重载货车驾驶人驾驶风格识别与量化研究

doi:10.16097/j.cnki.1009-6744.2022.04.016

交通运输系统工程与信息 ›› 2022, Vol. 22 ›› Issue (4): 137-148.DOI: 10.16097/j.cnki.1009-6744.2022.04.016

重载货车驾驶人驾驶风格识别与量化研究

覃文文^{1, 2}，鄢祺阳¹，谷金晶³，李武⁴，戢晓峰^{* 1, 2}

1. 昆明理工大学，交通工程学院，昆明 650504；2. 云南省现代物流工程研究中心，昆明 650504； 3. 同济大学，电子与信息工程学院，上海 201804；4. 大连理工大学，建设工程学部，辽宁大连 116024

收稿日期:2022-02-06 修回日期:2022-03-29 接受日期:2022-04-14 出版日期:2022-08-25 发布日期:2022-08-23
作者简介:覃文文(1986- )，男，广西柳州人，讲师，博士。
基金资助:
国家自然科学基金；云南省基础研究计划青年项目

Driving Style Recognition and Quantification for Heavy-duty Truck Drivers

QIN Wen-wen^{1, 2} , YAN Qi-yang¹ , GU Jin-jing³ , LI Wu⁴ , JI Xiao-feng^{* 1, 2}

1. Faculty of Transportation Engineering, Kunming University of Science and Technology, Kunming 650504, China; 2. Yunnan Modern Logistics Engineering Research Center, Kunming 650504, China; 3. College of Electronics and Information Engineering, Tongji University, Shanghai 201804, China; 4. Faculty of Infrastructure Engineering, Dalian University of Technology, Dalian 116024, Liaoning, China

Received:2022-02-06 Revised:2022-03-29 Accepted:2022-04-14 Online:2022-08-25 Published:2022-08-23
Supported by:
National Natural Science Foundation of China (52002161, 52062024)；Yunnan Fundamental Research Project (202101AU070166)。

摘要/Abstract

摘要： 重载货车驾驶人的激进驾驶风格具有强烈的习惯性特征和风险性特征，一旦养成很难矫正，且极易诱发交通事故。针对现有研究极少关注重载货车驾驶人驾驶风格的不足，本文基于某全国货运监管平台提供的云南省重载货车低频轨迹数据，从风格聚类、风格识别和风格评估这3个方面，提出综合考虑疲劳驾驶特征和超速驾驶特征的重载货车驾驶人驾驶风格分析方法。首先，基于轨迹数据蕴含驾驶人驾驶行为模式的特点，构建表征重载货车驾驶人驾驶风格的疲劳驾驶和超速驾驶特征集；其次，利用因子分析进行特征约简，并采用K-均值聚类方法划分重载货车驾驶人的驾驶风格；然后，构建基于支持向量机的驾驶风格识别模型，并与梯度提升决策树的识别结果进行对比；最后，基于疲劳驾驶特征和超速驾驶特征的累积分布，建立基于 CRITIC (Criteria Importance Though Intercriteria Correlation)赋权法的重载货车驾驶人驾驶风格量化评估模型。研究结果表明：经过特征约简，提取的疲劳因子和超速因子能综合反映上述两类特征集 80.838%的信息；根据疲劳因子和超速因子可将驾驶风格划分为4种类别，即稳健型、超速型、疲劳型和危险型，相应重载货车驾驶人比例依次为62.60%、25.02%、7.40%和4.98%；基于支持向量机的重载货车驾驶人驾驶风格识别模型对不同风格的识别准确率均大于97%，整体表现优于梯度提升决策树；基于CRITIC赋权法的驾驶风格评估模型能有效量化重载货车驾驶人的驾驶风格，其中稳健型驾驶人表现最好，75%以上的驾驶人风格评估总分高于60分；危险型驾驶人表现最差，75%以上的驾驶人风格评估总分低于20分。研究结果可为重载货车驾驶人不良驾驶行为的监测、干预和管理提供理论依据和技术支撑。

