Short-Time Traffic Flow Prediction of Highway Toll Station Based on Combined GRU-MLP Model

In this paper, based on the traffic flow data of a toll station on a highway in Fujian Province for 7 days, the combined GRU-MLP model is used to predict the short-time traffic flow at a time interval of 15 min, and compared with the single models such as SVM, LSTM, GRU, MLP, etc., and the 4 evaluation indexes of MAE, MAPE, RMSE, and R2 are adopted to analyze the 5 models, and the results show that the paper The proposed GRU-MLP combination model performs better on the four evaluation indexes, in which the MAE value is reduced by 14.25%, the MAPE value is reduced by 36.61%, the RMSE value is reduced by 8.16%, and the R2 value is improved by 0.73%, and the combination model is better than the other four models in terms of prediction, which verifies the superiority of the combination model, and it is able to provide the highway management department with better It verifies the superiority of the combined model and can provide better decision support for highway management.