A racing car reinforcement learning automatic driving method based on action space safety mapping adopts the technical scheme that an automatic driving controller of a racing car is established by using a neural network, motion state data of the racing car and data of a front racing track are input, and corresponding accelerator/brake control quantity and steering wheel control quantity are output. And according to the dynamic model of the racing car, establishing an action space safety mapping function, converting the neural network output quantity possibly causing the car to enter a dangerous state into a safe control quantity, and inputting the safe control quantity into a control system of the racing car. And training the neural network according to the reward feedback signal by using a reinforcement learning algorithm, continuously updating neural network parameters, and finally obtaining an optimized racing car automatic driving strategy.

    一种基于动作空间安全映射的赛车强化学习自动驾驶方法,本发明采用的技术方案是使用神经网络建立赛车的自动驾驶控制器,输入赛车自身的运动状态数据和前方赛道的数据,输出对应的油门/刹车控制量和方向盘控制量。根据赛车的动力学模型,建立动作空间安全映射函数,将可能导致车辆进入危险状态的神经网络输出量转化为安全的控制量再输入赛车的控制系统。利用强化学习算法,根据奖励反馈信号对神经网络进行训练,不断更新神经网络参数,最终获得优化的赛车自动驾驶策略。


    Zugriff

    Download


    Exportieren, teilen und zitieren



    Titel :

    Racing car reinforcement learning automatic driving method based on action space safety mapping


    Weitere Titelangaben:

    一种基于动作空间安全映射的赛车强化学习自动驾驶方法


    Beteiligte:
    WANG YUANDA (Autor:in) / YUAN XIN (Autor:in) / JIA YUBIN (Autor:in) / SUN CHANGYIN (Autor:in)

    Erscheinungsdatum :

    2023-12-08


    Medientyp :

    Patent


    Format :

    Elektronische Ressource


    Sprache :

    Chinesisch


    Klassifikation :

    IPC:    G05D SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES , Systeme zum Steuern oder Regeln nichtelektrischer veränderlicher Größen / B60W CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION , Gemeinsame Steuerung oder Regelung von Fahrzeug-Unteraggregaten verschiedenen Typs oder verschiedener Funktion



    Automatic driving decision uncertainty estimation method based on reinforcement learning

    ZOU QIANYING / LIAO YUAN | Europäisches Patentamt | 2023

    Freier Zugriff

    Automatic driving method and device based on hierarchical reinforcement learning

    MAO ZHIQI / PENG LIHUI / YAO DANYA et al. | Europäisches Patentamt | 2023

    Freier Zugriff

    Automatic driving decision-making method based on deep reinforcement learning

    LIU CHENGQI / LIU SHAOWEIHUA / ZHANG YUJIE et al. | Europäisches Patentamt | 2024

    Freier Zugriff

    Racing toward safety

    Online Contents | 2002


    Automatic driving vehicle control method based on multi-agent reinforcement learning

    MA HONGBIN / LIU PING / JIN YING et al. | Europäisches Patentamt | 2023

    Freier Zugriff