-10-25 发布日期:2023-10-22
作者简介:张铎(1994- )，男，山东菏泽人，博士生。
基金资助:
国家自然科学基金(52125208，52232015)；中央高校学科交叉重点项目 (2022-5-ZD-02)。

Intra-driver Heterogeneity Prediction and Modeling Based on Naturalistic Driving Experiment

ZHANG Duo¹,RAO Hong-yu^{1, 2},LIU Jia-qi¹,WANG Jun-hua¹,SUN Jian^{* 1}

1. The Key Laboratory of Road and Traffic Engineering, Ministry of Education, Tongji University, Shanghai 201804, China; 2. Hangzhou Hikvision Digital Technology Co. Ltd., Hangzhou 310051, China

Received:2023-04-26 Revised:2023-06-25 Accepted:2023-07-10 Online:2023-10-25 Published:2023-10-22
Supported by:
National Natural Science Foundation of China (52125208，52232015)；Fundamental Research Funds for the Central Universities of Ministry of Education of China (2022-5-ZD-02)。

摘要/Abstract

摘要： 为了对交通流建模研究和高级辅助驾驶开发等提供可靠跟驰行为支持，本文针对同一位驾驶人个体内部发生的从正常到异常的跟驰行为模式的转变，提出一种以Transformer深度学习模型为基础的驾驶人跟驰行为内在异质性预测和建模方法。本文基于超过20万km的大规模自然驾驶实验，首先，利用对驾驶人的全要素长期行为观测，为 41 位驾驶人建立基线模型，完成3194次驾驶人内在异质性事件识别与提取；其次，设计基于Transformer多头自注意力机制的深度学习预测器，实现对驾驶人内在异质性事件的准确预测。结果表明，预测器在跟驰内在异质性三分类时间点预测的F1分数(F1-Score)相较于长短时记忆网络表现更优，达到了87.13%，基于预测结果的动态跟驰参数切换可以降低21.08%的驾驶行为建模误差。研究结果有助于更加准确地重现驾驶人跟驰行为，进一步提高交通流仿真精度和高级辅助驾驶控制策略拟人化水平。

关键词: 交通工程, 驾驶人内在异质性, Transformer模型, 跟驰行为, 自注意力机制

Abstract: To provide reliable support for car-following behavior in traffic flow modeling research and advanced driving assistance systems, this study proposes a method for predicting and modeling the intrinsic heterogeneity of driver following behavior based on the Transformer deep learning model and considers individual driver's behavior changes in the car-following process. This study is based on a large-scale naturalistic driving experiment, which involved over 200000 kilometers of naturalistic driving records. First, the baseline models are developed using longterm behavioral observations of 41 drivers, and 3194 intra-driver heterogeneity events are identified and extracted according to the baseline model. Further, a deep learning predictor based on the Transformer multi-head self-attention mechanism is designed to accurately predict intra-driver heterogeneity events of drivers. The results show that the predictor performs better than the long short-term memory network in predicting the three-class time points of carfollowing intra-driver heterogeneity, with an F1 score of 87.13%. Based on prediction results, dynamic car-following parameter switching can reduce the driving behavior modeling error by 21.08%. The research results help to understand the behavioral response mechanism of drivers, and further improve the accuracy of traffic flow simulation and the design of personalized control strategies.

Key words: traffic engineering, intra-driver heterogeneity, Transformer model, car-following behavior, self-attention mechanism

中图分类号:

U491.2

张铎, 饶红玉, 刘佳琦, 王俊骅, 孙剑. 基于自然驾驶数据的驾驶人跟驰行为内在异质性预测与建模[J]. 交通运输系统工程与信息, 2023, 23(5): 33-44.

ZHANG Duo, RAO Hong-yu, LIU Jia-qi, WANG Jun-hua, SUN Jian. Intra-driver Heterogeneity Prediction and Modeling Based on Naturalistic Driving Experiment[J]. Journal of Transportation Systems Engineering and Information Technology, 2023, 23(5): 33-44.

导出引用管理器 EndNote|Ris|BibTeX

链接本文: http://www.tseit.org.cn/CN/10.16097/j.cnki.1009-6744.2023.05.004

http://www.tseit.org.cn/CN/Y2023/V23/I5/33

参考文献

[1] 鲍琼, 屈琦凯, 唐涵润, 等. 网联环境下基于多驾驶人风险评价的不良行为主动干预研究[J]. 交通运输系统工程与信息, 2022, 22(4): 283-292. [BAO Q, QU Q K, TANG H R, et al. Multi-drivers risk evaluation based proactive intervention of drivers' risky behavior under connected transportation contexts[J]. Journal of Transportation Systems Engineering and Information Technology, 2022, 22(4): 283-292.]

[2] HAN J, WANG X, WANG G. Modeling the car-following behavior with consideration of driver, vehicle, and environment factors: A historical review[J]. Sustainability, 2022, 14(13): 8179.

[3] WEI D, LIU H. Analysis of asymmetric driving behavior using a self-learning approach[J]. Transportation Research Part B: Methodological, 2013, 47: 1-14.

[4] LAVAL J A, LECLERCQ L. Microscopic modeling of the relaxation phenomenon using a macroscopic lane-changing model[J]. Transportation Research Part B: Methodological, 2008, 42(6): 511-522.

[5] HIGGS B, ABBAS M. Segmentation and clustering of car-following behavior: Recognition of driving patterns[J]. IEEE Transactions on Intelligent Transportation Systems, 2014, 16(1): 81-90.

