基于高斯混合隐马尔可夫模型的路怒攻击性驾驶行为辨识

doi:10.16097/j.cnki.1009-6744.2026.02.025

交通运输系统工程与信息 ›› 2026, Vol. 26 ›› Issue (2): 268-279.DOI: 10.16097/j.cnki.1009-6744.2026.02.025

基于高斯混合隐马尔可夫模型的路怒攻击性驾驶行为辨识

万平^*1，陈盈¹，邓鑫焰²，马晓凤³

1. 华东交通大学，交通运输工程学院，南昌330013；2.南昌交通学院，交通与土木工程学院，南昌330013； 3. 武汉理工大学，智能交通系统研究中心，武汉430063

收稿日期:2025-11-11 修回日期:2026-01-10 接受日期:2026-01-15 出版日期:2026-04-25 发布日期:2026-04-20
作者简介:万平（1984—），男，湖北黄冈人，副教授，博士。
基金资助:
江西省社会科学基金(23GL15)；江西省自然科学基金(20252BAC240100)。

Identification of Road Rage Aggressive Driving Behaviors Based on Gaussian Mixture-Hidden Markov Model

WAN Ping^*1, CHEN Ying¹, DENG Xinyan², MA Xiaofeng³

1. School of Transportation Engineering, East China Jiaotong University, Nanchang 330013, China; 2. School of Transportation and Civil Engineering, Nanchang Jiaotong Institute, Nanchang 330013, China; 3. Intelligent Transport Systems Research Center, Wuhan University of Technology, Wuhan 430063, China

Received:2025-11-11 Revised:2026-01-10 Accepted:2026-01-15 Online:2026-04-25 Published:2026-04-20
Supported by:
Social Science Foundation of Jiangxi Provincial, China (23GL15)；Natural Science Foundation of Jiangxi Province, China (20252BAC240100)。

摘要/Abstract

摘要： 为有效干预“路怒症”诱发的攻击性驾驶行为，本文提出一种基于车辆运动状态特征的路怒驾驶行为辨识模型。首先，在交通繁忙路段开展基于道路事件刺激的愤怒情绪诱导实验，获取攻击性与正常驾驶行为数据；其次，通过基于相对非约束最小二乘重要性拟合（RuLSIF）密度比估计的时间序列数据变点检测方法及简式情绪量表，完成正常驾驶行为与攻击性驾驶行为样本的标定，并对驾驶行为数据进行Welch'st检验分析，验证本文选取的速度、纵向加速度、横向加速度与航向角速度这4个特征在不同驾驶状态下具有显著差异性；最后，构建基于高斯混合-隐马尔可夫模型（GHMM）的路怒攻击性驾驶行为辨识模型，采用Viterbi 算法与Baum-Welch算法优化模型参数，通过合成少数类过采样技术-编辑最近邻规则（SMOTEENN）组合采样策略平衡样本，并引入似然比检验和阈值设定等方法提升模型辨识能力。研究结果表明，该模型整体准确率达84.17%，较逻辑回归（LR）模型和支持向量机（SVM）模型分别提升了6.90%和6.27%;特征分析验证其对横向加速度和航向角速度等关键特征的辨识效果显著。本文研究结果可为开发基于车辆运动状态特征的路怒攻击性驾驶行为检测预警系统提供理论支撑。

关键词: 智能交通, 驾驶行为辨识, 高斯混合-隐马尔可夫模型, 攻击性驾驶行为, 实车实验

Abstract: To effectively intervene in aggressive driving behaviors induced by road rage, this paper proposes a road rage driving behavior identification model considering vehicle motion state features. An anger emotion induction experiment was conducted based on road event stimuli on busy traffic sections to obtain data on aggressive and normal driving behaviors. Then, the samples of normal driving behavior and aggressive driving behavior were calibrated through the time series data change point detection method based on the Relative Unconstrained Least-Squares Importance Fitting (RuLSIF) density ratio estimation and the short form of the Simple Differential Emotions Scale (SDES). Welch's t-test analysis was conducted for the driving behavior data and verified that the four features, including speed, longitudinal acceleration, lateral acceleration, and yaw rate, have significant differences under different driving states. A road rage aggressive driving behavior identification model based on the Gaussian Mixture-Hidden Markov Model (GHMM) was then developed. The Viterbi algorithm and Baum-Welch algorithm were adopted to optimize the model parameters. The Synthetic Minority Over-sampling Technique combined with Edited Nearest Neighbors (SMOTE-ENN) hybrid sampling strategy was used to balance the samples. The likelihood ratio test and threshold setting methods were introduced to enhance the model's recognition ability. The results show that the model has an overall accuracy of 84.17%, which is respectively 6.90% and 6.27% higher than that of the Logistic Regression (LR) model and Support Vector Machine (SVM) model. Feature analysis verifies that the model shows significant identification effects on key features such as lateral acceleration and yaw rate. The research results provide theoretical support for the development of a road rage aggressive driving behavior detection and early warning system based on vehicle motion state features.

