低成本高精度驾驶分神类型与程度分层识别方法

doi:10.16097/j.cnki.1009-6744.2026.03.035

交通运输系统工程与信息 ›› 2026, Vol. 26 ›› Issue (3): 393-402.DOI: 10.16097/j.cnki.1009-6744.2026.03.035

• 系统工程理论与方法 • 上一篇

低成本高精度驾驶分神类型与程度分层识别方法

李鹏辉^1,2，逯航天¹，常乃心¹ ，郭戎格^1,2 ，马彬³，董春娇^*1

1. 北京交通大学，交通运输学院，北京100044；2.道路交通安全管控技术国家工程研究中心，北京102600； 3. 北京信息科技大学，机电工程学院，北京100192

收稿日期:2025-11-30 修回日期:2026-03-19 接受日期:2026-03-30 出版日期:2026-06-25 发布日期:2026-06-24
作者简介:李鹏辉（1991—），男，湖北孝感人，副教授。
基金资助:
国家自然科学基金 (52302425)；道路交通安全管控技术国家工程研究中心开放课题 (2025GCZXKFKT07A)。

A Low-Cost and High-Precision Method for Hierarchical Recognition of Driver Distraction Types and Levels

LI Penghui^1,2, LU Hangtian¹, CHANG Naixin¹, GUO Rongge^1,2, MA Bin³, DONG Chunjiao^*1

1. School of Traffic and Transportation, Beijing Jiaotong University, Beijing 100044, China; 2. National Engineering Research Center for Road Traffic Safety Control Technology, Ministry of Public Security, Beijing 102600, China; 3. School of Mechanical and Electrical Engineering, Beijing Information Science and Technology University, Beijing 100192, China

Received:2025-11-30 Revised:2026-03-19 Accepted:2026-03-30 Online:2026-06-25 Published:2026-06-24
Supported by:
National Natural Science Foundation of China (52302425)；Open Fund Project of the National Engineering Research Center for RoadTraffic Safety Management and Control Technologies (2025GCZXKFKT07A)。

摘要/Abstract

摘要： 针对现有驾驶分神识别方法存在的精度与成本兼顾难、分神程度识别能力弱的问题，本文提出基于麻雀搜索优化XGBoost算法的驾驶员分神类型与程度分层识别方法。首先，系统性提取视觉、认知分神驾驶下驾驶行为、眼动与生理等46项特征指标，采用重复测量方差分析与效应量比较遴选视觉、认知分神高灵敏度指标，构建面向驾驶分神类型与程度识别的高灵敏性指标集。其次，采用递归特征消除算法对驾驶性能、视觉及生理等多源指标进行二次优选，构建仅涵盖驾驶行为指标的单一特征集与涵盖驾驶行为、眼动、生理多源指标的多源特征集。而后，建立驾驶分神类别与程度分层识别框架，采用麻雀搜索算法对XGBoost模型进行超参数寻优构建各层分类器。最后，系统性对比单一特征集与多源特征集作为输入时模型的识别效果，构建低成本高精度的驾驶分神类型与程度识别模型。结果表明，在分神类型识别上，单一特征集与多源特征集模型的准确率分别达到了92.4%和95.9%，均表现出优异性能；此外，经麻雀搜索算法优化的分神类型识别模型相比于支持向量机、随机森林与常规XGBoost等模型准确率分别提升4%~6%、11%、7%。在分神程度识别上，多源特征集模型对认知与视觉分神程度识别准确率较高，分别达到了81.0%和84.9%，而单一特征集模型识别准确率较低。因此，推荐采用单一特征集模型实现分神类型的低成本识别，采用多源特征集模型实现分神程度的高精度识别。所提出的方法可实现驾驶分神类型与程度的低成本高精度识别，可为不同应用场景下的驾驶员状态监测系统开发提供理论支撑。

