The invention provides a traffic flow speed prediction method and system based on a directed hypergraph and an attention mechanism, and belongs to the technical field of traffic state prediction. Based on the directed hypergraph, a relationship between nodes in a road network is constructed; aggregating directed hyperedge information to represent complex spatial-temporal characteristics of the road network; building an encoder and decoder architecture; an attention mechanism is fused, an attention module is constructed, and an improved dense connection structure is introduced to improve the precision of the method. The method can effectively overcome the defects of a method based on a traditional graph structure, achieves the extraction and fusion of time sequence and spatial features, solves the problems of gradient disappearance and the like in a deep neural network, can be suitable for the multi-step prediction of the traffic flow in a real road network, and obtains good prediction precision.

    本发明提供一种基于有向超图及注意力机制的交通流速度预测方法及系统,属于交通状态预测技术领域,基于有向超图,构建道路网中各节点之间的关系;聚合有向超边信息,以表征路网复杂时空特征;搭建编码器、解码器架构;融合注意力机制、构建注意力模块,并引入改进密集连接结构,以提高方法精度。本发明能够有效克服基于传统图结构的方法缺陷,并实现时序和空间特征的提取及融合,同时解决深层神经网络中梯度消失等问题,可适用于真实路网中的交通流多步预测,并将获得良好的预测精度。


    Access

    Download


    Export, share and cite



    Title :

    Traffic flow speed prediction method and system based on directed hypergraph and attention mechanism


    Additional title:

    基于有向超图及注意力机制的交通流速度预测方法及系统


    Contributors:
    WANG JIANGFENG (author) / DING WEIDONG (author) / LI YUNFEI (author) / XIONG HUIYUAN (author) / LUO DONGYU (author) / SONG YUCHAO (author)

    Publication date :

    2023-11-28


    Type of media :

    Patent


    Type of material :

    Electronic Resource


    Language :

    Chinese


    Classification :

    IPC:    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 / G06N COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS , Rechnersysteme, basierend auf spezifischen Rechenmodellen / G08G Anlagen zur Steuerung, Regelung oder Überwachung des Verkehrs , TRAFFIC CONTROL SYSTEMS



    Directed hypergraph attention network for traffic forecasting

    Xiaoyi Luo / Jiaheng Peng / Jun Liang | DOAJ | 2022

    Free access

    Directed hypergraph attention network for traffic forecasting

    Luo, Xiaoyi / Peng, Jiaheng / Liang, Jun | Wiley | 2022

    Free access

    Transform-fused hypergraph neural network-based high-speed station traffic flow prediction method

    LI LING / LIU NING / SU BO et al. | European Patent Office | 2024

    Free access

    Equivalent hypergraph construction method for traffic flow prediction

    WANG FENG / CAO YUHANG / ZHANG JIEYIN et al. | European Patent Office | 2025

    Free access

    Deep learning traffic flow prediction method based on time perception hypergraph

    HE ZHIXIANG / GONG MENGZAN | European Patent Office | 2024

    Free access