The invention relates to the technical field of traffic flow prediction, in particular to a highway network multi-task flow prediction method and system based on deep learning. According to the method, the correlation between the inbound traffic and the outbound traffic and the similarity of time features are concerned, the common features of the inbound traffic and the outbound traffic are represented by using a multi-task learning technology, and the represented common features are used as the input of feature fusion, so that the multi-task collaborative prediction of the inbound traffic and the outbound traffic of the highway network target station can be realized; the characteristics of the spatial-temporal correlation of the highway network traffic and the change rule of the influence of different external factors on the characteristics are deeply mined through the deep learning traffic prediction model, and the accuracy of traffic prediction is improved.
本发明涉及交通流量预测技术领域,特别是一种基于深度学习的高速路网多任务流量预测方法及系统;本发明关注进站流量和出站流量的关联以及时间特征的相似性,使用多任务学习技术将两者的共同特征表征出来,表征的共同特征作为特征融合的输入,能实现对高速路网目标站点的进站流量和出站流量的多任务协同预测,通过深度学习流量预测模型深入挖掘高速路网流量时空相关性的特点及不同外部因素对其影响的变化规律,提高了流量预测的准确性。
Highway network multi-task traffic prediction method and system based on deep learning
基于深度学习的高速路网多任务流量预测方法及系统
2024-06-18
Patent
Electronic Resource
Chinese
IPC: | G08G Anlagen zur Steuerung, Regelung oder Überwachung des Verkehrs , TRAFFIC CONTROL SYSTEMS |
Highway short-time traffic flow prediction method based on deep learning
European Patent Office | 2023
|