The invention discloses an associated multi-intersection traffic flow prediction method. The method comprises the following steps: 1, collecting traffic flow data of an associated intersection, preprocessing the traffic flow data of the intersection, and dividing the traffic flow data into a training set, a verification set and a test set; 2, extracting spatial features from the input intersection traffic flow data by using CNN (Convolutional Neural Network); 3, taking the spatial features as input, and extracting time features by using Transform; 4, after the Decoder layer is completely executed, finally respectively outputting three vectors to the data of the three time windows, stacking the three vectors, and inputting the stacked three vectors to an average pooling layer; 5, setting model parameters; and 6, training the model until the maximum training period, and performing a traffic flow prediction task on the associated multiple intersections by using the final model. According to the invention, CNN and Transform are respectively utilized to extract space and time features of the associated multiple intersections. A learnable time code is embedded into a position code of a Transform, position information and time information are jointly injected into a model, and the model is helped to better learn time characteristics of traffic volume.

    本发明公开了一种关联多交叉口交通流量预测方法。本发明步骤如下:1、收集关联交叉口交通流量数据,对交叉口车流数据进行预处理后划分为训练集、验证集和测试集;2、使用CNN对输入的交叉口车流数据提取空间特征;3、将空间特征作为输入,使用Transformer提取时间特征;4、当Decoder层全部执行完毕后,三个时间窗口数据最终分别输三个向量,三个向量堆叠后输入到平均池化层;5、设置模型参数;6、训练模型直至最大训练周期,使用最终模型对关联多交叉口做交通流量预测任务。本发明分别利用CNN和Transformer提取关联多交叉口的空间和时间特征。使用可学习的时间编码嵌入Transformer的位置编码,将位置信息和时间信息共同注入模型,帮助模型更好地学习到交通量的时间特征。


    Zugriff

    Download


    Exportieren, teilen und zitieren



    Titel :

    Associated multi-intersection traffic flow prediction method


    Weitere Titelangaben:

    一种关联多交叉口交通流量预测方法


    Beteiligte:
    FU TINGTING (Autor:in) / YU QIANWEN (Autor:in)

    Erscheinungsdatum :

    2022-11-18


    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



    Distributed Multi-Intersection Traffic Flow Prediction using Deep Learning

    Moumen Idriss / Mahdaoui Rabie / Raji Fatima Zahra et al. | DOAJ | 2024

    Freier Zugriff

    Traffic flow prediction method and device for traffic intersection

    CAI WEIBIN / GAO WEI / ZHANG JINYUAN et al. | Europäisches Patentamt | 2023

    Freier Zugriff

    Signal intersection traffic prediction method considering intersection channelization and multi-factor correction

    LIAN PEIKUN / BAO RIYONG / ZHANG KANGYI et al. | Europäisches Patentamt | 2023

    Freier Zugriff

    Intersection traffic prediction method and intersection traffic behavior and static prediction method on path

    ZHANG LI / LIANG KAI / YIN PENG et al. | Europäisches Patentamt | 2022

    Freier Zugriff

    Traffic flow prediction method and device for single intersection

    LI SHENGYANG / ZHAO HONGJUN / LI JINZHAO | Europäisches Patentamt | 2022

    Freier Zugriff