| According to the subject about civilian vehicle population forecasting in automobile industry strategy of Development and Reform Commission of Dadong District in Shenyang, Liaoning province,(No.70472032), this paper analyses dynamic combination forecasting method and put it into practice.Traditional combination model only combines with quantitative forecasting models according to deviation, and ignores the information from the time and the judgment from expert. When it is facing to the diversified environment, the traditional combination can not solve the complex forecasting problems.Due to the limitations of combination forecast models, this paper takes China’s civilian vehicle population forecast for an example, and carries on four methods to research:gray model, multiple regression model, three points model and expert forecasting model. Under the idea of comprehensive evaluation, the paper makes comprehensive evaluation on the selected forecasting models, and calculates the dynamic index signs. The dynamic combination forecasting method is constructed which contains the features about information integration of objectivity and subjectivity. Finally, a calculation example explains the application of this method. |
