The invention relates to the field of path planning, in particular to a vehicle path planning method and system based on reinforcement learning, and the method comprises the steps: constructing a directed graph through a traffic road network, one node in the directed graph being an intersection in the traffic road network, and if road connection exists between two nodes, an edge exists between the two nodes; real-time traffic flow information is obtained, weights are given to edges in the directed graph according to the congestion condition of traffic flow, and the weight of the edges is lower when the congestion condition is serious; inputting the directed graph into a graph convolutional network, and extracting a road condition flow characteristic matrix from the directed graph; the vehicle is regarded as an intelligent agent, the intelligent agent selects an execution action according to the current state every time the vehicle arrives at a node, and the state of the intelligent agent is updated according to the selected action; the action selected by the intelligent agent is that the road condition flow characteristic matrix, the position matrix of the current intelligent agent and the position matrix of the destination are spliced together to serve as the state of the intelligent agent in the advancing direction of the node.

    本发明涉及路径规划领域,特别涉及一种基于强化学习的车辆路径规划方法及系统,方法包括通过交通路网构建有向图,有向图中一个节点为交通路网中的一个路口,若两个节点之间存在道路连接则两个节点之间存在一条边;获取实时的交通流信息,根据交通流的拥堵情况为有向图中的边赋予权重,拥堵情况越严重边权重越低;将有向图输入到图卷积网络中,从有向图中提取得到路况流量特征矩阵;将车辆视为一个智能体,每到达一个节点智能体根据当前状态选择执行的动作,并根据选择的动作更新智能体的状态;智能体选择的动作为在该节点处的前进方向,将路况流量特征矩阵、当前智能体所在的位置矩阵和目的地所在的位置矩阵拼接在一起作为智能体的状态。


    Access

    Download


    Export, share and cite



    Title :

    Vehicle path planning method and system based on reinforcement learning


    Additional title:

    一种基于强化学习的车辆路径规划方法及系统


    Contributors:
    HU SINAN (author) / LU XINGYU (author) / LIU YANBING (author) / MENG KAI (author) / TU QILIN (author)

    Publication date :

    2024-03-15


    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 / G06N COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS , Rechnersysteme, basierend auf spezifischen Rechenmodellen



    Intention-driven reinforcement learning-based path planning method

    ZHANG HUA / SU NA / WANG JUNBO | European Patent Office | 2024

    Free access

    Parking path planning method based on reinforcement learning

    WANG ZHONGLI / CAI BOGEN / WANG HAO et al. | European Patent Office | 2020

    Free access

    INTENTION-DRIVEN REINFORCEMENT LEARNING-BASED PATH PLANNING METHOD

    ZHANG HUA / SU NA / WANG JUNBO | European Patent Office | 2024

    Free access


    Deep Reinforcement Learning-Based UAV Path Planning Algorithm

    Wang, Kunfu / Hui, Ma / Hou, Jiajun et al. | IEEE | 2024