College of Energy and Traffic Engineering, Inner Mongolia Agricultural University, Hohhot 010018, China
Show less
文章历史+
Received
Published
Revised
2026-03-06
2026-11-08
2026-04-21
Online Date
2026-11-19
PDF (10796K)
摘要
针对典型草原公路长直路段施工作业区引发驾驶人工作负荷瞬时跃迁的安全风险,构建一种多模态数据驱动的工作负荷识别模型,以提升非常规场景下的行车安全监测能力。选取典型草原公路长直路段,布设内侧车道封闭与外侧路肩封闭两种施工场景开展驾驶模拟实验,同步采集眼动(eye movement, EM)、心电(electrocardiogram, ECG)及车辆操作与运行(vehicle operation and control, VOC)等多模态数据,并结合NASA(National Aeronautics and Space Administration)任务负荷指数(NASA task load index, NASA-TLX)量表获取主观负荷评分作为监督标签。通过相关性分析与因子分析筛选核心特征,并结合生理有效性校验剔除不适配短时间窗的频域指标,筛选出12 项核心指标构建特征体系。基于贝叶斯优化算法对5 种机器学习模型进行参数寻优,对比分析不同模型及特征组合的识别性能。结果显示,极端梯度提升(extreme gradient boosting, XGBoost)模型在草原公路施工作业区场景下表现最优,分类准确率达91.60%,受试者工作特征曲线下面积(receiver operating characteristic-area under curve, ROC-AUC)值为0.948,显著优于随机森林(random forest, RF)与支持向量机(support vector machine, SVM)等模型。特征组合对比发现,EM-ECG-VOC全维度特征融合的识别效能最高;而在双模态组合中,心电与车辆操作组合(ECG-VOC)表现优于包含眼动指标的组合。研究表明,多模态特征融合结合优化的集成学习算法可有效识别草原公路施工区驾驶人认知状态的瞬时变化,可为高风险路段的主动安全预警系统开发与自动驾驶接管策略优化提供理论依据与算法支撑。
Abstract
To address the safety risks associated with the instantaneous transition of driver workload triggered by work zones on typical long straight sections of prairie highways, a multimodal data-driven workload recognition model was developed to enhance traffic safety monitoring capabilities in unconventional scenarios. Representative long straight sections of prairie highways were selected to conduct driving simulation experiments, featuring two specific work zone scenarios: inner lane closure and outer shoulder closure. Multimodal data, including eye movement (EM), electrocardiogram (ECG), and vehicle operation and control (VOC), were synchronously collected. Meanwhile, subjective workload scores were obtained as supervised labels using the NASA(National Aeronautics and Space Administration) task load index (NASA-TLX). Core features were identified through correlation and factor analysis. Frequency-domain indicators unsuitable for short time windows were excluded via physiological validity verification, resulting in a refined feature system comprising 12 core indicators. Five machine learning models were parameterized and optimized using a Bayesian optimization algorithm, followed by a comparative analysis of recognition performance across different models and feature combinations. The results indicate that the extreme gradient boosting (XGBoost) model exhibited superior performance in prairie highway work zone scenarios, achieving a classification accuracy of 91.60% and an Area Under the receiver operating characteristic curve (ROC-AUC) of 0.948, which significantly outperformed random forest (RF) and support vector machine (SVM) models. Regarding feature combinations, the fusion of all dimensions (EM-ECG-VOC) yielded the highest recognition efficiency. Among dual-modal combinations, the ECG-VOC combination demonstrated better performance than those incorporating eye movement indicators. It is concluded that the fusion of multimodal features combined with optimized ensemble learning algorithms can effectively identify instantaneous fluctuations in driver workload within prairie highway work zones. The result provides theoretical evidence and algorithmic support for the development of active safety warning systems and the optimization of autonomous driving takeover strategies for high-risk road sections.
LÜ Zhen, YANG Xin-yu, GUO Chen.
Driver Workload Recognition in Prairie Highway Work Zones Based on Multimodal Data[J].
Science Technology and Engineering, 2026, 26(31): 13711-13723 DOI:10.12404/j.issn.1671-1815.2601173
MaJ, WuY P, RongJ Z, et al. A systematic review on the influence factors, measurement, and effect of driver workload[J]. Accident Analysis and Prevention, 2023, 192: 1-17.
[2]
MaS W, YanX D, BillingtonJ, et al. Cognitive load during driving: EEG microstate metrics are sensitive to task difficulty and predict safety outcomes[J]. Accident Analysis and Prevention, 2024, 207: 107769.