Key words: intelligent transportation, driving behavior identification, Gaussian Mixture-Hidden Markov Model (GHMM), aggressive driving behavior, on-road experiments

中图分类号:

U491.254

万平, 陈盈, 邓鑫焰, 马晓凤. 基于高斯混合隐马尔可夫模型的路怒攻击性驾驶行为辨识[J]. 交通运输系统工程与信息, 2026, 26(2): 268-279.

WAN Ping, CHEN Ying, DENG Xinyan, MA Xiaofeng. Identification of Road Rage Aggressive Driving Behaviors Based on Gaussian Mixture-Hidden Markov Model[J]. Journal of Transportation Systems Engineering and Information Technology, 2026, 26(2): 268-279.

导出引用管理器 EndNote|Ris|BibTeX

链接本文: http://www.tseit.org.cn/CN/10.16097/j.cnki.1009-6744.2026.02.025

http://www.tseit.org.cn/CN/Y2026/V26/I2/268

参考文献

[1]和圆圆.面向路怒症分析的表情识别算法研究[D].重庆: 重庆科技学院, 2023. [HE Y Y. Research on expression recognition algorithm for road rage analysis [D]. Chongqing: Chongqing University of Science and Technology, 2023.]

[2] 万平,吴超仲,马晓凤.基于ROC曲线和驾驶行为特征的驾驶愤怒强度判别阈值[J].吉林大学学报(工学版), 2020, 50(1): 121-131. [WAN P, WU C Z, MA X F. Discriminating threshold of driving anger intensity based on driving behavior features by ROC curve analysis[J]. Journal of Jilin University (Engineering and Technology Edition), 2020, 50(1): 121-131.]

[3]LIU K, SUN C, LI C. Research on abnormal driving behavior identification of freeway based on XGBoost method[M]// Proceedings of the 2023 COTA International Conference of Transportation Professionals (CICTP 2023), Reston, USA: American Society of Civil Engineers (ASCE), 2023: 284-289.

[4]WAN P, WU C Z, LIN Y Z, et al. Driving anger states detection based on incremental association Markov blanket and least square support vector machine [J]. Discrete Dynamics in Nature and Society, 2019, 2019: 1-17.

[5]ALKINANI, MONAGI H, WAZIR Z K, et al. Detecting human driver inattentive and aggressive driving behavior using deep learning: Recent advances, requirements and open challenges[J]. IEEE Access, 2020, 8: 105008 105030.

[6]赵朔,奇格奇,李培豪,等.基于脑电通道注意力机制的驾驶行为识别研究[J].交通运输系统工程与信息, 2024, 24(4): 283-291. [ZHAO S, QI G Q, LI P H, et al. Driving behavior recognition based on EEG channel attention mechanism[J]. Journal of Transportation Systems Engineering and Information Technology, 2024, 24(4): 283-291.]

[7] DENG Q, SOFFKER D. A review of HMM-based approaches of driving behaviors recognition and prediction[J]. IEEE Transactions on Intelligent Vehicles, 2021, 7(1): 21-31.

[8] 罗强,刘绍鎏,罗诗琦,等.基于混合高斯-隐马尔可夫模型的驾驶意图识别方法研究[J].交通工程,2024,24 (9): 29-33. [LUO Q, LIU S L, LUO S Q, et al. Research on driving intention recognition method based on hybrid Gaussian-Hidden Markov model[J]. Journal of Traffic Engineering, 2024, 24(9): 29-33.]

[9] 万平,邱港澳,陈培栋,等. 极端通勤群体驾驶愤怒表达与诱因关联模型[J].交通信息与安全,2024, 42(4): 154-163. [WAN P, QIU G A, CHEN P D, et al. Correlation model of driving anger expression and inducing factors among extreme commuters[J]. Journal of Traffic Information and Safety, 2024, 42(4): 154-163.]

[10] 万平, 吴超仲,林英姿,等.基于置信规则库的驾驶人愤怒情绪识别模型[J]. 交通运输系统工程与信息, 2015, 15(5): 96-102. [WAN P, WU C Z, LIN Y Z, et al. A recognition model of driving anger based on belief rule base[J]. Journal of Transportation Systems Engineering and Information Technology, 2015, 15(5): 96-102.]

[11] SHINAR D. Aggressive driving: The contribution of the drivers and the situation[J]. Transportation Research Part F: Traffic Psychology and Behaviour, 1998, 1(2): 137 160.

