公路客运量的回归分析和研究预测

Regression Analysis and Prediction of Highway Passenger Volume

摘要: 随着我国经济的发展、国民生活水平的日益提高，我国对公路交通基础建设也越加重视。本文利用我国1981~2015年客运量各项有关数据，对公路客运量影响因素进行分析，以总人口、国内生产总值、农业生产总值以及民用载客汽车拥有量为自变量，公路客运量作为因变量。运用MATLAB建立多元回归模型，并对回归模型的合理性进行了分析。其中多元回归模型综合考虑了影响客运量的众多因素，对这些因素进行模型拟合，并用逐步回归法对模型进行了检验和改进，大大提高了模型的科学性和准确性。最后运用所得的回归模型拟合2011~2015公路客运量，并与实际客运量数据进行对比，分析误差。

Abstract: With the development of economy and the increase of living standard, China has paid more attention on the infrastructure of highway transportation. This paper applies the related data of the passenger capacity of the highway transportation from 1981 to 2015 to analyze the factors influenced the passenger capacity of the highway transportation. It takes the population, GDP, agricultural GDP and the civil car ownership as the independent variables, and the passenger capacity of the highway transportation as the dependent variable to establish the multivariate regression model with MATLAB. At the same time, the rationality of the regress model is also analyzed in this essay. We have fitted various factors of the multivariate regress model in consideration of the complexity, verified and improved the multivariate regress model with stepwise regression in order to enhance the scientificity and accuracy of the model. In the final step, the passenger capacity of the highway transportation from 2011 to 2015 has been calculated by the regress model of univariate and multivariate, and a comparison has been made between the calculated date and the actual data which aims to analyze the difference and errors.