The invention discloses a traffic flow prediction method based on a Transform adaptive adversarial graph neural network, and the method comprises the steps: carrying out the embedding operation of a degree embedding algorithm and a distance embedding algorithm on original data, and obtaining traffic flow data; inputting the traffic flow data into a TDN module of a generator to carry out attention mechanism operation to obtain aggregation time information Y; y serves as input of an SDGCN graph neural network to obtain aggregation space information Z, then Z is input into an MLP with two hidden layers to obtain prediction data # imgabs0 # to obtain real data Xr, connection operation is conducted on the prediction data # imgabs1 # and traffic flow data X, the connection operation is recorded as Xp, and Xp and the real data Xr are input into a discriminator together; and calculating loss functions Ld and Lg of the discriminator and the generator, then calculating a loss function L of the whole model, and finally updating parameters of the model through stochastic gradient descent. According to the method, the accuracy of traffic flow prediction and the global consistency of prediction results are improved, and the method has a certain advantage in the aspect of convergence speed.
本发明公开了一种基于Transformer自适应对抗图神经网络的交通流预测方法,包括:将原始数据进行度嵌入算法和距离嵌入算法的嵌入操作,得到交通流数据;将交通流数据输入生成器的TDN模块进行注意力机制操作得到聚合时间信息Y;将Y作为SDGCN图神经网络的输入得到聚合空间信息Z,再将Z输入进一个具有两个隐藏层的MLP得到预测数据#imgabs0#获取真实数据Xr,将预测数据#imgabs1#与交通流数据X进行连接操作记为Xp,并将Xp和真实数据Xr一同输入到鉴别器中;计算鉴别器和生成器的损失函数Ld和Lg,再计算整个模型的损失函数L,最后通过随机梯度下降来更新模型的参数。本发明提高了在交通流预测上的准确性以及预测结果的全局一致性,并且在收敛速度方面具有一定优势。
Traffic flow prediction method based on Transform adaptive adversarial graph neural network
一种基于Transformer自适应对抗图神经网络的交通流预测方法
2024-09-27
Patent
Electronic Resource
Chinese
European Patent Office | 2024
|Traffic flow prediction method based on graph neural network
European Patent Office | 2024
|Node adaptive learning graph attention neural network traffic flow prediction method
European Patent Office | 2024
|Highway traffic flow prediction method based on Transform and graph attention network
European Patent Office | 2023
|