The invention discloses a traffic flow prediction model method based on a discrete graph structure, and belongs to the technical field of traffic management, and the method comprises the steps: 1, collecting data; 2, preprocessing the data; 3, building and optimizing a model; 4, performing model training; 5, deploying the model; according to the method, under the GNN framework, the structure of the graph and prediction of the time sequence are learned at the same time, the method is suitable for the situation that the graph structure is unknown, learning is achieved by optimizing performance distribution of the graph model, differential re-parameterization sampling of the discrete graph structure is achieved through neural network parameterization, and the method has the advantages of being high in robustness, high in robustness and the like. And the time expandability, the prediction accuracy and the calculation efficiency are improved.
本发明公开了一种基于离散图结构的交通流预测模型方法,属于交通管理技术领域,包括步骤一、采集数据;步骤二、进行数据的预处理;步骤三、进行模型的搭建与优化;步骤四、进行模型的训练;步骤五、进行模型的部署;步骤六、进行模型的预测及使用,本发明,在GNN框架下,同时学习图的结构和时间序列的预测,适用于图结构未知的情况,通过优化图模型的性能分布实现学习,通过神经网络参数化,实现对离散图结构的可微分重参数化采样,提高时间上的可扩展性、预测准确性以及计算效率。
Traffic flow prediction model method based on discrete graph structure
一种基于离散图结构的交通流预测模型方法
09.08.2024
Patent
Elektronische Ressource
Chinesisch
IPC: | G08G Anlagen zur Steuerung, Regelung oder Überwachung des Verkehrs , TRAFFIC CONTROL SYSTEMS |
Traffic flow prediction method based on regional graph structure
Europäisches Patentamt | 2025
|Traffic flow prediction method based on graph neural network
Europäisches Patentamt | 2024
|Traffic flow prediction method based on graph convolutional network
Europäisches Patentamt | 2025
|Traffic flow prediction method based on double graph convolution
Europäisches Patentamt | 2024
|Traffic flow prediction method based on interactive space enhanced graph convolution model
Europäisches Patentamt | 2024
|