Combined Forecasting Model of Traffic Flow Based on DGM(1,1) and GRNN

Abstract

In order to improve the prediction accuracy of traffic flow, this paper proposes a combined forecasting models with residual correction based on DGM(1,1). After introducing the modeling steps of DGM(1,1), DGM-GRNN combination model is established. In the model, the GRNN is used to predict the residual of DGM(1,1) model. Finally, we sum up the predict result of DGM(1,1) model and the residual correction model, and get the final prediction result of the combined forecasting model. We predicted the traffic flow of shangsan expressway by the model in this paper and then compared the results with the experimental results of DGM(1,1). The validity and feasibility of the proposed method are verified.