[12] VANLAAR W, SIMPSON H, MAYHEW D, et al. Aggressive driving: A survey of attitudes, opinions and behaviors[J]. Journal of Safety Research, 2008, 39(4): 375-381.

[13] MAKOTO Y, TAIJI S, TAKAFUMI K, et al. Relative density-ratio estimation for robust distribution comparison[J]. Neural Computation, 2013, 25(5): 1324 1370.

[14] 彭娜. 基于行车状态数据的激进型驾驶行为检测[D]. 西安: 长安大学, 2019. [PENG N. Aggressive driving behavior detection based on driving status data[D]. Xi'an: Chang'an University, 2019.]

[15] HONG J H, MARGINES B, DEY A K. A smartphone based sensing platform to model aggressive driving behaviors[C]//Proceedings of the SIGCHI Conference on Human Factors in Computing Systems, New York, USA: Association for Computing Machinery (ACM), 2014: 4047-4056.

[16] ZHANG X, ZHAO X, RONG J. A study of individual characteristics of driving behavior based on hidden Markov model[J]. Sensors & Transducers, 2014, 167(3): 194.

基于高斯混合隐马尔可夫模型的路怒攻击性驾驶行为辨识

Identification of Road Rage Aggressive Driving Behaviors Based on Gaussian Mixture-Hidden Markov Model

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]	王亦兵, 陈安妮, PAPAGEORGIOU Markos, 余宏鑫, 郭静秋, 章立辉. 智能网联无车道线城市道路内边界与交通信号协同控制[J]. 交通运输系统工程与信息, 2026, 26(2): 81-90.
[2]	谷远利, 宇泓儒, 陈龙, 邓社军, 陆文琦. 网联自动驾驶车辆专用车道动态宏微观协同部署方法[J]. 交通运输系统工程与信息, 2026, 26(2): 125-136.
[3]	孙健, 纪裕伟, 于珂伟, 李子豪, 赵昱霖. 融合双注意力机制的快速路协同深度强化学习方法[J]. 交通运输系统工程与信息, 2026, 26(2): 137-147.
[4]	刘涛, 李林. 融合贝叶斯优化与深度学习的机场公交专线短时客流预测[J]. 交通运输系统工程与信息, 2026, 26(2): 300-308.
[5]	王福建, 马佳豪, 李廷浩, 马东方. 混合交通环境下基于动态决策间隔的强化学习信号控制方法[J]. 交通运输系统工程与信息, 2026, 26(1): 45-54.
[6]	蒋贤才, 曲悦, 魏贺迪. 人机混合驾驶协同合流的多智能体近端策略优化算法[J]. 交通运输系统工程与信息, 2026, 26(1): 65-75.
[7]	谢金苹, 崔洪军, 朱敏清, 马新卫. 不同共乘参与程度下异构自动驾驶车队规模研究[J]. 交通运输系统工程与信息, 2026, 26(1): 76-89.
[8]	甘佐贤, 刘雅新, 秦严严. 面向混合交通流的逆向可变车道左转车辆轨迹优化控制[J]. 交通运输系统工程与信息, 2026, 26(1): 90-103.
[9]	李熙莹, 陈泽, 李锦, 刘静宇, 盘婳燕, 江倩殷. 融合时序感知与边界损失的车辆轨迹重构方法[J]. 交通运输系统工程与信息, 2026, 26(1): 104-114.
[10]	马健霄, 王雨, 陆涛, 白莹佳, 王羽尘, 赵顗. 融合多源势场的公路作业区智能驾驶动态路径规划[J]. 交通运输系统工程与信息, 2026, 26(1): 115-124.
[11]	尚婷, 胥浩, 毛慧涵, 何军. 低能见度下智能汽车双因子避障轨迹优化研究[J]. 交通运输系统工程与信息, 2026, 26(1): 125-134.
[12]	单肖年, 胡颖, 成嘉琪, 田大新. 网联混行下城市干道动态公交专用道管控策略[J]. 交通运输系统工程与信息, 2026, 26(1): 135-147.
[13]	谭一帆, 唐瑞雪, 姚志洪, 蒲云. 稀疏轨迹下结构-行为联合建模的生成式路径推理[J]. 交通运输系统工程与信息, 2026, 26(1): 252-260.
[14]	缪鸿志, 王俊朋, 吴佳雨, 李歆蔚, 郑建风. 考虑交通-电力两网交互的电动集卡超充设施规划方法[J]. 交通运输系统工程与信息, 2026, 26(1): 305-317.
[15]	邹复民, 陈培烨, 蔡祈钦, 廖律超, 罗永煜. 基于两级自适应空间建模的多视距高速公路交通流预测[J]. 交通运输系统工程与信息, 2026, 26(1): 318-328.