抵近信控交叉口分心驾驶识别模型

doi:10.16097/j.cnki.1009-6744.2022.01.023

交通运输系统工程与信息 ›› 2022, Vol. 22 ›› Issue (1): 217-224.DOI: 10.16097/j.cnki.1009-6744.2022.01.023

抵近信控交叉口分心驾驶识别模型

张玉婷^{* 1}，陈波佑¹，张双焱¹，闫学东²，李晓梦³

1. 长安大学，运输工程学院，西安 710021；2. 北京交通大学，综合交通运输大数据应用技术交通运输行业重点实验室，北京 100044；3. 昆士兰科技大学，交通事故与道路安全研究中心，昆士兰 4059，澳大利亚

收稿日期:2021-07-28 修回日期:2021-10-19 接受日期:2021-11-23 出版日期:2022-02-25 发布日期:2022-02-23
作者简介:张玉婷(1991- )，女，甘肃白银人，讲师，博士。
基金资助:
国家自然科学基金；国家重点研发计划；国家级大学生创新创业训练计划

Recognition Model of Distracted Drivers When Approaching Signalized Intersections

ZHANG Yu-ting^{* 1} , CHEN Bo-you¹ , ZHANG Shuang-yan¹ , YAN Xue-dong² , LI Xiao-meng³

1. College of Transportation Engineering, Chang'an Universtiy, Xi'an 710021, China; 2. MOT Key Laboratory of Transportation Industry of Big Data Application Technologies for Comprehensive Transport, Beijing Jiaotong University, Beijing 100044, China; 3. Centre for Accident Research and Road Safety-Queensland (CARRS-Q), Queensland University of Technology, QLD 4059, Australia

Received:2021-07-28 Revised:2021-10-19 Accepted:2021-11-23 Online:2022-02-25 Published:2022-02-23
Supported by:
National Natural Science Foundation of China (71901036)；National Key Research and Development Program of China (2019YFE0108000)；National Innovation and Entrepreneurship Training Program for College Students(S202010710019)

摘要/Abstract

摘要： 为降低驾驶人抵近交叉口过程因分心导致的交通事故，本文基于双向长短时记忆网络 (BILSTM)建立分心驾驶识别模型。依托驾驶模拟实验，采集了45位驾驶人抵近信控交叉口过程中的横纵行为数据，通过方差分析研究分心任务对驾驶行为的影响。结果表明，分心驾驶人需要更长的制动反应时间，制动操作时间缩短，踩压制动踏板的力度下降，同时操纵方向盘的稳定性变差。然后，筛选有显著性影响的6个特征行为指标作为模型的输入，结果表明：BILSTM模型分心状态识别的精确率最高，达到92.6%，F1值为88.7%；准确率、精确率、召回率、F1、AUC和ROC曲线等模型性能均优于单向长短期记忆网络、支持向量机和决策树5.0分心识别模型。研究结果说明BILSTM模型能有效判别抵近信控交叉口驾驶人分心状态，可为交叉口驾驶人分心预警系统的优化设计提供依据和指导。

关键词: 交通工程, 信控交叉口, 分心状态识别, BILSTM模型, 驾驶模拟

Abstract: To reduce the traffic accidents caused by distraction when approaching signalized intersections, this study developed a driver distraction detection model based on bi-directional long short-term memory (BILSTM) method. A series of experiment scenarios approaching intersections were performed using a high- driving simulator. The longitudinal and horizontal behaviors of 45 drivers were analyzed to determine the impacts of driver distractions on driving behaviors using the analysis of variance (ANOVA). The results indicate that the distracted driver needs longer braking response time, operates the braking in shorter time and exerts less pressure on the pedal. Meanwhile, the districted driver shows deteriorated stability when operating the steering wheel. Six indicators that have significant impacts on driving behavior were selected as the input to the distraction recognition model. The results show that the recognition accuracy of the model is 92.6% and the F1-scores is 88.7%. Compared with the long short-term memory (LSTM), support vector machines (SVM) and decision tree 5.0 distraction recognition models, the proposed model shows the best performance in terms of accuracy, recall, the F1-scores, the area under the curve (AUC) and the receiver operating characteristic (ROC) curve. The BILSTM model can effectively distinguish the distraction driving status when the driver approaching signalized intersections, which provides basis and guidance for the optimal design of the driver distraction warning assistance system at the intersection.

