The invention discloses a large-scale adaptive traffic signal control method for peak congestion. The method comprises the following steps: constructing a road network structure as a subsequent training and testing scene; obtaining road section-level congestion index data of the road network in a to-be-studied time period, and constructing a peak traffic flow file as traffic flow data of a simulation stage; carrying out traffic network clustering; performing multi-agent reinforcement learning training on each sub-road network; and inputting the trained MADQN controller into a corresponding sub-road network to execute a signal control task, and outputting an optimal signal phase selection in a current state in real time. The method can adapt to various road network scenes with complex road structures, dense traffic flow and time variation, efficiently and adaptively completes traffic signal control tasks, shows remarkable advantages in various traffic evaluation indexes such as average vehicle speed, average vehicle driving time and average intersection parking times, has application potential and value, and is high in expansibility.

    本发明公开一种面向高峰时段拥堵的大规模自适应交通信号控制方法,包括以下步骤:构建路网结构作为后续训练及测试场景;获取该路网在待研究时段的路段级拥堵指数数据并构建高峰车流量文件,作为仿真阶段的车流量数据;进行交通网络聚类;对每个子路网进行多智能体强化学习训练;将训练好的MADQN控制器投入对应子路网执行信号控制任务,实时输出当前状态下最优的信号相位选择。本发明方法能够适应各种道路结构复杂、车流密集时变的路网场景,高效、自适应地完成交通信号控制任务,在平均车速、平均车辆行驶时间、平均路口停车次数等各项交通评价指标上都表现出显著优势,具有应用潜力和价值,扩展性强。


    Access

    Download


    Export, share and cite



    Title :

    Large-scale adaptive traffic signal control method for congestion in peak period


    Additional title:

    一种面向高峰时段拥堵的大规模自适应交通信号控制方法


    Contributors:
    FANG YI (author) / TAO ZHICHENG (author) / LI CHAO (author) / CHEN JIMING (author) / HE SHIBO (author)

    Publication date :

    2024-09-06


    Type of media :

    Patent


    Type of material :

    Electronic Resource


    Language :

    Chinese


    Classification :

    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



    Peak-period traffic congestion : options for current programs

    Remak, Roberta / Rosenbloom, Sandra | TIBKAT | 1976



    Dynamic model of peak period congestion

    Ben-Akiva, Moshe / Cyna, Michèle / de Palma, André | Elsevier | 1983


    Congestion Control of Bursty Traffic in ATM Networks by Adaptive Peak Rate Adjustment

    Jiang, Z. / Li, L. | British Library Conference Proceedings | 1992