The invention provides a traffic flow prediction method based on multi-scale spatial feature mining. According to the method, a multi-scale spatial relationship capture and attention mechanism-dependent spatial correlation weight model is introduced to capture an omnibearing spatial dependency relationship in a traffic flow. According to the method, the spatial-temporal feature extraction capability of a model is expanded step by step through a time convolution network and a graph convolution gating mechanism, and an associated feature relationship in a historical traffic spatial-temporal flow is fully extracted. In the solving process of the prediction problem, spatial features are extracted through inter-hierarchy graph convolution network operation, a multi-scale spatial relationship is constructed by adopting graph convolution networks of different hierarchies, and the spatial relationship changing along with time is modeled. And finally, aggregating layer-by-layer hidden features through an attention mechanism, and modeling historical spatio-temporal features from the perspective of global features, thereby more accurately predicting future traffic conditions.
本发明提出了一种基于多尺度空间特征挖掘的交通流量预测方法。该方法引入多尺度空间关系捕捉和依赖于注意力机制的空间关联权重模型来抓取交通流中全方位的空间依赖关系。该方法通过时间卷积网络和图卷积门控机制来逐级扩充模型的时空特征提取能力,充分提取历史交通时空流中的关联特征关系。在预测问题的求解过程中,通过层级间的图卷积网络操作提取空间特征,采用不同层级的图卷积网络构建多尺度的空间关系,建模随时间变化的空间关系。最后,通过注意力机制将逐层隐藏特征进行聚合,从全局特征的角度建模历史时空特征,从而更加精准地预测未来的交通状况。
Traffic flow prediction method based on multi-scale spatial feature mining
一种基于多尺度空间特征挖掘的交通流量预测方法
2023-05-05
Patent
Elektronische Ressource
Chinesisch
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 mixed feature mining
Europäisches Patentamt | 2020
|Multi-scale traffic flow prediction model and method based on contextual feature injection
Europäisches Patentamt | 2020
|Traffic flow prediction method based on multi-feature fusion
Europäisches Patentamt | 2024
|Time-space traffic flow prediction method based on hidden time-space feature mining
Europäisches Patentamt | 2024
|Europäisches Patentamt | 2023
|