Research on Short-term Traffic Flow Prediction Based on Mutual Learning
Traffic flow prediction is the core of intelligent transportation systems(ITS),with spatiotemporal characteristics being the most important feature.Due to the complex spatial correlations and time dependencies between different roads,the traffic flow prediction has become a challenging task.Currently,prediction methods based on graph convolutional neural networks are further op-timized on the feature perception and extraction of local and global networks.To address above issues,a diffusion mutual convolution-al recurrent neural network(DMCRNN)optimized model based on graph neural networks is proposed.The model is based on the DCRNN as a benchmark model,and the mutual learning strategy is utilized to optimize it.During training,two DCRNN networks learn from and guide each other to enhance their respective feature learning capabilities.The effectiveness of the optimization strategy is verified on two real datasets of the METR-LA and PEMS-BAY.The results show that the optimized model significantly reduces the prediction errors,with a decrease in the MAE of 0.15 and 0.12 for one hour on the two datasets than that in the DCRNN,respective-ly,indicating that the mutual learning optimization strategy has a good performance.
万方数据
维普
