The invention discloses a Conv1D + LSTM-based multi-step traffic flow prediction method, and the method comprises the steps of collecting available traffic flow data, and carrying out the preliminary screening; performing data preprocessing on the screened data, for example, data abnormal points are repaired, noise existing in the data is removed, and normalization processing is carried out on the data; outputting the processed data to a one-dimensional convolution Conv1D to extract the characteristics of the traffic flow, and extracting the influence of external factors on the traffic flow at the same time; inputting the features extracted by the Conv1D layer into an LSTM layer for learning so as to predict traffic flow values of a plurality of time intervals in the future; and carrying out reverse normalization processing on the prediction value generated by the model. According to the method, external factors such as weather, time information and holidays and festivals are combined, Conv1D is used for modeling time features, periodic features and local related features of traffic flow data, and the extracted features are sent to LSTM for multi-step prediction. According to the invention, the influence of external factors on the traffic flow is fully considered, so that the multi-step prediction of the traffic flow is effectively realized.

    本发明公开了一种基于Conv1D+LSTM的多步交通流量预测方法,其步骤是:收集可用的交通流量数据,并进行初步地筛选;对筛选好的数据进行数据预处理,例如对数据异常点进行修复,去除数据中存在的噪声和对数据进行归一化处理等;将处理好的数据输到一维卷积Conv1D中提取交通流的特征,同时提取外部因素对交通流量的影响;将Conv1D层提取的特征输入到LSTM层进行学习,以预测未来多个时间间隔的交通流量值;反归一化处理模型产生的预测值。本方法结合天气、时间信息和节假日等外部因素,利用Conv1D来建模交通流数据的时间特征、周期特征和局部相关特征,将提取的特征送入LSTM进行多步预测。本发明通过充分考虑外部因素对交通流量的影响,有效地实现了对交通流量的多步预测。


    Access

    Download


    Export, share and cite



    Title :

    Conv1D + LSTM-based multi-step traffic flow prediction method


    Additional title:

    一种基于Conv1D+LSTM的多步交通流量预测方法


    Contributors:
    ZHENG WEN (author) / JIAO LELE (author) / ZHANG YUE (author) / KANG LUOYAO (author)

    Publication date :

    2021-07-23


    Type of media :

    Patent


    Type of material :

    Electronic Resource


    Language :

    Chinese


    Classification :

    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



    Short-term traffic flow prediction method based on Conv1D-LSTM model

    ZHANG ZHIPENG / LIU YUHANG / DAI LEI et al. | European Patent Office | 2023

    Free access

    Intersection traffic flow prediction method based on LSTM

    ZHANG HUI / LI ZHAOCHUAN / WANG GUANJUN et al. | European Patent Office | 2024

    Free access

    Traffic flow prediction method based on LSTM-Attention

    QIN XIAOLIN / LIU JIACHEN / SONG LIXIANG et al. | European Patent Office | 2021

    Free access

    Instant traffic flow prediction method based on Conv-LSTM

    WU LIJUN | European Patent Office | 2023

    Free access

    Short-term traffic flow prediction method based on improved LSTM

    GAN YONGHUA / JIANG XUEFENG / HU JINGSONG et al. | European Patent Office | 2021

    Free access