According to the multi-intersection traffic signal control method based on deep reinforcement learning, traffic lights in a road network serve as an intelligent agent, the intelligent agent operates a deep reinforcement learning algorithm, phase selection of the traffic lights is trained according to the specific condition of a multi-intersection, and traffic flow adjustment can be completed by adjusting the phase selection of the traffic lights. According to the method, the action with the highest score value is obtained by inputting the traffic condition of the five-inlet intersection into a deep reinforcement learning algorithm, so that the phase of the traffic light is adjusted, and the adopted action is continuously adjusted, so that the traffic flow in the road network is optimally adjusted, the optimal traffic efficiency of the road is ensured, and meanwhile, the traffic flow can be properly simplified, so that the traffic efficiency is improved. The method is applied to 8-phase and 6-phase intersections. According to the method, the problems that the vehicle waiting time is too long, the lane occupancy rate is too high, and model training is unstable and unreasonable are solved.

    一种基于深度强化学习的多交叉路口交通信号控制方法,路网中红绿灯作为智能体,该智能体运行一个深度强化学习算法,根据多叉路口的具体情况对红绿灯的相位选择进行训练,通过对红绿灯相位选择的调节可以完成车流调节。根据五进路口的交通状况输入深度强化学习算法中获得评分值最高的行动,从而对红绿灯的相位进行调节,并且不断的对采取行动进行调整,从而使得路网中的车流得到最优的调节,保证道路的通行效率最优,同时也可以经过适当的简化,应用于8相位、6相位的交叉路口。本发明解决车辆等待时间过长、车道被占有率过高、模型训练不稳定不合理的问题。


    Zugriff

    Download


    Exportieren, teilen und zitieren



    Titel :

    Multi-intersection traffic signal control method based on deep reinforcement learning


    Weitere Titelangaben:

    基于深度强化学习的多交叉路口交通信号控制方法


    Beteiligte:
    LIU LIJUAN (Autor:in) / BAI GUANGMING (Autor:in)

    Erscheinungsdatum :

    2023-11-07


    Medientyp :

    Patent


    Format :

    Elektronische Ressource


    Sprache :

    Chinesisch


    Klassifikation :

    IPC:    G08G Anlagen zur Steuerung, Regelung oder Überwachung des Verkehrs , TRAFFIC CONTROL SYSTEMS / G06F ELECTRIC DIGITAL DATA PROCESSING , Elektrische digitale Datenverarbeitung / G06N COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS , Rechnersysteme, basierend auf spezifischen Rechenmodellen



    Multi-intersection traffic signal control method based on deep reinforcement learning

    DENG HENG / WANG YULONG / GAO YANG et al. | Europäisches Patentamt | 2024

    Freier Zugriff


    Single-intersection signal control method based on deep reinforcement learning algorithm

    HUANG YIWANG / WU QIAN | Europäisches Patentamt | 2023

    Freier Zugriff

    Deep Reinforcement Learning-based Traffic Signal Control

    Ruan, Junyun / Tang, Jinzhuo / Gao, Ge et al. | IEEE | 2023


    Multi-intersection graph deep reinforcement learning signal lamp control method under data loss

    XU DONGWEI / YU ZEFENG / LIAO XIANGWANG et al. | Europäisches Patentamt | 2024

    Freier Zugriff