关键词: 交通工程, 驾驶风格识别, 支持向量机, 重载货车, 轨迹数据

Abstract: Aggressive driving styles of heavy-duty truck drivers may lead to serious traffic accidents with mass casualties, since aggressive styles have strong habitual characteristics and potential behavior risks, which are difficult to correct. However, most of the studies focus on driving style recognition of passenger cars rather than those of heavyduty trucks. In this paper, based on low- frequency trajectory data of heavy- duty trucks in Yunnan Province obtained from a national road freight platform, a framework of driving style analysis for heavy-duty truck drivers consideringfatigue driving and speeding driving characteristics is developed from the perspectives of style clustering, recognition, and evaluation. First, considering the driving behavior patterns conveyed by trajectory data, fatigue driving and speeding driving features were constructed to characterize the driving styles of heavy-duty truck drivers. Then, factor analysis for feature reduction was employed, and K-means clustering was used to classify the driving styles of heavyduty truck drivers. In addition, driving style recognition models based on support vector machine were constructed and compared with the recognition results of gradient boosting decision trees. Finally, based on the cumulative distributions of fatigue and speeding features, a quantitative assessment model for driving styles of heavy-duty truck drivers using the CRITIC (Criteria Importance Though Intercriteria Correlation) assignment method was proposed. The results show that the extracted fatigue factor and speeding factor can contain 80.838% information from the mentioned above two types of features when dimension reduction has been conducted. After that, according to the two types of features, the driving style can be divided into four categories: robust, speeding, fatigue, and dangerous. The proportion of the corresponding heavy-load truck drivers is 62.60%, 25.02%, 7.40% and 4.98%, respectively. Besides, the accuracy of the support vector machine-based driving style recognition model for heavy goods vehicle drivers is greater than 97% for all different styles, and the overall performance is better than the gradient boosting decision tree. Moreover, the proposed assessment model with the CRITIC assignment method can effectively quantify risk score for the driving style of each heavy- duty truck driver, in which the robust drivers performed the best in terms of more than 75% of drivers having a total style risk score higher than 60, whereas the risky drivers performed the worst with more than 75% of them scoring less than 20. This study could provide a theoretical basis and technical support for risky behavior monitoring, intervention, and management.

Key words: traffic engineering, driving style recognition, support vector machine, heavy-duty truck, trajectory data

中图分类号:

U491.31

覃文文, 鄢祺阳, 谷金晶, 李武, 戢晓峰. 重载货车驾驶人驾驶风格识别与量化研究[J]. 交通运输系统工程与信息, 2022, 22(4): 137-148.

QIN Wen-wen , YAN Qi-yang , GU Jin-jing , LI Wu , JI Xiao-feng. Driving Style Recognition and Quantification for Heavy-duty Truck Drivers[J]. Journal of Transportation Systems Engineering and Information Technology, 2022, 22(4): 137-148.

导出引用管理器 EndNote|Ris|BibTeX

链接本文: http://www.tseit.org.cn/CN/10.16097/j.cnki.1009-6744.2022.04.016

http://www.tseit.org.cn/CN/Y2022/V22/I4/137

参考文献

[1] MOHAMMADNAZAR A, ARVIN R, KHATTAK A J. Classifying travelers' driving style using basic safety messages generated by connected vehicles: Application of unsupervised machine learning[J]. Transportation Research Part C: Emerging Technologies, 2021, 122: 1- 18.

[2] TOLEDO T, MUSICANT O, LOTAN T. In-vehicle data recorders for monitoring and feedback ondrivers' behavior [J]. Transportation Research Part C: Emerging Technologies, 2008, 16(3): 320-331.

[3] MEHDIZADEH A, YAZDI M, CAI M, et al. Predicting unsafe driving risk among commercial truck drivers using machine learning: Lessons learned from the surveillance of 20 million driving miles[J]. Accident Analysis & Prevention, 2021, 159: 1-12.

[4] NIU S, UKKUSURI V. Risk assessment of commercial dangerous-goods truck drivers using geolocation data: Acase study in China [J]. Accident Analysis & Prevention, 2020, 137: 1-14.

[5] GSUL M, 胡予红, 周旋, 等. 道路交通运输安全发展报告(2017) [J]. 中国应急管理, 2018, 12(2): 48-58. [GSUL M, HU Y H, ZHOU X, et al. Road traffic safety development report(2017) [J]. China Emergency Management, 2018, 12(2): 48-58.]

