Transport volume forecast based on GRNN network

As there is a close relationship among transportation, local economy and enterprise development, the forecast of the traffic volume has become an important research project of transport market and economic development. The structure and algorithm of the Generalized Regression Neural Network (GRNN) are induced in this paper. The mathematical background of GRNN network is also described in detail. As a case, a GRNN network is built taking a number of important parameters that affect transport capacity as sample data. After learning and training to meet the minimum error, this network will forecast the future traffic volume. The result demonstrates the effectiveness of using GRNN to forecast transport volume. Finally, the advantages of GRNN network in forecasting the traffic volume are summarized.