The invention provides a traffic flow prediction method based on a space-time diagram convolutional network, and aims to solve the difficult-to-solve problems of high dimension, nonlinearity, uncertainty and the like of data in traffic flow prediction. The provided prediction model comprises a full-connection neural network layer, a space-time convolution block and a one-dimensional convolution layer. Traffic flow data passes through a full connection layer and then enters a space-time convolution block so as to capture time features and space features of the flow data. And finally, inputting a result of the space-time convolution block into a one-dimensional convolution layer to obtain a final prediction result. The space-time convolution block comprises a time convolution network layer, a time attention layer, a graph convolution layer and a gating fusion layer. The time convolution network layer and the time attention layer are used for capturing time features of different scales of traffic flow, the graph convolution layer is used for dynamically capturing spatial features, and the gating fusion layer is used for fusing feature matrixes of different scales. The space-time convolution block dynamically captures the space-time characteristics of different scales of traffic flow and screens important parts, so that the prediction precision is improved.

    本发明提出了一种基于时空图卷积网络的交通流量预测方法,旨在解决交通速流量预测中数据高维、非线性、不确定性等难以解决的问题。所提出的预测模型包括全连接神经网络层、时空卷积块和一维卷积层。交通流量数据经过全连接层后进入时空卷积块,以捕获流量数据的时间特征与空间特征。最后,时空卷积块的结果输入一维卷积层得到最后的预测结果。时空卷积块包括时间卷积网络层、时间注意力层、图卷积层、门控融合层。时间卷积网络层与时间注意力层用于捕获交通流量不同尺度的时间特征,图卷积层用于动态捕获空间特征,门控融合层用于融合不同尺度的特征矩阵。时空卷积块动态捕获了交通流量不同尺度的时空特征且筛选了重要的部分,提高了预测精度。


    Access

    Download


    Export, share and cite



    Title :

    Traffic flow prediction method based on space-time diagram convolutional network


    Additional title:

    一种基于时空图卷积网络的交通流量预测方法


    Contributors:
    JIANG CONG (author) / SONG YUN (author) / DENG ZELIN (author) / FAN WENDONG (author) / LUO YU (author)

    Publication date :

    2023-04-25


    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



    Traffic flow prediction method based on space-time diagram convolutional network

    JIANG YUNLIANG / XIA RENHUAN / ZHANG XIONGTAO et al. | European Patent Office | 2022

    Free access

    Traffic flow prediction method based on multi-space-time diagram convolutional network

    SHI QUAN / DAI JUNMING / SHEN QINQIN et al. | European Patent Office | 2021

    Free access

    Traffic flow prediction method of space-time diagram convolutional network

    TENG FEI / WANG ZIDAN / QIAO LU et al. | European Patent Office | 2024

    Free access

    Optimized traffic flow prediction model based on space-time diagram convolutional network

    ZHANG HONG / ZHAO YUANMENG / CHEN LINLONG et al. | European Patent Office | 2021

    Free access

    Traffic flow prediction method and system based on space-time diagram convolutional neural network

    WU JIANQING / XU YE / HUO YANQIANG et al. | European Patent Office | 2023

    Free access