The Research On Sales Forecast For New Energy Vehicle Based On SARIMA And BP Neural Network

In recent years,China's automobile industry has developed at a high speed,and its industrial scale has ranked first in the world for many years.Nowadays,the automobile industry has become an important measure of the pillar industries of the national economy and the national economy,in order to avoid the environmental side effects of such industries,the environmental protection for new energy automobile industry has received increasing attention internationally.This paper aims at the problem of industrial upgrading brought about by the changes in the sales volume of new energy vehicles in China,and mathematical statistical models and machine learning algorithms are used for predicting the monthly and annual sales volume of new energy vehicles in China.The accurate forecasts of sales volume on new energy vehicles can provide the support of theoretical statistics and predictive data for government sector to make decision and automotive industry to develop marketing strategies.This paper takes the monthly sales volume data of new energy vehicles in China as a sample.Firstly,the monthly sales volume data of China's new energy vehicles are executed factor decomposition to generate non-random wave sequences and random wave sequences,and a combined model of SARIMA(Seasonal Autoregressive Integrated Moving Average)and BP neural networks are constructed to fit and predict the monthly sales volume of new energy vehicles,simultaneously,for comparison,a single SARIMA model is constructed to fit and predict.Then,from the perspective of fitting and prediction value,the error value and data verification,the predictive effect of single model and combined model is compared.Then,based on the better performance of the combined model,the combined model is used for predicting the monthly sales volume of China's new energy vehicles in 2019,and the reasons and rationality of forecast results are analyzed.In addition,taking the annual sales volume data of new energy vehicles in China as a sample,the gray correlation analysis method is used for analyzing the influencing factors such as sales policy,power battery output,and traditional automobile manufacturing,and the BP neural network model is used for fitting and predicting the annual sales volume of new energy vehicles,at the same time,compared with the traditional ARIMA(Autoregressive Integrated Moving Average)model from the view of fitting value,prediction value,error value,etc..Finally,on the basis of the better performance of BP neural network model,BP neural network model is used for predicting the annual sales volume of China's new energy vehicles in the year of 2019 to 2023,and the prediction results are explained.