Study On Short-term Forecast Method Of Highway Toll Revenue

Short-term forecast of highway toll revenue is an important basis for investors to judge the short-term recovery ability of their funds.Also,it is the basis for expressway managers to make business planning and decision,to prepare the expenditure budget and to assess management tasks,especially for the annual,quarterly and monthly operating tasks.Short-term forecast of highway toll revenue has become an important issue in the highway decision-making and management of highway management department and investors,and it is urgent to set up a scientific,accurate and effective method to forecast short-term income of highway tolls.First of all,through the analysis of the main influencing factors of expressway toll revenue,the most reasonable idea is to select the traffic volume or toll revenue as the forecast factor from the perspective of system analysis.Secondly,through the analysis and research on the general forecasting methods of four expressway toll revenue,it is concluded that they are more suitable for the long-term forecast of toll revenue but are not suitable for short-term forecast.Thirdly,constrcuting the nonlinear autoregressive prediction method and model based on Levenberg-Marquardt algorithm according to the toll revenue characteristics and short-term forecasting requirements,and describing the combination form and construction process of the model.Finally,taking the NH Highway,GH Highway and GX Highway as examples to forecast annual,quarterly and monthly short-term toll revenue by MATLAB software,and drawing the conclusion that the model proposed in this paper is in accordance with the prediction accuracy.In this paper,the non-linear autoregressive model based on Levenberg-Marquardt algorithm is used to meet the requirements of accuracy,high efficiency and practicality of short-term highway toll revenue forecast,this paper provides a effective short-term forecast method of toll revenue for highway management department and investors.