The invention belongs to a traffic flow parameter prediction method, and particularly relates to a traffic flow prediction method based on combination of empirical mode decomposition and deep learning. Comprising the following steps: performing empirical mode decomposition on a collected traffic flow data time sequence to obtain a plurality of eigenmode function component sequences; forming a cross-scale sub-sequence of each component sequence, and sequentially splicing the cross-scale sub-sequences line by line to form an input matrix; inputting the input matrix into a convolutional neural network model with an attention mechanism to extract time sequence features, multiplying the output feature matrix and the saliency feature matrix element by element to obtain a fusion feature matrix, reducing the fusion feature matrix into fusion feature sequences of each component line by line, and taking each fusion feature sequence as the input of a long-short-term memory network to obtain a long-short-term memory network; performing time sequence prediction of each component; and reconstructing the time sequence prediction result of each component into a final prediction result. The technical problem that a traditional method is not ideal in effect under the condition that the traffic flow changes slowly in the peak period of traffic is solved.

    本发明属于交通流参数预测方法,具体涉及一种基于经验模态分解与深度学习结合的交通流预测方法。包括对采集到的交通流数据时间序列的进行经验模态分解得到多个本征模函数分量序列;形成各分量序列的跨尺度子序列,依次将跨尺度子序列逐行拼接为输入矩阵;将输入矩阵输入一带有注意力机制的卷积神经网络模型提取时间序列特征,将所述输出特征矩阵与显著性特征矩阵逐元素相乘得到融合特征矩阵,再逐行还原为各分量的融合特征序列,以各融合特征序列作为长短期记忆网络的输入,进行各分量的时间序列预测;将各分量的时间序列预测结果重构为最后的预测结果。解决了传统的方法在交通高峰期时段交通流变化缓慢的情况下,效果不理想的技术问题。


    Access

    Download


    Export, share and cite



    Title :

    Traffic flow prediction method based on combination of empirical mode decomposition and deep learning


    Additional title:

    一种基于经验模态分解与深度学习结合的交通流预测方法


    Contributors:
    LI YONG (author) / WANG SENZHENG (author) / LIU MEIQI (author) / WU JIANGUANG (author) / WU JINZE (author) / LIU RUI (author)

    Publication date :

    2024-01-16


    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 / G06F ELECTRIC DIGITAL DATA PROCESSING , Elektrische digitale Datenverarbeitung / 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




    Space-time convolution short-time traffic flow prediction method based on complementary integrated empirical mode decomposition

    ZHANG LINGYUN / ZHANG KAI / DING YUJIE et al. | European Patent Office | 2022

    Free access

    Multi-node Mode Decomposition Based Deep Learning Model for Road Section Traffic Prediction

    Pholsena, Khouanetheva / Pan, Li / Zheng, Zhenpeng | IEEE | 2019


    Short-term traffic flow prediction method combining data decomposition and deep learning

    ZHANG LINGYUN / ZHANG KAI / DING YUJIE et al. | European Patent Office | 2022

    Free access

    Traffic flow prediction method based on deep learning

    XIE GANG / WANG HAIYING / XIE RUIQI | European Patent Office | 2024

    Free access