关键词: 交通工程, 分神识别, 麻雀搜索算法, 驾驶分神, 类别, 程度

Abstract: To improve the existing detection methods for driver distraction in balancing accuracy with cost and identifying distraction levels, this paper proposes a hierarchical recognition method using XGBoost and Sparrow Search Algorithm. The study extracted 46 features from driving behavior, eye movement, and physiological data during visual and cognitive distraction. Highly sensitive indicators for distraction type and degree were selected through repeated-measures Analysis of Variance (ANOVA) and effect size comparisons to build a high-sensitivity feature set. A recursive feature elimination algorithm was then applied for secondary feature selection from multi-source indicators, resulting in two distinct sets: a single-source feature set containing only driving behavior features and a multi-source feature set integrating driving behavior, eye movement, and physiological features. A hierarchical recognition framework was established, employing the Sparrow Search Algorithm to optimize hyperparameters of the XGBoost classifiers at each level. The study also evaluated the model performance using the single-source and multi-source feature sets as input to develop a low-cost, high-accuracy model. Results demonstrate that for distraction type recognition, the single-source and multi-source models achieved accuracies of 92.4% and 95.9%, respectively. The Sparrow Search Algorithm optimized model for type recognition consistently outperformed Support Vector Machine, Random Forest, and standard XGBoost models, with accuracy improvements of 4%~6%, 11%, and 7%. For distraction degree recognition, the multi-source model achieved the accuracies of 81.0% for cognitive and 84.9% for visual distraction, while the single-source model showed relatively poor performance. Therefore, the single-source model is recommended for low-cost type recognition, and the multi-source model for high-precision degree assessment. The proposed method enables cost-effective and accurate identification of driver distraction, providing theoretical support for developing context-aware driver state monitoring systems.

Key words: traffic engineering, distraction recognition, sparrow search algorithm, driver distraction, type, degree

中图分类号:

U491

李鹏辉, 逯航天, 常乃心, 郭戎格, 马彬, 董春娇. 低成本高精度驾驶分神类型与程度分层识别方法[J]. 交通运输系统工程与信息, 2026, 26(3): 393-402.

LI Penghui, LU Hangtian, CHANG Naixin, GUO Rongge, MA Bin, DONG Chunjiao. A Low-Cost and High-Precision Method for Hierarchical Recognition of Driver Distraction Types and Levels[J]. Journal of Transportation Systems Engineering and Information Technology, 2026, 26(3): 393-402.

导出引用管理器 EndNote|Ris|BibTeX

链接本文: http://www.tseit.org.cn/CN/10.16097/j.cnki.1009-6744.2026.03.035

http://www.tseit.org.cn/CN/Y2026/V26/I3/393

参考文献

[1] NationalCenter for Statistics andAnalysis.Overviewof motor vehicle traffic crashes in 2022[R].Washington, DC: National Highway Traffic Safety Administration, 2024.

[2] CHANGNC, DONGC J, ZHUS F, et al. Towards recognizingcognitivedistractionlevelswithlow-costand high-sensitive measures: The effectiveness of sample, approximate, and traditional steering entropies[J]. TransportationResearchPart F: TrafficPsychologyand Behaviour,2025,109:1063-1079.

[3]REIMER B, MEHLER B, WANG Y, et al. A field study on the impact of variations in short-term memory demands on drivers' visual attention and driving performance across three age groups[J]. Human Factors, 2012, 54(3): 454-468.

[4]LI P H, MARKKULA G, LI Y, et al. Is improved lane keeping during cognitive load caused by increased physical arousal or gaze concentration toward the road center?[J]. Accident Analysis and Prevention, 2018, 117: 65-74.

[5]MEHLER B, REIMER B, COUGHLIN J F. Sensitivity of physiological measures for detecting systematic variations in cognitive demand from a working memory task: An on-road study across three age groups[J]. Human Factors, 2012, 54(3): 396-412.

[6]JIN L S, NIU Q N, HOU H J, et al. Driver cognitive distraction detection using driving performance measures [J]. Discrete Dynamics in Nature and Society, 2012, 2012 (1): 432634.

[7]MIYAJI M, KAWANAKA H, OGURI K. Effect of pattern recognition features on detection for driver's cognitive distraction[C]//13th International IEEE Conference on Intelligent Transportation Systems, New York: IEEE, 2010: 605-610.

[8] 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. [ZHOU Y, FU R, LIU Z F. Research on distraction recognition considering different distraction types of drivers[J]. Journal of Transportation Systems Engineering and Information Technology, 2022, 22(3): 132-139.]

[9] SHAJARI A, ASADI H, ALSANWY S, et al. Detection of driver cognitive distraction using driver performance measures, eye-tracking data and a D-FFNN model[C]// 2023 IEEE International Conference on Systems, Man, and Cybernetics (SMC), New York: IEEE, 2023: 2093 2099.

[10] CHEN Y, BOUKERCHE A. A novel lane departure warning system for improving road safety[C]//2020 IEEE International Conference on Communications (ICC), Dublin: IEEE, 2020: 1-6.