[6] 侯海晶, 金立生, 关志伟, 等. 驾驶风格对驾驶行为的影响[J]. 中国公路学报, 2018, 31(4): 18-27. [HOU H J, JIN L S, GUAN Z W, et al. Effects of driving style on driver behavior[J]. China Journal of Highway Transportation, 2018, 31(4): 18-27.]

[7] 熊坚, 施锦浩, 万华森. 人车路综合风险场模型构建及驾驶风格评估[J]. 交通运输系统工程与信息, 2021, 21(6): 105-114. [XIONG J, SHI J H, WAN H S. Modeling of driver-vehicle-road integrated risk field and driving style assessment[J]. Journal of Transportation Systems Engineering and Information Technology, 2021, 21(6): 105-114.]

[8] 郝景贤, 余卓平, 赵治国, 等. 混合动力汽车驾驶风格识别的研究[J]. 汽车工程, 2017, 39(12): 1444-1450. [HAO J X, YU Z P, ZHAO Z G, et al. A Study on the driving style recognition of hybrid electric vehicle[J]. Automotive Engineering, 2017, 39(12): 1444-1450.]

[9] 徐婷, 张香, 张亚坤, 等. 基于 AdaBoost 算法的货车驾驶人安全倾向性分类[J]. 安全与环境学报, 2019, 19 (4): 1273-1281. [XU T, ZHUANG X, ZHANG Y K, et al. Truck driver safety tendency classification based on the AdaBoost algorithm[J]. Journal of Safety and Environment, 2019, 19(4): 1273-1281.]

[10] 朱冰, 蒋渊德, 邓伟文, 等. 基于KL散度的驾驶员驾驶习性非监督聚类[J]. 汽车工程, 2018, 40(11): 1317- 1323. [ZHU B, JIANG Y D, DENG W W, et al. Unsupervised clustering of driving styles based on KL divergence[J]. Automotive Engineering, 2018, 40 (11): 1317-1323.]

[11] FIGUEREDO G P, AGRAWAL U, MASE J M M, et al. Identifying heavy goods vehicle driving styles in the United Kingdom[J]. IEEE Transactions on Intelligent Transportation Systems, 2019, 20(9): 3324-3336.

[12] 孙川, 吴超仲, 褚端峰, 等. 基于车联网数据挖掘的营运车辆驾驶速度行为聚类研究[J]. 交通运输系统工程与信息, 2015, 15(6): 82-87. [SUN C, WU C Z, CHUD F, et al. Driving speed behavior clustering for commercial vehicle based on connected vehicle data mining[J]. Journal of Transportation Systems Engineering and Information Technology, 2015, 15(6): 82-87.]

[13] 王海星, 王翔宇, 王招贤, 等. 基于数据挖掘的危险货物运输风险驾驶行为聚类分析[J]. 交通运输系统工程与信息, 2020, 20(1): 183-189. [WANG H X, WANG X Y, WANG Z X, et al. Dangerous driving behavior clustering analysis for hazardous materials transportation based on data mining[J]. Journal of Transportation Systems Engineering and Information Technology, 2020, 20(1): 183-189.]

[14] ZHENG K, ZHENG Y, XIE X, et al. Reducing uncertainty of low-sampling-rate trajectories[C]// 28th International Conference on Data Engineering, Arlington, VA: IEEE, 2012: 1144-1155.

[15] 中华人民共和国交通运输部.道路运输车辆动态监督管理办法[EB/OL]. (2019-12-31) [2022-02-03]. http:// jtw. beijing. gov. cn/xxgk/flfg/zcfg/201912/t20191231_ 1547577.html. [Ministry of Transport of the People's Republic of China. Measures for dynamic supervision and management of road transport vehicles [EB/OL]. (2019- 12- 31) [2022- 02- 03]. http: //jtw.beijing.gov.cn/ xxgk/flfg/zcfg/201912/t20191231_1547577.html.]

[16] YU R, LONG X, LI J. Driving style analyses for carsharing users utilizing low-frequency trajectory data[C]// 5th International Conference on Transportation Information and Safety (ICTIS), Liverpool: IEEE, 2019: 927-933.

