The invention discloses a time-space traffic state prediction method based on a Transform network. The method comprises the steps of data code embedding, global spatial feature extraction, local spatial feature information extraction, global-local feature fusion and single-step traffic state prediction. According to the method, the dynamic nature of the dependency relationship of the traffic state in space is fully considered, and step-by-step feature extraction and modeling are carried out hierarchically, so that the prediction of the method is more accurate; the dynamic traffic space dependency relationship extracted by the method can be explicitly expressed, has interpretability, and can provide reference for traffic planning; moreover, the method can extract and construct the global spatial features of the traffic network in a pure data-driven manner, still obtains a better prediction result in the absence of spatial feature data of the traffic network, and is reliable. Therefore, the method is more accurate and more reliable in traffic state prediction, has interpretability, has lower requirements on data, and is more suitable for popularization and application.

    本发明公开了一种基于Transformer网络的时空交通状态预测方法,包括数据编码嵌入、全局空间特征提取、局部空间特征信息提取、全局‑局部特征融合以及单步交通状态预测。该方法充分考虑到了交通状态在空间上依赖关系的动态性,并层次化的进行分步的特征提取与建模,使得该方法的预测更精确;该方法提取的动态交通空间依赖关系能够显式表达,具有可解释性,可以为交通规划提供参考;且该方法能够纯数据驱动地提取和构建交通网络的全局空间特征,在缺少交通网络空间特征数据的情况下依然得到较好的预测结果,具有可靠性。因此,该方法对交通状态预测更精确、更可靠,具有可解释性,对数据的要求更低,更适合推广应用。


    Access

    Download


    Export, share and cite



    Title :

    Time-space traffic state prediction method based on Transform network


    Additional title:

    一种基于Transformer网络的时空交通状态预测方法


    Contributors:
    MA XIAOLEI (author) / YAN HAOYANG (author) / LI YUJIE (author)

    Publication date :

    2021-09-03


    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



    Transform-based adaptive space-time diagram neural network traffic flow prediction method and system

    WANG RONG / ZHAO JIANKUAN / LI MIAOFEI et al. | European Patent Office | 2022

    Free access

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

    ZHENG LINJIANG / CHEN YILING / LIU WEINING et al. | European Patent Office | 2022

    Free access

    Traffic flow prediction method based on improved space-time Transform

    GAO RONG / WAN YILIANG / SHAO XIONGKAI | European Patent Office | 2022

    Free access

    Traffic speed prediction method based on multi-spatial scale space-time Transform

    ZHANG YUE / LIU BAILONG / AN JIYONG et al. | European Patent Office | 2023

    Free access

    Traffic state prediction method based on adaptive dynamic space-time diagram convolutional network

    ZHUANG XUFEI / MAO RUI / GAO XUDONG et al. | European Patent Office | 2024

    Free access