[3]
TarabayR, AbouZeidM. A dynamic hybrid choice model to quantify stress in a simulated driving environment[J]. IEEE Transactions on Intelligent Transportation Systems, 2021, 22(9): 5624-5635.
[4]
ZhaoC, ZhangS Y, ZhangZ R, et al. Deriving workload from driving behavior and psycho-physiology in work zones[J]. Accident Analysis and Prevention, 2025, 222: 108240.
[5]
YangL, ZhouR, SunY, et al. Recognizing and explaining driving stress using a Shapley additive explanation model by fusing EEG and behavior signals[J]. Accident Analysis and Prevention, 2024, 209: 107835.
[6]
LüL, ShengY T, SongC C, et al. Driving safety assurance method in work zone crossovers of highway reconstruction and expansion project[J]. Journal of Advanced Transportation, 2021, 12: 1214512.
[7]
KummethaV C, KondyliA, ChrysikouE G, et al. Safety analysis of work zone complexity with respect to driver characteristics-a simulator study employing performance and gaze measures[J]. Accident Analysis and Prevention, 2020, 142: 105566.
[8]
TorreF L, DomenichiniL, NocentiniA. Effects of stationary work zones on motorway crashes[J]. Safety Science, 2017, 92: 148-159.
[9]
HangJ Y, LiX M, YanX D, et al. Driving risks assessment and in-vehicle warning design for improving work zone safety[J]. Accident Analysis & Prevention, 2025, 215: 107991.
[10]
WangB, ZhangC, ChenT, et al. Toward safer highway work zones: an empirical analysis of crash risks using improved safety potential field and machine learning techniques[J]. Accident Analysis and Prevention, 2024, 194: 1-15.
[11]
交通运输部. 2022年交通运输行业发展统计公报[R]. 北京: 交通运输部, 2023.
[12]
Ministry of Transport (MOT). 2022 statistical communiqué on transportation industry development[R/OL]. Beijing: MOT, 2023.
[13]
LioiA, PorteraA, TefaL, et al. How cognitive distraction affects motorway short-term work zone safety along curves: a driving simulation study[J]. Journal of Transportation Safety & Security, 2025, 17(5): 489-506.
[14]
ThapaD, MishraS, AdeelK M. Assessing driver behavior in work zones: a discretized duration approach to predict speeding[J]. Accident Analysis and Prevention, 2024, 196: 107427.
[15]
ShakouriM, IkumaL H, AghazadehF, et al. Analysis of the sensitivity of heart rate variability and subjective workload measures in a driving simulator: the case of highway work zones[J]. International Journal of Industrial Ergonomics, 2018, 66: 136-145.
[16]
ZhangC, MaY, ChenS, et al. Exploring the occupational fatigue risk of short-haul truck drivers: effects of sleep pattern, driving task, and time-on-task on driving behavior and eye-motion metrics[J]. Transportation Research, Part F Traffic Psychology and Behaviour, 2024, 100(1): 37-56.
[17]
VicenteK J, ThorntonD C, MorayN. Spectral analysis of sinus arrhythmia: a measure of mental effort[J]. Human Factors: The Journal of the Human Factors and Ergonomics Society, 1987, 29(2): 171-182.
[18]
VosughN, BahmaniZ, MohammadianA. Distracted driving recognition based on functional connectivity analysis between physiological signals and perinasal perspiration index[J]. Expert Systems with Applications, 2023, 231: 120707.
[19]
YunJ, JangJ, JungA, et al. Evaluating driving safety of road alignment conditions by predicted driver's psychophysiological workload using vehicle maneuvering data[J]. Transportation Research Record, 2024, 2678(5): 479-490.
[20]
LeeB G, ChungW Y. Wearable glove-type driver stress detection using a motion sensor[J]. IEEE Transactions on Intelligent Transportation Systems, 2017, 18(11): 3122-3131.
[21]
RigasG, GoletsisY, FotiadisD I. Real-time driver's stress event detection[J]. IEEE Transactions on Intelligent Transportation Systems, 2012, 13(1): 221-234.
[22]
IdewuW A, WolshonB. Joint merge and its impact on merging speeds in lane reduction areas of construction zone[J]. Transportation Research Record Journal of the Transportation Research Board, 2010, 2169(1): 31-39.
[23]
NahedR, NassarE, KhouryJ, et al. Assessing the effects of geometric layout and signing on drivers' behavior through work zones[J]. Transportation Research Interdisciplinary Perspectives, 2023, 21: 100901.
LiuYanling, ChengChong, ZhangShi, et al. Mental workload of drivers at curve section of interchange ramp based on natural driving data[J]. Science Technology and Engineering, 2025, 25(16): 6942-6952.
