OD passenger flow prediction method based on dynamic hypergraph convolutional neural network

The invention discloses an OD passenger flow prediction method based on a dynamic hypergraph convolutional neural network, and relates to the fields of deep learning and the like, in particular to an OD passenger flow prediction task oriented to hypergraph representation and a graph convolutional network. According to the method, on the basis of high-order representation of a traffic topological structure by using a hypergraph, a hypergraph convolutional neural network is introduced, spatial relevance of OD passenger flow on a starting point and a destination is mined through a historical OD matrix to construct a dynamic hyperedge, and dynamic modeling of complex spatial relevance of OD flow is realized. Compared with a traditional mathematical model and a machine learning method, the method is deeper and more accurate in OD flow feature modeling, and the prediction accuracy is improved. 一种基于动态超图卷积神经网络的OD客流预测方法，涉及深度学习等领域，尤其是面向超图表示以及图卷积网络的OD客流预测任务。该方法在利用超图对交通拓扑结构的高阶表示基础上，引入超图卷积神经网络，并通过历史的OD矩阵挖掘OD客流在起始点(origina