[6] ZHANG D, CHEN X, WANG J, et al. A comprehensive comparison study of four classical car-following models based on the large-scale naturalistic driving experiment [J]. Simulation Modelling Practice and Theory, 2021, 113: 102383.

[7] RAO H, ZHANG D, QIN G, et al. Investigating the intradriver heterogeneity in car following behavior based on large-scale naturalistic driving study[J]. Transportmetrica B: Transport Dynamics, 2023, 11(1): 1363-1383.

[8] TREIBER M, HENNECKE A, HELBING D. Congested traffic states in empirical observations and microscopic simulations[J]. Physical Review E, 2000, 62(2): 1805.

[9] AHMED A, NGODUY D, ADNAN M, et al. On the fundamental diagram and driving behavior modeling of heterogeneous traffic flow using UAV-based data[J]. Transportation Research Part A: Policy and Practice, 2021, 148: 100-115.

[10] 张文娟, 杨皓哲, 张彬, 等. 考虑多时间尺度特征的城市轨道交通短时客流量预测模型[J]. 交通运输系统工程与信息, 2022, 22(6): 212-223. [ZHANG W J, YANG H Z, ZHANG B, et al. Short-time passenger flow prediction model of urban rail transit considering multi-timescale features[J]. Journal of Transportation Systems Engineering and Information Technology, 2022, 22(6): 212-223.]

[11] VASWANI A, SHAZEER N, PARMAR N, et al. Attention is all you need[J]. Advances in Neural Information Processing Systems, 2017, 30: 6000-6010.

[12] AHSAN H. A study on how data quality influences machine learning predictability and interpretability for tabular data[D]. Youngstown: Youngstown State University, 2022.

[13] LU Y, RAI H, CHANG J, et al. Context- aware scene graph generation with seq2seq transformers[C]// Proceedings of the IEEE/CVF international conference on computer vision. Online: IEEE Xplore, 2021. 15931- 15941.

[14] AKIBA T, SANO S, YANASE T, et al. Optuna: A nextgeneration hyperparameter optimization framework[C]// Proceedings of the Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining: Anchorage: ACM, 2019: 2623-2631.

[15] HOCHREITER S, SCHMIDHUBER J. Long short-term memory [J]. Neural Computation, 1997, 9(8): 1735-1780.

基于自然驾驶数据的驾驶人跟驰行为内在异质性预测与建模

Intra-driver Heterogeneity Prediction and Modeling Based on Naturalistic Driving Experiment

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]	曲大义, 孟奕名, 王韬, 宋慧, 陈意成. 基于分子力场的网联自主车辆跟驰安全特性及模型[J]. 交通运输系统工程与信息, 2023, 23(6): 33-41.
[2]	王嘉文, 孙晨晨, 赵靖, 杭佳宇. 交通事件下路段通行能力多层模糊估计方法[J]. 交通运输系统工程与信息, 2023, 23(6): 100-110.
[3]	李岩, 陈姜会, 曾明哲, 徐金华, 汪帆. 考虑天气影响的高速公路交织区交通运行状态识别[J]. 交通运输系统工程与信息, 2023, 23(6): 111-119.
[4]	林建新, 刘依妮, 朱雪超, 刘博. 基于时空认知的违章停车行为治理策略研究[J]. 交通运输系统工程与信息, 2023, 23(6): 133-140.
[5]	孙超, 尹浩为, 宋茂灿, 张鹏. 考虑潜在需求的区域共享泊位滚动时域分配研究[J]. 交通运输系统工程与信息, 2023, 23(6): 141-152.
[6]	陈信强, 王美琳, 李朝锋, 杨洋, 梅骁峻, 周亚民. 基于深度学习与多级匹配机制的港区人员轨迹提取[J]. 交通运输系统工程与信息, 2023, 23(4): 70-79.
[7]	高云峰, 席建伟, 孙科. 面向客货分离交叉口的网联货车队列车速引导与信号优先组合优化[J]. 交通运输系统工程与信息, 2023, 23(4): 88-101.
[8]	管德永, 徐越, 王可. 基于动态灵敏度的感应控制优化方法[J]. 交通运输系统工程与信息, 2023, 23(4): 124-133.
[9]	柏强, 江瑛, 李毅帆, 徐誌蔓. 基于改进时空轨迹理论的机场出发层车道边通行能力评估方法研究[J]. 交通运输系统工程与信息, 2023, 23(4): 228-236.
[10]	郝嘉田, 吴忠广, 田万利, 韩峰, 刘博文. 高速公路改扩建中央分隔带开口长度通用计算模型构建与验证[J]. 交通运输系统工程与信息, 2023, 23(4): 262-269.
[11]	秦严严, 肖腾飞, 贺正冰. 雾天场景下高速公路通行能力分析及提升策略[J]. 交通运输系统工程与信息, 2023, 23(3): 39-47.
[12]	王殿海, 黄莉莎, 曾佳棋, 蔡正义. 降雨对城市路网交通状态的影响机理[J]. 交通运输系统工程与信息, 2023, 23(3): 48-55.
[13]	葛慧敏, 臧文铠, 董磊, 周礼军. 大型多环道涡轮环交车道级信号控制方法[J]. 交通运输系统工程与信息, 2023, 23(3): 123-132.
[14]	付锐, 许清津, 葛振振, 陈桃, 吴付威, 袁伟. 弯道路侧标志梯度提示策略研究[J]. 交通运输系统工程与信息, 2023, 23(3): 133-144.
[15]	李思贤, 邢增勇, 钱大琳, 李鹏程, 袁梦. 危险货物道路运输车辆驾驶人风险倾向性分类研究[J]. 交通运输系统工程与信息, 2023, 23(3): 161-173.