An Improved Discrete Grey Model Based on BP Neural Network for Traffic Flow Forecasting

The forecasting of traffic flow is an important part of intelligent transportation system; actual and accurate forecasting of traffic flow can give scientific support for urban traffic guidance and control. As there is big forecast error when modeling toward traffic flow data with discrete grey model DGM (1, 1), this paper amends the equal interval time sequence. According to the characteristic of time coefficient and backpropagation (BP) neural network, we propose an improved grey model by combining DGM (1, 1) model with BP neural network model. The experimental result indicates that the improved grey model is scientific and effective for the forecasting of traffic flow.