A New Forecasting Model for Electrical Vehicle Sales with Accumulative Effects of Multiple Events

An accurate prediction of electric vehicles sales can provide effective guidance for the production and development of the electric vehicle industry. In this paper, the influence factors of the electric vehicles events were analyzed, and the use of double difference model of quantitative research the subsidy policy, back filling, charging infrastructure construction three events in China the influence of the electric car sales as a result, and combining the SARIMA, BASS, BP neural network prediction model to build the integrated forecast model, under the influence of the major events of significance is verified through empirical analysis. The results indicate that the integrated model, considering the impact of major events, significantly outperforms single forecasting models in terms of accuracy; The electric car subsidy policy and charging infrastructure development has a promoting effect on sales, sales increased by 5450 and 7000 respectively; Subsidies to exit the events have a negative impact on sales, sales of 3290 units. This article for the electric cars under the influence of the superposition of multiple events forecast work provides a new thought of forecast, enrich the electric car prediction model system in China, but the electric car industry policies for science under complex conditions provide reliable decision making tool.