[11] 李鹏辉.驾驶分神对行车安全影响的建模研究[D].北京: 清华大学, 2020. [LI P H. Modeling study on the impact of driver distraction on driving safety[D]. Beijing: Tsinghua University, 2020.] [12] LI P H, MERAT N, ZHENG Z X, et al. Does cognitive distraction improve or degrade lane keeping performance? Analysis of time-to-line crossing safety margins[J]. Transportation Research Part F: Traffic Psychology and Behaviour, 2018, 57: 48-58.

[13] LI P H, HU M X, CHEN F, et al. Improve adaptive cruise control under driver distraction: Time headway compensation based on a random-effect crash risk model [J]. IET Intelligent Transport Systems, 2021, 15(10): 1241-1255.

[14] COHEN J. Statistical power analysis for the behavioral sciences(2nd ED)[M]. New York: Routledge, 2013.

[15] 薛建凯. 一种新型的群智能优化技术的研究与应用: 麻雀搜索算法[D]. 上海: 东华大学,2019. [XUE J K. Research and application of a new swarm intelligence optimization technique: Sparrow search algorithm[D]. Shanghai: Donghua University, 2019.]

低成本高精度驾驶分神类型与程度分层识别方法

A Low-Cost and High-Precision Method for Hierarchical Recognition of Driver Distraction Types and Levels

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]	焦方通, 郭佩佩, 王殿海, 杜志刚, 国洪奇, 孙锋. 隧道地下合流区驾驶人视觉特性及其对行车安全的影响[J]. 交通运输系统工程与信息, 2026, 26(3): 93-102.
[2]	张琦, 张思雨, 李得伟. 应用虚拟现实技术的行人交通研究综述[J]. 交通运输系统工程与信息, 2026, 26(3): 176-191.
[3]	王笑蓉, 李引珍, 孙颖杰, 县勇, 李雯, 巨玉祥. 不确定扰动下的区域道路交通网络韧性演化及提升方法研究[J]. 交通运输系统工程与信息, 2026, 26(2): 47-59.
[4]	向往, 唐司杨, 王云, 薛晴婉. 动态引导与语音对高速收费站前车辆主线换道行为的影响[J]. 交通运输系统工程与信息, 2026, 26(2): 102-112.
[5]	吴焕, 曹成涛, 林晓辉, 吴璠. 基于改进人口合成方法的居民出行调查数据扩样[J]. 交通运输系统工程与信息, 2026, 26(2): 342-353.
[6]	杨阳, 卫倩, 于谦, 吴芳. 短行程时长阈值对行驶工况构建的影响研究[J]. 交通运输系统工程与信息, 2026, 26(2): 354-362.
[7]	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.
[8]	杜太升, 彭正鍾, 张源凯, 田琼, 蒋晓桐. 考虑乘客偏好的需求响应定制公交线路优化[J]. 交通运输系统工程与信息, 2026, 26(1): 217-227.
[9]	杨洋, 陈冠华, 王明涛, 黄海博. 考虑数据平衡影响的道路交通事故建模与致因分析[J]. 交通运输系统工程与信息, 2026, 26(1): 261-269.
[10]	郑展骥, 吴程宇, 王振科, 饶嘉强, 凃强, 徐进. 城市地下道路接入口驾驶人眼动特征影响机制研究[J]. 交通运输系统工程与信息, 2026, 26(1): 329-339.
[11]	罗霜, 陈宽, 李善兴, 徐进. 基于实车试验的小净距“隧道-互通”心理负荷量化分析[J]. 交通运输系统工程与信息, 2025, 25(6): 341-349.
[12]	韩宝睿, 纪宇轩, 李根, 杨政, 颜荣添, 徐圣睿. 基于粘滞性车辆组团识别的道路偶发拥堵预警研究[J]. 交通运输系统工程与信息, 2025, 25(5): 91-102.
[13]	张文会, 乔梓凡, 陈德启. 出租车轨迹数据驱动的充电站选址定容方法[J]. 交通运输系统工程与信息, 2025, 25(5): 291-301.
[14]	郑立勇, 孙剑, 饶红玉, 邵健轩, 赵威, 郝勇刚. 基于物理信息深度学习的交叉口车辆轨迹补全方法[J]. 交通运输系统工程与信息, 2025, 25(4): 116-125.
[15]	荣建, 吴培佳, 高亚聪, 王益, 窦灏. 融合仿真与机器学习的交织区通行能力协同估计方法[J]. 交通运输系统工程与信息, 2025, 25(4): 206-218.