The invention discloses an active automobile suspension control system and method based on model prediction and reinforcement learning, the system comprises a data acquisition module, a modeling processing module, a reinforcement modeling module, a reinforcement learning module and a control output module, the data acquisition module is in control connection with the modeling processing module, the modeling processing module is in control connection with the reinforcement modeling module, and the reinforcement learning module is in control connection with the reinforcement modeling module. According to the method, model prediction and reinforcement learning are combined on the basis of pavement condition preview, and an organic model and algorithm integration framework is established; in the aspect of model prediction, a seven-degree-of-freedom mathematical model is used for predicting the future motion state trend of the suspension, so that an accurate reference is provided for subsequent control decision making; in the aspect of reinforcement learning, the PPO algorithm is used for dynamic vehicle control, the optimization processing speed of a control strategy is increased, the state indexes such as the vertical acceleration, the pitch angle acceleration and the roll angle acceleration of a vehicle body are reduced by optimizing the damping output of an automobile suspension shock absorber, and the suspension stability of an automobile is improved.

    本发明公开了基于模型预测和强化学习的主动汽车悬架控制系统及方法,包括数据采集模块、建模处理模块、强化建模模块、强化学习模块和控制输出模块,所述数据采集模块控制连接建模处理模块,建模处理模块控制连接强化建模模块,本发明,通过在路面状况预瞄的基础上将模型预测和强化学习相结合,建立有机的模型与算法集成框架;在模型预测方面,使用七自由度数学模型对悬架未来运动状态趋势进行预测,从而为后续的控制决策提供准确的参考;在强化学习方面,使用PPO算法进行动态化车辆控制,提高了控制策略的优化处理速度,通过优化汽车悬架减震器的阻尼输出来降低车身垂向加速度、俯仰角加速度和侧倾角加速度等状态指标,提高汽车的悬架稳定性。


    Access

    Download


    Export, share and cite



    Title :

    Active automotive suspension control system and method based on model prediction and reinforcement learning


    Additional title:

    基于模型预测和强化学习的主动汽车悬架控制系统及方法


    Contributors:
    JI QIJIN (author) / LI GUIXI (author)

    Publication date :

    2024-02-06


    Type of media :

    Patent


    Type of material :

    Electronic Resource


    Language :

    Chinese


    Classification :

    IPC:    B60G VEHICLE SUSPENSION ARRANGEMENTS , Radaufhängungen und Federungen für Fahrzeuge / G06N COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS , Rechnersysteme, basierend auf spezifischen Rechenmodellen



    Suspension control system based on reinforcement learning

    CHOI JONGEUN / YONG HWANMOO | European Patent Office | 2020

    Free access


    Neural network learning control of automotive active suspension systems

    Watanabe,Y. / Sharp,R.S. / Cranfield Univ.,GB | Automotive engineering | 1999


    Automotive suspension system and automotive suspension control method

    ZENG CHAO | European Patent Office | 2020

    Free access

    Moderated reinforcement learning of active and semi-active vehicle suspension control laws

    Frost, G.P. / Gordon, T.J. / Howell, M.N. et al. | Tema Archive | 1996