The invention discloses a regional flow prediction method based on deep learning, and the method comprises the following steps: carrying out the mesh generation of a prediction region according to the longitude and latitude, carrying out the preprocessing of regional flow original data, and obtaining the inflow and outflow data of each mesh at each moment; taking the convolutional long-term and short-term memory network as a basic composition unit of the prediction network and the coding network, taking the preprocessed data as the input of the coding network, and obtaining the output state of the coding network at each moment; adopting an attention mechanism, calculating to obtain corresponding attention weights according to the state of the prediction network at the previous moment and the obtained output states of the encoding network at all moments, and calculating a weighted sum of the attention weights and the output states of the encoding network at all moments to obtain different vectors at all moments; and combining the vector with the input of the prediction network, and inputting the combined vector into the prediction network to obtain a regional flow prediction result. The method disclosed by the invention can effectively extract the spatial-temporal features and improve the prediction precision.

    本发明公开了一种基于深度学习的区域流量预测方法,包括如下步骤:根据经纬度对预测区域进行网格划分,对区域流量原始数据进行预处理,得到每个时刻下每个网格的流入量和流出量数据;将卷积长短期记忆网络作为预测网络和编码网络的基本组成单元,将预处理后的数据作为编码网络的输入,得到编码网络各个时刻的输出状态;采用注意力机制,根据上一时刻预测网络的状态与得到的编码网络各个时刻的输出状态计算得到对应的注意力权重,并对注意力权重与编码网络各个时刻的输出状态求加权和,得到各个时刻下不同的向量;将向量与预测网络的输入结合起来,输入到预测网络,得到区域流量预测结果。本发明所公开的方法能够有效提取时空特征、提高预测精度。


    Zugriff

    Download


    Exportieren, teilen und zitieren



    Titel :

    Regional flow prediction method based on deep learning


    Weitere Titelangaben:

    一种基于深度学习的区域流量预测方法


    Beteiligte:
    LI QIQIANG (Autor:in) / LIU JIAXIN (Autor:in) / SONG WEN (Autor:in)

    Erscheinungsdatum :

    2021-04-16


    Medientyp :

    Patent


    Format :

    Elektronische Ressource


    Sprache :

    Chinesisch


    Klassifikation :

    IPC:    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 / G08G Anlagen zur Steuerung, Regelung oder Überwachung des Verkehrs , TRAFFIC CONTROL SYSTEMS



    Traffic flow prediction method based on deep learning

    ZHANG SHICHAO / DUAN HONGXIANG / ZHANG CHENGYUAN | Europäisches Patentamt | 2022

    Freier Zugriff

    Traffic flow prediction method based on hybrid deep learning

    XIA DAWEN / CHEN YAN / LI HUAQING et al. | Europäisches Patentamt | 2023

    Freier Zugriff

    New Deep Learning-Based Passenger Flow Prediction Model

    Utku, Anıl / Kaya, Sema Kayapinar | Transportation Research Record | 2022


    Short-term traffic flow prediction method based on deep learning

    LI ZHUANGZHUANG / GUI ZHIMING / GUO LIMIN et al. | Europäisches Patentamt | 2020

    Freier Zugriff

    Traffic flow prediction method based on clustering and evolution deep learning

    ZHANG WANFENG / ZHENG YONGSHUN / PEI JIAYONG et al. | Europäisches Patentamt | 2024

    Freier Zugriff