Sales Forecast of Chinese New Energy Vehicles Based on Wavelet and BP Neural Network

Abstract

In recent years, new energy vehicles have occupied more and more market shares. If its monthly sales are predicted, it can provide reference for the production of enterprises and the formulation of national policies. However, the current prediction model does not consider the various factors that affect the sales, and the prediction is not accurate enough. Firstly, according to the periodicity of monthly sales, the wavelet neural network prediction model is established in this paper, and compared with the actual value, the result is ideal. The BP neural network prediction model is established to obtain more accurate prediction results by comprehensively considering four factors, including season, oil price change, technologies progress of new energy vehicles and government policy. The results show that the average error of prediction is 5%, which has good application value.