The invention discloses a traffic flow prediction method for traffic flow spatio-temporal data information. According to the invention, a prediction model based on an attention mechanism is provided for a traffic flow prediction task, and the prediction precision is improved. According to the method, the time features are pre-learned by using a cross attention mechanism of Transform and an original feature extraction block, and the time features and the space features in a traffic flow task are more fully fused by intersecting with a previous model; according to the method, the result is effective, high performance and generalization ability are shown on multiple data sets, the accuracy of traffic flow prediction is further improved, and a new thought is provided for intelligent traffic.
本发明公开了一种面向交通流时空数据信息的交通流预测方法。本发明针对交通流预测任务提出了一种基于注意力机制的预测模型,实现了预测精准度上的提升。本发明利用Transformer的交叉注意力机制和独创的特征提取块对时间特征进行预学习,相交于以往模型更充分地融合了交通流任务中的时间特征和空间特征;本发明结果有效,在多个数据集上面展现出较强性能和泛化能力,进一步提高交通流预测的准确率,为智慧交通提供了一种新思路。
Traffic flow prediction method for traffic flow spatio-temporal data information
一种面向交通流时空数据信息的交通流预测方法
2024-12-24
Patent
Electronic Resource
Chinese
IPC: | G08G Anlagen zur Steuerung, Regelung oder Überwachung des Verkehrs , TRAFFIC CONTROL SYSTEMS |
Spatio-Temporal AutoEncoder for Traffic Flow Prediction
IEEE | 2023
|Traffic flow spatio-temporal data prediction method, system, device and medium
European Patent Office | 2023
|Short-term traffic flow prediction method based on spatio-temporal correlation
European Patent Office | 2020
|Traffic flow prediction method based on spatio-temporal information fusion and storage medium
European Patent Office | 2024
|Global spatio‐temporal dynamic capturing network‐based traffic flow prediction
DOAJ | 2023
|