Highlights Three types of learning styles—stable, adaptive and changeful are discovered in a reinforced driving task. The evolutionary index is introduced to distinguish learning styles and depict the learning curve. The β band EEG signal at parietal region is significantly activated in the whole learning process.

    Abstract Overtaking maneuvers occur when vehicle drivers pursue higher driving speeds or comfort scenarios through back-to-back lane-changing behaviors, which require active participation of mental resources and certain self-learning practices. However, few studies have investigated how brain activities change during overtaking. Moreover, the learning process, which indicates the heterogeneity of drivers from a process-based perspective, has been neglected. In this work, we studied varied overtaking and learning styles using electroencephalogram (EEG) signals collected from drivers during a simulated driving task with a possible learning process. The average speed, standard deviation of speed, steering wheel angle and lateral movement distance of overtaking behaviors are analyzed in these reinforced tasks to evaluate overtaking performance. Four types of overtaking styles (i.e., low-speed type, low-speed & strong-oscillation type, high-speed & strong-steering type, and high-speed & close-distance type) and three types of learning styles (i.e., stable, adaptive and changeful) are discovered, not only from eventual overtaking behaviors but also from behavioral changes in a certain learning process. EEG features, such as the power spectral density (PSD) in the θ, α, β and γ bands, are extracted to characterize driver mental states and to correlate with heterogeneous learning styles. The obtained results show that fatigue and fatigue confrontation are more likely with a stable learning style, and the mental workload is reduced with an adaptive learning style, whereas no significant changes in brain activity are apparent with a changeful learning style. Understanding and recognizing heterogeneous overtaking and learning styles with varying EEG patterns will be extremely useful in the future for deep integration of advanced driving assistance systems (ADASs) and brain computer interface (BCI) systems.


    Zugriff

    Zugriff prüfen

    Verfügbarkeit in meiner Bibliothek prüfen

    Bestellung bei Subito €


    Exportieren, teilen und zitieren



    Titel :

    Heterogeneous overtaking and learning styles with varied EEG patterns in a reinforced driving task


    Beteiligte:
    Zhao, Shuo (Autor:in) / Guan, Wei (Autor:in) / Qi, Geqi (Autor:in) / Li, Peihao (Autor:in)


    Erscheinungsdatum :

    2022-04-02




    Medientyp :

    Aufsatz (Zeitschrift)


    Format :

    Elektronische Ressource


    Sprache :

    Englisch




    Driving strategies in overtaking

    Wilson, Tay / Best, W. | Elsevier | 1981


    Learning Dynamic Graph for Overtaking Strategy in Autonomous Driving

    Hu, Xuemin / Liu, Yanfang / Tang, Bo et al. | IEEE | 2023


    Overtaking Maneuvers in Simulated Highway Driving using Deep Reinforcement Learning

    Kaushik, Meha / Prasad, Vignesh / Krishna, K Madhava et al. | IEEE | 2018



    Automatic driving overtaking control method and system

    LIU BAOXIANG | Europäisches Patentamt | 2020

    Freier Zugriff