Key words: traffic engineering, signalized intersection, recognition of distracted driving, bi-directional long short-term memory (BILSTM) model, driving simulator

中图分类号:

U491

张玉婷, 陈波佑, 张双焱, 闫学东, 李晓梦. 抵近信控交叉口分心驾驶识别模型[J]. 交通运输系统工程与信息, 2022, 22(1): 217-224.

ZHANG Yu-ting , CHEN Bo-you , ZHANG Shuang-yan, YAN Xue-dong , LI Xiao-meng. Recognition Model of Distracted Drivers When Approaching Signalized Intersections[J]. Journal of Transportation Systems Engineering and Information Technology, 2022, 22(1): 217-224.

导出引用管理器 EndNote|Ris|BibTeX

链接本文: http://www.tseit.org.cn/CN/10.16097/j.cnki.1009-6744.2022.01.023

http://www.tseit.org.cn/CN/Y2022/V22/I1/217

参考文献

[1] NHTSA. National highway traffic safety administration, distracted driving 2019[R]. (2019-01-01)[2021-05-01]. https: // www．nhtsa．gov/ risky-driving/ distracteddriving#topic-consequences.

[2] CHOUDHARY P, VELAGA N. Mobile phone use during driving: Effects on speed and effectiveness of driver compensatory behaviour[J]. Accident Analysis & Prevention, 2017, 106: 370-378.

[3] 肖书影, 张志清, 李诗芸. 驾驶人弯道分心对行车安全的影响[J]. 武汉理工大学学报(交通科学与工程版), 2021, 45(2): 213-218. [XIAO S Y, ZHANG Z Q, LI S Y. Effect of driver distraction on driving safety in curved section[J]. Journal of Wuhan University of Technology (Transportation Science & Engineering), 2021, 45(2): 213-218.]

[4] KOUNTOURIOTIS G, SPYRIDAKOS P, CARSTEN O, et al. Identifying cognitive distraction using steering wheel reversal rates[J]. Accident Analysis & Prevention, 2016, 96: 39-45.

[5] 李鹏辉, 胡孟夏, 张文会, 等. 分心对驾驶人交通冲突反应时间的影响[J]. 中国公路学报, 2018, 31(4): 36- 42. [LI P H, HU M X, ZHANG W H, et al. Influence of distraction driver's reaction time to traffic conflicts[J]. China Journal of Highway and Transport, 2018, 31(4): 36-42.]

[6] NAVEEN E, SHAMSUNNAHAR Y. Disentangling the influence of cell phone usage in the dilemma zone: An econometric approach[J]. Accident Analysis & Prevention, 2016, 96: 280-289.

[7] 罗毅, 高岩, 尤志栋. 基于换道行为的驾驶分心识别方法[J]. 中国安全科学学报, 2018, 28(10): 25-30. [LUO Y, GAO Y, YOU Z D. A lane changing behavior based method for detecting driver distraction[J]. China Safety Science Journal, 2018, 28(10): 25-30.]

[8] 张辉, 钱大琳, 邵春福, 等. 驾驶人分心状态判别支持向量机模型优化算法[J]. 交通运输系统工程与信息, 2018, 18(1): 127-132. [ZHANG H, QIAN D L, SHAO C F, et al. Identification of driver distraction states with optimized support vector machine method[J]. Journal of Transportation Systems Engineering and Information Technology, 2018, 18(1):127-132.]

[9] 孙剑, 张一豪, 王俊骅. 基于自然驾驶数据的分心驾驶行为识别方法[J]. 中国公路学报, 2020, 33(9): 225- 235. [SUN J, ZHANG Y H, WANG J Y. Detecting distraction behaviour of drivers using naturalistic driving data[J]. China Journal of Highway and Transport, 2020, 33(9): 225-235.]

[10] 周扬, 付锐, 袁伟, 等. 驾驶人认知分心识别随机森林模型研究[J]. 中国安全科学学报, 2018, 28(1): 20-25. [ZHOU Y, FU R, YUAN W, et al. Research on drivers' cognitive distracted recognition model based on random forest[J]. China Safety Science Journal, 2018, 28(1): 20- 25.]

[11] MASOOD S, RAI A, AGGARWAL A, et al. Detecting distraction of drivers using convolutional neural network [J]. Pattern Recognition Letters, 2020, 139: 79-85.

