Freight volume forecasting method based on Gray Neural Network model

One of the key preconditions for setting down long-term transport planning and doing reasonable transport investments scientifically is forecasting volume of freight reliably. This paper builds Grey Neural Network model according to characteristics of grey model GM(1,1) and artificial neural networks (ANN) model. Grey Neural model overcomes the shortcoming of singularity forecasting model, thereby it increases the reliability level of forecasting results. The paper verifies Grey Neural Network model though comparing with raw statistics date of calendar year in Jilin province, and it also forecasts volume of freight for future. The results show that forecasting precision of the model is high and it shows much faster convergence rate.