[17] WANG Q, ZHANG R, WANG Y, et al. Machine learningbased driving style identification of truck drivers in openpit mines[J]. Electronics, 2019, 9(1): 19-42.

[18] NADERI H, NASSIRI H, SAHEBI S. Assessing the relationship between heavy vehicle driver sleep problem and confirmed driver behavior measurement tools in Iran [J]. Transportation Research Part F: Traffic Psychology and Behaviour, 2018, 59: 57-66.

[19] 中华人民共和国国务院. 中华人民共和国道路交通安全法实施条例[EB/OL]. (2020-12-27). [2022-02- 03]. http: //www. gov. cn/ zhengce/ 2020-12/27/content_ 5574617.html. [The State Council of the People's Republic of China. Regulations on the implementation of the road traffic safety law of the People's Republic of China[EB/OL]. (2020-12-27). [2022-02-03] http://www. gov.cn/zhengce/2020-12/27/content_5574617.html.]

重载货车驾驶人驾驶风格识别与量化研究

Driving Style Recognition and Quantification for Heavy-duty Truck Drivers

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]	覃文文 , 李欢 , 李武 , 谷金晶 , 戢晓峰. 货车驾驶人驾驶行为与行车安全研究进展[J]. 交通运输系统工程与信息, 2022, 22(5): 55-74.
[2]	徐进 , 杨子邈 , 陈钦 , 陈正委. 基于电子不停车收费数据的山区高速公路车速分布与车型分类研究[J]. 交通运输系统工程与信息, 2022, 22(5): 75-84.
[3]	马庆禄, 牛圣平, 曾皓威, 段学锋. 网联环境下混合交通流偶发拥堵演化机理研究[J]. 交通运输系统工程与信息, 2022, 22(5): 97-106.
[4]	胡立伟, 吕一帆, 赵雪亭, 薛宇, 张成杰, 雷国庆, 刘凡. 基于数据驱动的交通事故伤害程度影响因素及其耦合关系研究[J]. 交通运输系统工程与信息, 2022, 22(5): 117-124.
[5]	聂庆慧 , 龙秀江 , 梁程 , 夏井新 , 欧吉顺. 多约束条件下城市道路环卫车优化配置与路径规划[J]. 交通运输系统工程与信息, 2022, 22(5): 273-284.
[6]	林培群 , 曾维嘉, 周楚昊, 庞崇浩. 新冠疫情防控对大湾区高速公路运输的影响分析[J]. 交通运输系统工程与信息, 2022, 22(5): 318-327.
[7]	焦朋朋 , 李汝鉴, 王健宇, 葛浩菁, 陈越. 考虑潜在类别的老年行人交通事故严重程度致因分析[J]. 交通运输系统工程与信息, 2022, 22(5): 328-336.
[8]	任其亮, 徐韬, 程龙春. 基于射频数据的道路交通流路径识别优化模型[J]. 交通运输系统工程与信息, 2022, 22(4): 89-95.
[9]	陈永恒, 李浩楠, 吴场建, 李婉宁. 基于元胞自动机的附加导流岛型出口仿真建模[J]. 交通运输系统工程与信息, 2022, 22(4): 96-105.
[10]	鲍琼, 屈琦凯, 唐涵润, 陈建明, 沈永俊. 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.
[11]	田顺, 郑博文, 孙健, 刘晶郁. 基于实车行驶工况的电动公交运营期碳足迹测算[J]. 交通运输系统工程与信息, 2022, 22(4): 149-157.
[12]	冯海霞, 王兴渝, 咸化彩, 刘新华, 李健, 宁二伟. 城市交通运行状况对机动车碳排放的影响研究[J]. 交通运输系统工程与信息, 2022, 22(4): 167-175.
[13]	陈政, 温惠英. 高速公路隧道临近段车辆换道持续时间生存分析[J]. 交通运输系统工程与信息, 2022, 22(4): 210-217.
[14]	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.
[15]	陈光会, 姬浩, 靳富强, 郭清娥, 贾斌, 苏兵. 考虑能源类型的城市出租车运营效率评价研究[J]. 交通运输系统工程与信息, 2022, 22(3): 36-44.