Forecast Analysis Of Private Car Ownership In China Based On Dynamic Combination Model

The number of Chinese private car ownership was only 284,900 in 1985.After three decades of development,the number has more than 140 million by the end of 2015.In the foreseeable future,the private car ownership will also remain high-speed and sustained growth.However,a series of urgent social problems,such as traffic congestion,air pollution and other issues is getting worse when the private cars are becoming more and more important for people to travel.Thus,forecasting the number of Chinese private car ownership and the Changing tendency of private car ownership in China has become a hot research topic.Scientific and accurate prediction of private car ownership can help the national decision-making departments to make the appropriate industrial policy adjustment,supply side reform to provide theoretical basis and data support,also help the automotive industry to provide the basis for planning and development direction.This paper focuses on the private car ownership and its influencing factors in China.The purpose of the study is to establish a dynamic combination model to predict the ownership of private cars in China and its development trend,and as a starting point,put forward reasonable suggestions.So,the main research work is as follows: Firstly,the regression model between private car ownership and its main factors is established based on the partial least squares regression theory.At the same time,the private car ownership is predicted,through the prediction of the influence factors.As the second part of the study,a time series seasonal model is established by the private car ownership of the timing data to predict the ownership of private cars.Finally,using the regression combined forecasting method to determine the combination weight,and considering the individual prediction model may change with time,and then the prediction effect will change.So,add the time parameters to change the individual prediction model so that to establish a dynamic combination forecasting model.Thus,a more reasonable prediction of private car ownership in China can be get.This article has two innovations.The one is the introduction of single-variable partial least squares regression method for the analysis of the relationship between private car ownership and its influencing factors.The method can solve the problem of multiple collinearity between influencing factors and maximize the amount of extracted components and private car ownership.The other one is considering the individual prediction model may change with time,and then the prediction effect will change,then the time parameters is added to modify the individual prediction model,so this paper establish the dynamic combination forecasting model.Thus make the research more meaningful than previous.