The invention discloses a traffic flow prediction method based on space-time Granger causality. The method comprises the following steps: modeling a traffic flow transfer causality TCR into a space-time Granger causality graph; the TCR is captured according to a space-time alignment algorithm; a space-time Granger causality test method is used for testing which two nodes have GDTi transmission from the aligned time sequence, so that the stable TCR under the dynamic traffic flow is captured; and inputting the obtained space-time Granger causality test diagram into a GNN-based traffic prediction model, and capturing the time dependence and the space dependence of a traffic network at the same time. According to the method, the TCR is modeled through space-time Granger causality, a global and dynamic spatial dependency relationship can be expressed, and GDTi can be captured, so that the long-range prediction effect is improved; the graph neural network traffic flow prediction framework using the space-time Granger causality graph is used for improving the long-time prediction effect.

    本发明公开了一种基于时空格兰杰因果的交通流预测方法,包括步骤:将交通流传递的因果关系TCR建模为时空格兰杰因果关系图;根据时空对齐算法捕获TCR;使用时空格兰杰因果检验方法从对齐后的时间序列中测试出哪两个节点间有GDTi的传递,从而捕获动态交通流下稳定的TCR;将得到的时空格兰杰因果检验图输入到基于GNN的交通预测模型中,同时捕获交通路网的时间依赖性和空间依赖性。本发明提出时空格兰杰因果来建模TCR,能表达一种全局、动态的空间依赖关系,能捕获GDTi从而提升长程预测效果;使用时空格兰杰因果图的图神经网络交通流预测框架用于提升长时预测效果。


    Access

    Download


    Export, share and cite



    Title :

    Traffic flow prediction method based on space-time Granger causality


    Additional title:

    一种基于时空格兰杰因果的交通流预测方法


    Contributors:
    LI HAIFENG (author) / HE SILU (author) / LUO QINYAO (author) / SHI TIANDONG (author)

    Publication date :

    2023-06-09


    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 / 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



    Exploring Vehicular Interaction from Trajectories Based on Granger Causality

    Liao, Xishun / Wu, Guoyuan / Barth, Matthew J. et al. | IEEE | 2023


    Out-of-Sample Tests for Granger Causality

    Chao, J. / Corradi, V. / Swanson, N. R. | British Library Online Contents | 2001


    Granger Causal Inference for Interpretable Traffic Prediction

    Zhang, Lei / Fu, Kaiqun / Ji, Taoran et al. | IEEE | 2022



    Granger-causality between transportation and GDP: A panel data approach

    Beyzatlar, Mehmet Aldonat | Online Contents | 2014