The Research On Sales Forecasting Models And Methods For Chinese Automobile Market

The automobile industry is the propellant industry and plays a positive role in driving thevigorous development of other industries,therefore,playing a significant role in Chinese economic development.At present,China has developed into the largest automobile market in the world.Although automobile has great market demand,Chinese new automobile market and automobile aftermarket are changing and become unstable with the changes of most consumers' mindsets,coupled with the slowdown of macroeconomic growth and the fierce market competition.In this case,the accurate sales forecasting will provide enterprises with strong data support,which help enterprises to formulate short-term and long-term planning and grab chances in the market.Therefore,the paper has certain theoretical and practical significance by using different forecasting methods to sales forecasting for Chinese automobile market.First of all,a study was conducted on the relatively macroscopic annual automobile sales in China.The main factors affecting Chinese automobile sales are selected and related forecasting models are established.And a hybrid intelligence algorithm integrating particle swarm optimization and ant colony optimization is used to optimize the weight coefficients in the forecasting model.Then the forecasting results are obtained and the performance analysis is performed.The performance analysis of the forecasting results shows that the annual sales forecasting model and algorithm proposed in this paper have advantages.Secondly,the paper takes the long-term forecasting of monthly sales as the main line based on the econometric models,and the automobile sales forecasting method based on structural relationship recognition is proposed.The specific work is divided into three parts:first,relevant economic variables through the structural relationships identification are selected to forecast Chinese automobile sales.Then,domestic brands sales and related economic variables through the structural relationships identification are used as explanatory variables to forecast Chinese automobiles sales.Finally,The brand automobile sales forecasting is focused,related brands sales and related economic variables are selected as explanatory variables through structural relationship recognition to further forecasting.In the empirical analysis,through the analysis of the forecasting results,it can be found that the proposed model and method have higher forecasting accuracy in terms of long-termforecasting.Finally,the relatively microscopic automobile spare parts sales forecasting in China is focused.Aiming at the variety of existing spare parts,difficulties in classification and inaccurate forecasting,a unified and dynamic model for automobile spare parts sales forecasting is proposed.The model integrates five classical forecasting models and expands the model's adaptability to different kinds of spare parts forecasting,then heuristic intelligence algorithm is used to solve the weight coefficients of the combined model,greatly improving the forecasting accuracy.Through the analysis of forecasting results,it can be shown that the model and method have wide applicability,effectiveness and certain advantages for the automobile spare parts sales forecasting in China.Through theoretical analysis,different forecasting methods are applied to different sales forecasting,and in the comparative analysis of forecasting results,validity and superiority of the proposed forecasting model are further proved.