Reinforcement learning (RL) is gradually being implemented in the hybrid electric vehicle (HEV) supervisory control. Even though RL exhibits significant fuel consumption saving, the long learning time makes it hardly applicable in real-world vehicles. This study aims to reduce the learning iterations of Q-learning in HEV application utilizing warm-start methods. Different from previous studies, which initiated Q-learning with zero or random Q values, this study initiates the Q-learning with different supervisory controls, and the detailed analysis is given. The results show that the proposed warm-start Q-learning requires 68.8% fewer iterations than cold-start Q-learning and improves 10%–16% MPG compared with equivalent consumption minimization strategy control. The results of this study can be used to facilitate the deployment of RL in vehicle applications.


    Zugriff

    Zugriff prüfen

    Verfügbarkeit in meiner Bibliothek prüfen

    Bestellung bei Subito €


    Exportieren, teilen und zitieren



    Titel :

    Learning Time Reduction Using Warm-Start Methods for a Reinforcement Learning-Based Supervisory Control in Hybrid Electric Vehicle Applications


    Beteiligte:
    Xu, Bin (Autor:in) / Hou, Jun (Autor:in) / Shi, Junzhe (Autor:in) / Li, Huayi (Autor:in) / Rathod, Dhruvang (Autor:in) / Wang, Zhe (Autor:in) / Filipi, Zoran (Autor:in)


    Erscheinungsdatum :

    01.06.2021


    Format / Umfang :

    5666720 byte




    Medientyp :

    Aufsatz (Zeitschrift)


    Format :

    Elektronische Ressource


    Sprache :

    Englisch





    Hybrid Electric Vehicle Powertrain Control Based on Reinforcement Learning

    Yoon, Hwan-Sik / Yao, Zhengyu | SAE Technical Papers | 2021



    Rule-based supervisory control of split-parallel hybrid electric vehicle

    Anbaran, Sajad A. / Idris, N. R. N. / Jannati, M. et al. | IEEE | 2014