The invention relates to a traffic flow prediction method based on a space-time diagram wavelet convolutional neural network. Comprising the steps of S1, preprocessing traffic flow data; s2, establishing a traffic flow prediction model; s3, training a traffic flow prediction model and adjusting and optimizing parameters; and S4, traffic flow prediction is carried out based on the traffic flow prediction model of parameter tuning. The model established by the method has the characteristics that structurally, the STGWCNN model adopts a'sandwich 'structure with two large ends and a small middle to fully apply a bottleneck strategy, and a graph wavelet convolutional neural network layer is used for performing down-scaling and up-scaling on a channel to realize scale compression and feature compression. Functionally, a graph wavelet convolutional neural network layer designed by an STGWCNN model captures a spatial topological structure of time-space sequence data, local and sparse feature expression can be effectively learned, and meanwhile, the expression effect and flexibility of the network are improved; the time gating convolution layer can well achieve the purpose of extracting the time dependence of the space-time sequence through gating linear unit stacking causal convolution.
本发明涉及一种基于时空图小波卷积神经网络的交通流预测方法。包括:S1、交通流数据预处理;S2、交通流预测模型的建立;S3、交通流预测模型训练及参数调优;S4、基于参数调优的交通流预测模型进行交通流预测。本发明方法建立模型具有特点:结构上,STGWCNN模型采用两头大中间小的“三明治”结构以充分应用瓶颈策略,由图小波卷积神经网络层对通道进行下缩放和上缩放,实现尺度压缩和特征压缩。功能上,STGWCNN模型设计的图小波卷积神经网络层捕捉时空序列数据的空间拓扑结构,可以有效地学习局部化的、稀疏的特征表达,同时提升网络的表达效果和网络的灵活性;时间门控卷积层通过门控线性单元堆叠因果卷积,能够较好实现提取时空序列的时间依赖性目的。
Traffic flow prediction method based on space-time diagram wavelet convolutional neural network
一种基于时空图小波卷积神经网络的交通流预测方法
2022-06-14
Patent
Electronic Resource
Chinese
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 |
Traffic flow prediction method based on improved space-time diagram convolutional neural network
European Patent Office | 2025
|Traffic flow prediction method based on space-time diagram convolutional network
European Patent Office | 2022
|Traffic flow prediction method and system based on space-time diagram convolutional neural network
European Patent Office | 2023
|Traffic flow prediction method based on space-time diagram convolutional network
European Patent Office | 2023
|Traffic flow prediction method of space-time diagram convolutional network
European Patent Office | 2024
|