The invention discloses a traffic adaptive control method based on multi-agent reinforcement learning. The method comprises the following steps: S1, acquiring traffic information of a road traffic environment and calculating the information of the road traffic environment; s2, intelligent agents corresponding to all the intersections are established, all the intersections in the traffic network are sorted, and the phase selection sequence of all the intersections is determined; s3, optimizing a signal control strategy of the intersection, and establishing a virtual simulation platform combined by UE and Simulink to simulate the signal control strategy; s4, real-time traffic characteristics of all intersections in the traffic network are obtained, and combined actions of adjacent agents are obtained; s5, obtaining observation information and rewards after the action is executed and observation track information obtained by simulation after each execution; s6, generating interaction experience; s7, updating a reviewer network and an actor network in the intelligent agent in real time by using interaction experience to find an optimal strategy; according to the method, the stability and the convergence speed in the whole agent training process are improved.

    本发明公开了一种基于多智能体强化学习的交通自适应控制方法,包括如下步骤:S1:获取道路交通环境的交通信息并计算道路交通环境的信息;S2:建立与各个交叉路口对应的智能体,对交通路网中各个交叉路口进行排序,确定各个交叉路口的相位选择顺序;S3:优化交叉口的信号控制策略,并搭建UE与Simulink联合的虚拟仿真平台对其仿真;S4:获取交通路网中各个交叉路口的实时交通特征,获得相邻智能体的联合动作;S5:获取执行动作后的观测信息、奖励以及每次执行后仿真所获得的观察轨迹信息;S6:生成交互经验;S7:利用交互经验,实时更新智能体中的评论者网络和行动者网络,以找到最优策略;本发明提高了整个智能体训练过程中的稳定性与收敛速度。


    Access

    Download


    Export, share and cite



    Title :

    Traffic adaptive control method based on multi-agent reinforcement learning


    Additional title:

    一种基于多智能体强化学习的交通自适应控制方法


    Contributors:
    HUANG HUA (author) / ZHANG WEI (author) / LI XIAOLONG (author) / LI CHUANG (author) / ZHANG MIN (author) / TAN SHUANG (author) / YANG ZIYI (author)

    Publication date :

    2024-06-07


    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



    Adaptive traffic signal control method based on multi-agent reinforcement learning

    ZHANG CHENGWEI / JIN SHAN / ZHENG KANGJIE | European Patent Office | 2021

    Free access

    Multi-agent reinforcement learning method for fair adaptive traffic signal control

    ZHANG CHENGWEI / FANG WANQING / ZHAO XINTIAN | European Patent Office | 2023

    Free access

    Multi-agent reinforcement learning for integrated and networked adaptive traffic signal control

    EL-TANTAWY SAMAH / ABDULHAI BAHER | European Patent Office | 2017

    Free access

    MULTI-AGENT REINFORCEMENT LEARNING FOR INTEGRATED AND NETWORKED ADAPTIVE TRAFFIC SIGNAL CONTROL

    EL-TANTAWY SAMAH / ABDULHAI BAHER | European Patent Office | 2018

    Free access

    Research on Multi-Agent Reinforcement Learning Traffic Control

    Fu, Xinpeng / Chen, Simin / Liang, Qixian et al. | IEEE | 2023