The invention provides a traffic flow prediction method of a recurrent neural network based on dynamic diffusion diagram convolution, which comprises the following steps of: establishing a topological structure model for a topological structure of a road, and learning a space inherent feature representation by utilizing a static topological structure from the consideration of the static topological structure; in view of a dynamic topological structure, the information of each node at each moment is subjected to own potential time feature learning by utilizing the feature representation of dynamic nodes; considering from the angle of diffusion diagram convolution, combining the inherent characteristics of the space and the own potential time characteristics, so that the model has the characteristics of the space and the time of the node at the moment, and has the capability of aggregating node neighbor information; on the basis of the perspective of a recurrent neural network, historical traffic flow information is integrated together under the condition that a model has space and time characteristics of a certain moment node, so that the model has a certain capability of extracting global time characteristics; and finally, accurate prediction of the traffic flow is completed.

    本发明提供了基于动态扩散图卷积的循环神经网络的交通流量预测方法,包括如下步骤:对于道路的拓扑结构建立拓扑结构模型,从静态拓扑结构考虑,利用静态拓扑结构来学习一个空间固有的特征表示;从动态拓扑结构考虑,利用动态节点的特征表示,对每个时刻的每个节点的信息进行一个自有潜在时间特征的学习;从扩散图卷积的角度考虑,结合空间固有的特征和自有潜在时间特征,使模型同时具有该时刻节点的空间和时间的特征性,具有聚合节点近邻信息的能力;从循环神经网络的角度考虑,在模型已经具有某一时刻节点的空间和时间的特征的情况下,将历史交通流量信息整合在一起,使模型具有一定的提取全局时间特征的能力;最终完成交通流量的准确预测。


    Zugriff

    Download


    Exportieren, teilen und zitieren



    Titel :

    Traffic flow prediction method of recurrent neural network based on dynamic diffusion diagram convolution


    Weitere Titelangaben:

    基于动态扩散图卷积的循环神经网络的交通流量预测方法


    Beteiligte:
    LI YUN (Autor:in) / TIAN YUDOU (Autor:in) / JING PEIGUANG (Autor:in)

    Erscheinungsdatum :

    2022-06-24


    Medientyp :

    Patent


    Format :

    Elektronische Ressource


    Sprache :

    Chinesisch


    Klassifikation :

    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



    Optimized Graph Convolution Recurrent Neural Network for Traffic Prediction

    Guo, Kan / Hu, Yongli / Qian, Zhen et al. | IEEE | 2021


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

    JI ZHENYUAN / HUAN YUEHUI / YANG GENKE et al. | Europäisches Patentamt | 2023

    Freier Zugriff

    Traffic prediction method based on dual dynamic space-time diagram convolution

    SUN YANFENG / JIANG XIANGHENG / HU YONGLI et al. | Europäisches Patentamt | 2022

    Freier Zugriff

    Traffic flow prediction method based on global diffusion convolution residual network

    ZHENG KAI / YE GUANYU / LI YUANMING et al. | Europäisches Patentamt | 2020

    Freier Zugriff

    Road section traffic flow prediction method based on time convolution neural network

    JIN SHENG / CHANG WEI | Europäisches Patentamt | 2020

    Freier Zugriff