Analysis And Modeling Of Highway Traffic Generation Considering Spatial Effects

With the continuous progress of the new stage of economy and the construction of urban agglomerations,profound and complex changes are taking place in the mechanism of highway traffic generation and socioeconomic factors.This paper focuses on the relationship between traffic volume and social development of highway construction under the new normal and urban agglomeration,and analyzes the relationship between traffic generation volume and socio-economic analysis by combining socio-economic factors and considering spatial effects.This article collected data on the generation of highway traffic generation volumes 17 cities in Shandong Province from 2008 to 2017,combined with the socio-economic index considering the new stage of economy.In the study,the principal component analysis method was used to extract socio-economic data which can represent the level of economy growth,industrial structure and citizen living standard for each city.The principle components were used as explanatory variables to analyze the relationship between passenger and freight traffic generation and social economy using linear regression models and panel data models.The fitting extents of the linear regression model and the panel data model of the traffic generation volume are relatively high.Both results show that the regional economic development factors and industrial structure factors have a significant impact on the passenger and freight traffic generation volume.With the continuous development of urban agglomerations,the regional highway traffic volume will be affected by the social and traffic characteristics of the surrounding areas,which can’t be considered in the linear regression models and panel data models.Therefore,the study establishes a spatial panel data model that considers the spatial effect of the panel data model for passenger traffic and freight traffic by adding the spatial weight matrix.The results of these models show that the passenger traffic generation are influenced by the development factors,industrial structure factors.The industrial structure factors and the traffic generation from surrounding cities also has effects.The freight traffic generation are highly connected with the industrial structure factors and the freight traffic generation from surrounding cities.Compared with linear regression model and panel data model,the fitting degree of spatial panel data model has been furtherly improved.The study used the 2008-2016 data to re-establish above models for predicting2017 data traffic generation data and calculating its mean absolute percentage error(MAPE).The error results show that the prediction accuracy of panel data model and spatial panel data model is significantly improved than inear regression models.This paper grasps the development characteristics of Chinese new stage of economy and urban agglomeration,collects data on highway traffic volume and socio-economic data,combines multiple types of data models to analyze the relationship between socio-economic and highway traffic volume,and forms a forecast method for traffic volume It provides important data support for the development and benefit evaluation of expressways.