Forecasting Model Of Car Ownership Based On Multivariate Statistical Analysis

The car ownership means the number of car which registered officially in an area. With the development of economy and increasing per capita income in China, the level of car ownership has also been increased in recent years. However, China encounter the serious burden of traffic jam, the lack of energy and environment pollution because of the increasing number of cars. Moreover, these problems will be more serious though the level of the car ownership in China is lower than that in other developed countries.A prediction model of the car ownership based on the multiple linear regression is surveyed in this paper to ensure the electric-vehicle industry and the economy keep in the same pace. Firstly, the elements which could affect the car ownership are selected. This kind of data is obtained from the China Statistical Yearbook, various years, the National Bureau of Statistics and China communique of the state of environment. For element includes two or more measures, the Principal Component Analysis(PCA) is used to reduce the dimensionality of the original data. Stepwise regression method is then used to input the data of elements. At this step, multicollinearity is avoided and the multiple linear regression model is conducted. Finally, the conducted model is utilized to predict the tendency of the car ownership in the later ten years. Meanwhile, corresponding measures for field of energy source, transport, environment pollution and the electric- vehicle industrial structural adjustment resulted from the increasing of the car ownership are proposed based on prediction results.