The driving style of an Autonomous Vehicle (AV) refers to how it behaves and interacts with other AVs. In a multi-vehicle autonomous driving system, an AV capable of identifying the driving styles of its nearby AVs can reliably evaluate the risk of collisions and make more reasonable driving decisions. However, there has not been a consistent definition of driving styles for an AV in the literature, although it is considered that the driving style is encoded in the AV's trajectories and can be identified using Maximum Entropy Inverse Reinforcement Learning (ME-IRL) methods as a cost function. Nevertheless, an important indicator of the driving style, i.e., how an AV reacts to its nearby AVs, is not fully incorporated in the feature design of previous ME-IRL methods. In this paper, we describe the driving style as a cost function of a series of weighted features. We design additional novel features to capture the AV's reaction-aware characteristics. Then, we identify the driving styles from the demonstration trajectories generated by the Stochastic Model Predictive Control (SMPC) using a modified ME-IRL method with our newly proposed features. The proposed method is validated using MATLAB simulation and an off-the-shelf experiment.


    Zugriff

    Download


    Exportieren, teilen und zitieren



    Titel :

    Identifying Reaction-Aware Driving Styles of Stochastic Model Predictive Controlled Vehicles by Inverse Reinforcement Learning


    Beteiligte:
    Dang, Ni (Autor:in) / Shi, Tao (Autor:in) / Zhang, Zengjie (Autor:in) / Jin, Wanxin (Autor:in) / Leibold, Marion (Autor:in) / Buss, Martin (Autor:in)

    Erscheinungsdatum :

    2023


    Medientyp :

    Preprint


    Format :

    Elektronische Ressource


    Sprache :

    Englisch





    Studies on Drivers’ Driving Styles Based on Inverse Reinforcement Learning

    Zhu, Bing / Wang, Jinsong / Deng, Weiwen et al. | SAE Technical Papers | 2018


    METHOD AND SYSTEM FOR MODELING PERSONALIZED CAR-FOLLOWING DRIVING STYLES WITH MODEL-FREE INVERSE REINFORCEMENT LEARNING

    WANG ZIRAN / HAN KYUNGTAE / GUPTA ROHIT | Europäisches Patentamt | 2024

    Freier Zugriff


    Learning driving styles for autonomous vehicles from demonstration

    Kuderer, Markus / Gulati, Shilpa / Burgard, Wolfram | IEEE | 2015