The invention discloses a region-level traffic prediction method based on a depth map convolutional network, and the method comprises the steps: constructing an undirected graph in a prediction region, and generating a spatial adjacency matrix of the prediction region based on the undirected graph; reading a historical feature sequence of each node in the undirected graph, and calculating the similarity between any two historical feature sequences based on space-time correlation to form a semantic adjacency matrix; fusing the spatial adjacency matrix and the semantic adjacency matrix to form a self-adaptive adjacency matrix; and inputting the self-adaptive adjacent matrix and the historical feature sequences of all nodes into a traffic feature prediction model, and outputting traffic features of the intersection corresponding to each node in a future set duration by the traffic feature prediction model. The traffic characteristics of the nodes are predicted based on the adaptive adjacency matrix, dynamic attributes of historical traffic data and semantic space relations between the nodes are fused, and the dynamically adjusted adjacency matrix can reflect spatial and temporal changes in an actual traffic network, so that the accuracy of the model is improved.

    本发明公开一种基于深度图卷积网络的区域级交通预测方法,包括:构建预测区域内的无向图,基于无向图生成预测区域的空间邻接矩阵;读取无向图中各节点的历史特征序列,基于时空关联计算任意两个历史特征序列间的相似性,形成的语义邻接矩阵;将空间邻接矩阵与语义邻接矩阵进行融合,形成自适应邻接矩阵;将自适应邻接矩阵及所有节点的历史特征序列输入交通特征预测模型,交通特征预测模型输出各节对应路口在未来设定时长内的交通特征。基于自适应邻接矩阵进行节点的交通特征预测,融合和了历史交通数据的动态属性和节点间的语义空间关系,这种动态调整的邻接矩阵可以反应实际交通网络中的时空变化,从而提高模型的准确性。


    Access

    Download


    Export, share and cite



    Title :

    Region-level traffic prediction method based on depth map convolutional network


    Additional title:

    基于深度图卷积网络的区域级交通预测方法


    Contributors:
    CHEN CHUANMING (author) / JIN QI (author) / WANG WENKAI (author) / YU QINGYING (author) / DENG XIANG (author) / DUAN RUIJIA (author) / NI TIANJIAO (author) / ZHENG XIAOYAO (author)

    Publication date :

    2025-05-16


    Type of media :

    Patent


    Type of material :

    Electronic Resource


    Language :

    Chinese


    Classification :

    IPC:    G06N COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS , Rechnersysteme, basierend auf spezifischen Rechenmodellen / G06Q Datenverarbeitungssysteme oder -verfahren, besonders angepasst an verwaltungstechnische, geschäftliche, finanzielle oder betriebswirtschaftliche Zwecke, sowie an geschäftsbezogene Überwachungs- oder Voraussagezwecke , DATA PROCESSING SYSTEMS OR METHODS, SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTING PURPOSES / G08G Anlagen zur Steuerung, Regelung oder Überwachung des Verkehrs , TRAFFIC CONTROL SYSTEMS / G06F ELECTRIC DIGITAL DATA PROCESSING , Elektrische digitale Datenverarbeitung



    Traffic prediction method using dynamic multi-level convolutional network

    LI HE / JIN DUO / LI XUEJIAO et al. | European Patent Office | 2024

    Free access


    Traffic accident prediction method based on graph convolutional network

    YANG QIAO / LI RUI / QI TIANJING | European Patent Office | 2023

    Free access

    Traffic flow prediction method based on graph convolutional network

    XU HUI / MENG FANYU / REN QIANQIAN et al. | European Patent Office | 2025

    Free access

    POI-based double-layer graph convolutional network traffic prediction method

    SU LI / ZHAO XINHAO / WU ZHE et al. | European Patent Office | 2023

    Free access