The Study On Perspective Traffic Volume Prediction Model Of The New Expressway

Traffic demand forecasting is critical to investment planning and operation decision of transportation infrastructure.For new expressways,perspective traffic volume prediction is a crucial part of engineering feasibility study and post-construction evaluation.It not only determines the construction standards of expressways,such as project scale and technical level,but also provides a basis for toll management after the project is opened to traffic.Therefore,it is vital to predict the perspective traffic volume scientifically and reasonably.However,with the completion and opening of expressways in China,the difference between the actual traffic volume and the predicted results in the engineering feasibility study becomes significant gradually,which not only decreases the return on investment of the project,but also affects the highway network planning.Hence,the incorrectness on highway perspective traffic volume predication is studied in this paper.Concerning this phenomenon,the analysis of the causes of incorrectness should be carried out scientifically and reasonably,which provides a basis for improving the prediction method in the engineering feasibility study.Therefore,the generation mechanism of perspective traffic volume is summarized,then the highway traffic volume prediction in Shandong province is taken as an example to analyze the shortcomings of the current prediction methods.From the perspective of the relationship between the social and economic level and the development of transportation,the method of traffic generation prediction in engineering feasibility study is improved for decreasing the prediction incorrectness of the perspective traffic volume.The social and economic indicators that affect the traffic generation are determined through comprehensive consideration of the national economy,the industrial structure and the transportation development.On this basis,the relationship between the explanatory variables and the traffic generation is constructed.Meanwhile,the annual average daily traffic forecasting methods of the highway are compared,support vector regression model is selected in this study finally,and the genetic algorithm is utilized to optimize the model parameters.In the case study,the traffic statistics of toll stations in Shandong province and the social and economic data over the years are adopted to forecast the traffic generation,which verifies the feasibility and reliability of the model.At last,in order to quantify the impact of social and economic development on the change of traffic volume,the ‘elastic coefficient method’ in the engineering feasibility study is developed by the ‘comprehensive elasticity coefficient’ in this research,which is utilized to calculate the growth rate of traffic volume in traffic zones.It reflects the social and economic factors comprehensively that influence the perspective traffic volume and has practical significances in engineering practice.