To forecast the traffic volume of the operation period highway scientifically and accurately becomes an important part of intelligent transportation system,which can provide key support for the optimization of highway infrastructure.In this paper,the wavelet neural network,GA-BP,SVM,GA-LSSVM and PSO-LSSVM models are used to forecast the short-term traffic vol-ume,and compared with the actual value.The results show that neural network forecasting model is better than support vector machines and their optimization models,which can improve the effectiveness and accuracy.Compared with other models,GA-BP model has the characteristics of high forecasting accuracy and fast convergence speed,which can meet the needs of highway oper-ation management and the intelligent transportation.
维普
万方数据
