The invention provides a deep learning-based graph neural network traffic flow prediction method. The method is composed of three modules fusing different deep learning methods: a feature attention fusion module, an information aggregation module and a multi-information combination module. The fusion feature attention module extracts influence factors among different traffic information through an attention mechanism and a Softmax function, combines traffic flow information with other traffic information in a time dimension, and fully considers time periodicity among different traffic information; the information aggregation module inputs the traffic flow information into a GRU network to extract historical time information, and realizes synchronous extraction of space-time dependency by using graph convolution; and the multi-information combination module adds the primary information and the secondary information by using a CONCAT (.) function and then performs graph convolution operation to mine a hidden relationship between the primary information and the secondary information. According to the invention, an efficient traffic flow prediction method is realized, and the traffic flow information can be rapidly and accurately predicted.

    本发明提出了一种基于深度学习的图神经网络交通流预测方法。该方法由融合了不同深度学习方法的三个模块:融合特征注意力模块、信息聚合模块和多信息结合模块组成。融合特征注意力模块通过注意力机制和Softmax函数提取不同交通信息间的影响因子,将交通流信息与其他各种交通信息在时间维度上结合,充分考虑了不同交通信息间的时间周期性;信息聚合模块将交通流信息输入GRU网络中提取历史时间信息,同时用图卷积实现时空依赖性的同步提取;多信息结合模块将主次信息用CONCAT(·)函数相加后进行图卷积运算,挖掘主次要信息间的隐藏关系。本发明实现了一个高效的交通流预测方法,能够快速准确的预测交通流信息。


    Access

    Download


    Export, share and cite



    Title :

    Graph neural network traffic flow prediction method based on deep learning


    Additional title:

    一种基于深度学习的图神经网络交通流预测方法


    Contributors:
    CHENG XIAOHUI (author) / HE YUHAO (author) / LU QIU (author)

    Publication date :

    2023-05-02


    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 / 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



    Traffic flow prediction method based on graph neural network

    PENG LAIHU / WU BAOWEN / QI YUBAO et al. | European Patent Office | 2024

    Free access

    Traffic flow prediction method based on federated learning and graph neural network

    MI BO / ZENG RAN / HUANG DARONG et al. | European Patent Office | 2024

    Free access

    Airport traffic flow prediction method based on graph neural network

    YAN ZHEN / YANG HONGYU / WU XIPING et al. | European Patent Office | 2023

    Free access

    Node adaptive learning graph attention neural network traffic flow prediction method

    LOU JUNGANG / HUANG XUXIANG / SHEN QING et al. | European Patent Office | 2024

    Free access

    Traffic flow prediction method based on dynamic graph neural network

    XU GUANGXIA / HU XINTING / CHEN LANG et al. | European Patent Office | 2022

    Free access