The invention relates to a fuel cell automobile energy management method combining working condition prediction and reinforcement learning. The fuel cell automobile energy management method is characterized by comprising the following steps: step 1, acquiring and processing a driving working condition of a driver and constructing and optimizing a neural network; 2, building a vehicle model of the fuel cell vehicle; and step 3, building a TD3-based deep reinforcement learning strategy, and carrying out training circulation of a fuel cell vehicle energy management algorithm strategy combining prediction and reinforcement learning. According to the method, the working condition speed, the acceleration, the charge state of the power battery and the equivalent factor adjustment coefficient are selected as state quantities, the TD3 algorithm is trained by utilizing standard working condition circulation until convergence, so that the optimal energy distribution of the power battery and the fuel battery is calculated, and the optimal energy distribution efficiency is improved while the working condition adaptability of the algorithm is improved. The fuel economy and the service life of the fuel cell can be effectively improved.

    本发明涉及了一种结合工况预测与强化学习的燃料电池汽车能量管理方法,其特征在于步骤如下:步骤1:驾驶员出行工况的采集、处理与神经网络的搭建及优化;步骤2:燃料电池汽车车辆模型搭建;步骤3:基于TD3深度强化学习策略的搭建,并进行结合预测与强化学习的燃料电池汽车能量管理算法策略的训练循环。其状态量选择工况速度、加速度、动力电池荷电状态和等效因子调节系数,利用标准工况循环对TD3算法进行训练,直到收敛,从而计算动力电池和燃料电池的最优能量分配,在提升算法的工况适应性的同时,能够有效的提升燃料经济性和燃料电池的寿命。


    Zugriff

    Download


    Exportieren, teilen und zitieren



    Titel :

    Fuel cell vehicle energy management method combining working condition prediction and reinforcement learning


    Weitere Titelangaben:

    结合工况预测与强化学习的燃料电池汽车能量管理方法


    Beteiligte:
    LU YANHUI (Autor:in) / ZHANG YUNRUI (Autor:in) / WANG PENGYU (Autor:in) / MA WEIYANG (Autor:in)

    Erscheinungsdatum :

    2024-03-05


    Medientyp :

    Patent


    Format :

    Elektronische Ressource


    Sprache :

    Chinesisch


    Klassifikation :

    IPC:    B60L PROPULSION OF ELECTRICALLY-PROPELLED VEHICLES , Antrieb von elektrisch angetriebenen Fahrzeugen / G06N COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS , Rechnersysteme, basierend auf spezifischen Rechenmodellen / H01M Verfahren oder Mittel, z.B. Batterien, für die direkte Umwandlung von chemischer in elektrische Energie , PROCESSES OR MEANS, e.g. BATTERIES, FOR THE DIRECT CONVERSION OF CHEMICAL ENERGY INTO ELECTRICAL ENERGY



    Fuel cell vehicle energy management optimization method based on working condition prediction

    SONG DAFENG / YANG LILI / ZENG XIAOHUA et al. | Europäisches Patentamt | 2021

    Freier Zugriff

    Rule fusion deep reinforcement learning energy management method based on working condition identification

    ZHOU XIAOCHUAN / CHANG CHENGCHENG / ZHANG ZIYU et al. | Europäisches Patentamt | 2021

    Freier Zugriff

    Fuel cell hybrid electric vehicle energy management method based on deep reinforcement learning

    MA YAN / LUAN YU | Europäisches Patentamt | 2024

    Freier Zugriff

    Fuel cell vehicle deep reinforcement learning energy management method based on experience pool migration

    ZHOU JIANHAO / XU DA / GUO AIJUN et al. | Europäisches Patentamt | 2023

    Freier Zugriff

    Hydrogen fuel unmanned aerial vehicle energy management method based on deep reinforcement learning

    WU YU / ZHANG MINGXUAN / WANG YUYANG et al. | Europäisches Patentamt | 2021

    Freier Zugriff