Graph convolution and multipath neural network prediction method for traffic prediction

The invention discloses a graph convolution and multipath neural network prediction method for traffic prediction, and the method comprises the steps: employing an improved graph convolution neural network for learning the spatial correlation of traffic conditions; secondly, a multi-path convolutional neural network model for increasing an attention mechanism is used for learning time correlation of traffic conditions; and the traffic flow of the urban road network is accurately predicted. The method can effectively predict the temporal and spatial change characteristics and rules of the traffic flow, is high in prediction precision, and improves the traffic flow prediction effect. 本发明公布了一种用于交通预测的图卷积和多路径神经网络预测方法，使用改进的图卷积神经网络，用于学习交通状况的空间相关性；其次，使用一个增加注意力机制的多路径卷积神经网络模型，用于学习交通状况的时间相关性。准确预测城市道路网络的交通流。本发明方法能够有效预测交通流的时空变化特征和规律，预测精度高，提升了交通流预测效果。