The invention provides a traffic flow prediction method and system based on time sequence dependent attention sparse convolution, and the method comprises the steps: dividing a target city into a plurality of grid regions, and defining the concepts of traffic inflow and traffic outflow at the same time, so as to calculate the traffic inflow and traffic outflow of any grid region; the method comprises the following steps of: acquiring traffic flow data of an area, generating historical flow data based on traffic inflow and traffic outflow of the area, extracting hidden features by adopting a convolutional layer so as to obtain a three-dimensional feature tensor, extracting a spatial-temporal feature sequence from the three-dimensional feature tensor based on an improved ConvLSTM network, and further calculating time correlation attention features among a plurality of time intervals. And then the spatio-temporal features and the time correlation attention features which are extracted deeply are fused to obtain a fused feature, and finally a three-dimensional feature tensor in a future time interval is predicted according to the fused feature. The method provided by the invention improves the accuracy of predicting the traffic flow result of the future time interval.

    本发明提出一种基于时序依赖注意力稀疏卷积的交通流量预测方法及系统,该方法包括:首先将目标城市划分为多个网格区域,同时定义交通流入和交通流出的概念,以计算出任一网格区域的交通流入和交通流出,而后再基于该区域的交通流入和交通流出生成历史流量数据,并采用卷积层提取隐藏特征,从而得到三维特征张量,再基于改进后的ConvLSTM网络从三维特征张量中提取时空特征序列,进而计算出多个时间间隔之间的时间相关性注意力特征,而后再将深度提取出的时空特征和时间相关性注意力特征进行融合,得到一融合特征,最后根据该融合特征预测在未来时间间隔下的三维特征张量。本发明提出的方法提高预测未来时间间隔的交通流量结果的准确度。


    Access

    Download


    Export, share and cite



    Title :

    Traffic flow prediction method and system based on time sequence dependent attention sparse convolution


    Additional title:

    基于时序依赖注意力稀疏卷积的交通流量预测方法及系统


    Contributors:
    LIAO HUICHUAN (author) / TANG JIE (author) / HUANG XIAOHUI (author)

    Publication date :

    2023-05-23


    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 of space-time convolution fusion probability sparse attention mechanism

    ZHANG HONG / CHEN LINBIAO / CHEN LINLONG et al. | European Patent Office | 2023

    Free access

    Traffic flow prediction method based on interactive dynamic graph convolution and probability sparse attention

    ZHANG HONG / CHEN LINBIAO / CHEN LINLONG et al. | European Patent Office | 2023

    Free access

    Attention-based dynamic space-time diagram convolution traffic flow prediction system and method

    JI ZHENYUAN / HUAN YUEHUI / YANG GENKE et al. | European Patent Office | 2023

    Free access

    Long-time-sequence traffic flow prediction method based on graph convolution-Informer model

    GUO ZIQIANG / CHENG BAOXI / YANG XIAOLEI et al. | European Patent Office | 2021

    Free access

    Traffic Flow Prediction Based on Interactive Dynamic Spatio-Temporal Graph Convolution with a Probabilistic Sparse Attention Mechanism

    Chen, Linlong / Chen, Linbiao / Wang, Hongyan et al. | Transportation Research Record | 2024