GuoChen, LüZhen, WangHaixiao, et al. Mental workload of drivers in grassland highway construction zones[J]. Journal of Inner Mongolia Agricultural University (Natural Science Edition), 2026, 47(2): 74-85.
LüZhen, QiChunhua, ZhuShoulin. The mental load of drivers on the prairie highway based on the eye movement behavior[J]. Science Technology and Engineering, 2021, 21(21): 9160-9167.
LiHangtian, YangFeng, LiuYuxin, et al. Analysis of the impact of information volume on guide signs at grassland highway intersections on driver cognitive characteristics[J]. Science Technology and Engineering, 2023, 23(21): 9296-9305.
[32]
ZhangS Y, ZhaoC, ZhangZ R, et al. Driving simulator validation studies: a systematic review[J]. Simulation Modelling Practice and Theory, 2025, 138: 102813.
LiuZhao, XiongJian, GuoFengxiang. Validation of speed effectiveness of driving simulator under different horizontal alignments[J]. Journal of East China Jiaotong University, 2012, 29(4): 69-72.
[35]
GroegerJ A, MurphyG. Driver performance under simulated and actual driving conditions: validity and orthogonality[J]. Accident Analysis & Prevention, 2020, 143: 105593.
[36]
赵婷. 草原道路简单景观环境下驾驶员疲劳特性研究[D]. 呼和浩特: 内蒙古农业大学, 2017.
[37]
ZhaoTing. Study on drivers' fatigue characteristics in simple landscape environments of prairie roads[D]. Hohhot: Inner Mongolia Agricultural University, 2017.
[38]
LedolterJ, KardonR H. Focus on data: statistical design of experiments and sample size selection using power analysis[J]. Investigative Ophthalmology & Visual Science, 2020, 61(8): 11.
HuoYueying. Study on the maximum straight length of prairie highways based on drivers' psychophysiological responses[D]. Hohhot: Inner Mongolia Agricultural University, 2010.
[41]
交通运输部公路科学研究院. 公路养护安全作业规程[S]. 北京: 人民交通出版社, 2015.
[42]
Research Institute of Highway Ministry of Transport. Safety rules for highway maintenance operation[S]. Beijing: China Communications Press, 2015.
WangJian, ChenYazhen, LiuXiaofei, et al. Research on the setting and optimization of traffic signs in expressway reconstruction and extension construction zones[J]. Journal of Highway and Transportation Research and Development: Applied Technology Edition, 2016, 12(10): 208-210.
[45]
Ezzati AminiR, Al HaddadC, BatabyalT, et al. Driver distraction and in-vehicle interventions: a driving simulator study on visual attention and driving performance[J]. Accident Analysis and Prevention, 2023, 191: 107195.
[46]
ZhangN, YangC, FardM. Measured increases in steering entropy may predict when performance will degrade: a driving simulator study[J]. Transportation Research, Part F. Traffic Psychology and Behaviour, 2022, 91(11): 87-94.
[47]
HartS G, StavelandL E. Development of NASA-TLX (Task Load Index): results of empirical and theoretical research[J]. Advances in Psychology, 1988, 52(6): 139-183.
[48]
GrierR A. How high is high? a meta-analysis of NASA-TLX global workload scores[J]. Proceedings of the Human Factors & Ergonomics Society Annual Meeting, 2015, 59(1): 1727-1731.
[49]
WeiW Y, FuX S, ZhongS Y, et al. Driver's mental workload classification using physiological, traffic flow and environmental factors[J]. Transportation Research, Part F Traffic Psychology and Behaviour, 2023, 94(4): 151-169.
[50]
MeteierQ, CapalleraM, RuffieuxS, et al. Classification of drivers' workload using physiological signals in conditional automation[J]. Frontiers in Psychology, 2021, 12: 596038.
[51]
BucchiF, BartolozziC. How the autonomic nervous system and driving style change with incremental stressing conditions during simulated driving[J]. IEEE Transactions on Intelligent Transportation Systems, 2015, 16(3): 1505-1517.
[52]
LiuT, ZhouZ Y, ChaiC, et al. Non-instinct detection of cellphone usage from lane-keeping performance based on eXtreme gradient boosting and optimal sliding windows[J]. IET Intelligent Transport Systems, 2022, 2022(16): 1600-1610.
[53]
RastgooM N, NakisaB, MaireF, et al. Automatic driver stress level classification using multimodal deep learning[J]. Expert Systems with Applications, 2019, 138: 127933.
[54]
HairJ F, TathamR L, AndersonR E, et al. Multivariate data analysis[M]. Upper Saddle River: Prentice-Hall, 1998.