[12] ATIQUZZAMAN M, QI Y, FRIES R. Real-time detection of drivers' texting and eating behavior based on vehicle dynamics [J]. Transportation Research Part F: Traffic Psychology and Behaviour, 2018, 58: 594-604.

[13] TANGO F, BOTTA M. Real-time detection system of driver distraction using machine learning[J]. IEEE Transactions on Intelligent Transportation Systems, 2013, 14(2): 894-905.

[14] CHOUDHARY P, PAWAR N M, VELAGA N R, et al. Overall performance impairment and crash risk due to distracted driving: A comprehensive analysis using structural equation modelling[J]. Transportation Research Part F: Traffic Psychology and Behaviour, 2020, 74: 120-138.

[15] OVIEDO-TRESPALACIOS O, HAQUE M M, KING M, et al. Effects of road infrastructure and traffic complexity in speed adaptation behaviour of distracted drivers[J]. Accident Analysis and Prevention, 2017, 101: 67-77.

[16] CHOUDHARY P, VELAGA N R. Effects of phone use on driving performance: A comparative analysis of young and professional drivers[J]. Safety Science, 2019, 111: 179-187.

抵近信控交叉口分心驾驶识别模型

Recognition Model of Distracted Drivers When Approaching Signalized Intersections

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]	孙启鹏, 乔佳璐, 张锴琦, 孙佳. 基于社会网络结构特征的网约车平台竞争策略研究[J]. 交通运输系统工程与信息, 2022, 22(3): 1-6.
[2]	李岩, 曾明哲, 朱才华, 汪帆, 邓亚娟. 基于多点线圈联合数据的高速公路匝道影响范围识别[J]. 交通运输系统工程与信息, 2022, 22(3): 53-62.
[3]	周扬, 付锐, 刘卓凡. 考虑驾驶人不同分心类型的分心识别研究[J]. 交通运输系统工程与信息, 2022, 22(3): 132-139.
[4]	刘海玥, 蒋朝哲, 付川云, 周悦. 出租车超速行为道路因素随机系数空间模型构建[J]. 交通运输系统工程与信息, 2022, 22(3): 140-146.
[5]	孙超, 陆建. 共享单车出行空间异质性特征及驱动因素研究[J]. 交通运输系统工程与信息, 2022, 22(3): 198-206.
[6]	袁振洲, 胡嫣然, 杨洋. 考虑多维动态特征交互的高速公路实时事故风险建模[J]. 交通运输系统工程与信息, 2022, 22(3): 215-223.
[7]	潘义勇, 吴静婷, 施颖, 缪炫烨. 建成环境对老年人交通事故严重程度异质性影响[J]. 交通运输系统工程与信息, 2022, 22(3): 224-230.
[8]	戢晓峰, 詹换勤, 普永明, 覃文文. 山区公路穿村镇路段过境车辆事故严重程度推理分析[J]. 交通运输系统工程与信息, 2022, 22(3): 231-237.
[9]	李逸昕, 倪颖, 孙剑. 混合交通流交叉口共享空间交互行为建模及仿真[J]. 交通运输系统工程与信息, 2022, 22(3): 255-266.
[10]	帅斌, 米荣伟, 张锐, 雷渝. 考虑驾驶行为异质性的城市快速路合流区仿真模型[J]. 交通运输系统工程与信息, 2022, 22(3): 267-275.
[11]	倪捷, 张凯铎, 刘志强, 葛慧敏. 自适应期望跟车间距和行为习惯的驾驶人跟驰模型[J]. 交通运输系统工程与信息, 2022, 22(3): 286-292.
[12]	林建新, 刘博, 赵霞, 张蕾. 基于混合约束自编码器的机动车工况智能构建方法[J]. 交通运输系统工程与信息, 2022, 22(2): 109-116.
[13]	戢晓峰, 耿昭师. 山区双车道公路货车碰撞预测的双变量冲突极值模型[J]. 交通运输系统工程与信息, 2022, 22(2): 230-238.
[14]	张建波, 孙建平, 徐春玲, 郭镜霞, 温慧敏, 宋国华. 考虑交通运行条件影响的驾驶员特征聚类[J]. 交通运输系统工程与信息, 2022, 22(2): 330-336.
[15]	马景峰, 任刚, 李豪杰, 曹奇, 杜建玮. 电动自行车与机动车事故严重性影响因素分析[J]. 交通运输系统工程与信息, 2022, 22(2): 337-348.