The Prediction Model of Highway Network Scale Based on Traffic Demand

The purpose of regional highway network construction is to meet the traffic demand, so the highway network scale should be suitable to the demand. The indexes of highway network scale and traffic demand are selected in this paper, and the Granger causality test is adopted to sift the indexes based on the analysis of relationship between highway network scale and traffic demand. The BP neural network prediction model is based on the filtered indexes. The data of developed countries are used as No.1 training sample while the data of fourteen provinces of China are used as No.2 training sample. The operation procedure is designed by MATLAB. Finally, it takes An-hui province as an illustration.