The study aims to assess the stress load levels of drivers on freeway tunnel sections and analyze the influencing factors. By recruiting 32 drivers for natural driving experiments across five tunnel segments, data were collected, including heart rate, Skin Conductance Response, the NASA Task Load Index questionnaire, as well as multimodal indicators such as vehicle speed, acceleration, and ambient illuminance during the driving process. To establish the evaluation model, the efficient LightGBM algorithm and the highly interpretable SHapley Additive exPlanations method were employed. Upon comparison with other prominent machine learning models, LightGBM demonstrated superior predictive performance. The results of the factor analysis revealed that the driving process, the minimum distance to tunnel entrances/exits, speed, longitudinal acceleration, and ambient illuminance were among the most critical factors influencing driver stress load. This study introduces an assessment approach for evaluating driver stress load levels on freeways, offering novel perspectives and methodologies to enhance tunnel safety and traffic efficiency.


    Access

    Check access

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    Evaluation of Driver Stress Load in Freeway Tunnel Based on LightGBM and SHAP Methods


    Contributors:
    Ge, Hongcheng (author) / Zhou, Mingfa (author) / Guo, Zhongyin (author) / Cheng, Lizhi (author)

    Conference:

    24th COTA International Conference of Transportation Professionals ; 2024 ; Shenzhen, China


    Published in:

    CICTP 2024 ; 2283-2292


    Publication date :

    2024-12-11




    Type of media :

    Conference paper


    Type of material :

    Electronic Resource


    Language :

    English