Model of Traffic Volume Forecasting Based on Fuzzy GMDH Network

In view of the undesirable forecasting results from the conventional group method of data handling(GMDH) network modeling that uses the least square method to identify coefficients,this paper introduces a fuzzy reasoning model into the GMDH network to take the place of partial descriptions of the traditional GMDH network,i.e.the integral binary quadratic polynomial forms.Based on the GMDH network,this paper proposes a model of traffic volume forecasting based on a fuzzy GMDH network.Simulation results show that the average relative error of the new model prediction is only 2.31%,and it is less than 3.35% which is the average relative error of prediction of the traditional GMDH network.The new model is proved effective.