The invention provides a road network short-term traffic flow prediction method based on a deep space-time residual network. The method comprises the following steps: respectively designing corresponding residual error network branches for the proximity and periodicity of two unique attributes of space-time data, and dynamically aggregating the outputs of two branch networks by distributing different weights to the same road in the two branches. Therefore, the influence degree of space-time attributes on traffic flow prediction of different road sections is adjusted, and then the aggregation result of the two residual networks is fused with external factors. Experimental verification is carried out by selecting RMSE and R2 as evaluation indexes of the model, and the DST-ResNet model has higher effectiveness and feasibility compared with a mainstream LSTM model.

    本发明提供一种基于深度时空残差网络的路网短时交通流预测方法,针对时空数据的两个独特属性邻近性和周期性分别设计相应的残差网络分支,通过为两个分支中相同的道路分配不同的权重动态聚合两个分支网络的输出,从而调整时空属性对不同路段交通流量预测的影响程度,其次将两个残差网络的聚合结果与外部因素进行融合。通过选择RMSE和R2为模型的评价指标进行了实验验证,该DST‑ResNet模型相较主流的LSTM模型具有更高的有效性和可行性。


    Zugriff

    Download


    Exportieren, teilen und zitieren



    Titel :

    Road network short-term traffic flow prediction method based on deep space-time residual network


    Weitere Titelangaben:

    基于深度时空残差网络的路网短时交通流预测方法


    Beteiligte:
    SHI QUAN (Autor:in) / DING XINYU (Autor:in) / SHI ZHENQUAN (Autor:in) / CAO YANG (Autor:in) / CAO ZHICHAO (Autor:in) / ZHU SENLAI (Autor:in)

    Erscheinungsdatum :

    09.03.2021


    Medientyp :

    Patent


    Format :

    Elektronische Ressource


    Sprache :

    Chinesisch


    Klassifikation :

    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



    Short-term traffic flow prediction of road network based on deep learning

    Han, Lei / Huang, Yi-Shao | IET | 2020

    Freier Zugriff

    Short‐term traffic flow prediction of road network based on deep learning

    Han, Lei / Huang, Yi‐Shao | Wiley | 2020

    Freier Zugriff

    Short-term traffic flow prediction method based on space-time convolutional network

    JIA CHAOLONG / KANG ZHEYI / PENG GANG et al. | Europäisches Patentamt | 2023

    Freier Zugriff

    A deep residual long short-term memory network model for traffic flow prediction

    Shi, Chang / Jiang, Bo / Huang, Zijun et al. | SPIE | 2022


    Traffic flow prediction method of attention-based deep residual space-time diagram convolutional network

    XU SHIJIAN / ZHANG XUHONG / WU YUEDONG et al. | Europäisches Patentamt | 2023

    Freier Zugriff