Convolutional network traffic flow prediction method based on space-time attention mechanism

The invention discloses a convolutional network traffic flow prediction method based on a space-time attention mechanism, and the method mainly comprises the steps: carrying out the modeling of the periodicity, spatial correlation and time dependence of a traffic flow through a space-time attention STA-Block, a graph convolutional network GCN and a standard convolutional network CN; performing modeling on space-time correlation between different time steps through a space-time attention mechanism and a gating fusion mechanism by STA-Block, and using GCN and CN to respectively capture space features and time features of traffic flow; and finally, predicting the output of the three components through the gating fusion mechanism; constructing an STAGCN method by combining the space-time attention mechanism and the space-time convolutional network, and inputting n historical time sequence traffic data into the STAGCN method to obtain n hidden states with space-time features. 一种基于时空注意力机制的卷积网络交通流预测方法，所述的交通流组合预测方法主要