Research On China’s New Energy Vechicle Markrt Analysis And Sales Forecast Based On Data Mining

With the rapid development of China’s economy,the living standard of the residents has been greatly improved,and the automobile has entered into thousands of households and become the main means of transportation.However,with the rapid economic development,urban environmental pollution and energy shortage are aggravated.Therefore,new-energy vehicles（NEV）have been regarded by the country as an important direction for the future development of the automobile industry.This thesis finds out what new energy vehicles users focus on by mining and analyzing the market information of new energy vehicles.By analyzing users’comments on vehicle use,this thesis obtains the regional differences,user characteristics,user concerns and other information of China’s new energy vehicles market.Based on sales data,the time series model of vehicle sales forecast is established.Firstly,Analyze the development of new energy vehicles at home and abroad,and describes the new energy vehicle market.According to the analysis,new-energy vehicles account for the largest proportion in first-tier cities such as Beijing,Shanghai and Guangzhou.These regions have considerable economic,transport and policy advantages.From the user analysis results,it can be seen that the car purchase is mainly used to pick up children,urban transportation and other short-distance driving.Secondly,150 types of vehicle models were clustered,100 types of pure electric vehicles and 50 types of plug-in hybrid electric vehicles were clustered into four types separately.According to configuration,price and other factors,pure electric vehicles are divided into four categories:high-end,medium and high-end,general and low-end.Plug-in hybrid cars are grouped into four categories,depending on the size of the vehicle.Each model has its own advantages and advantages.Users can choose different models according to their actual needs.Then use Naive Bayes classification method to divide the user comments into two aspects:satisfaction and dissatisfaction,and analyze the specific reasons with the LDA theme.It can be seen that users are satisfied with the appearance,power consumption,range and environmental protection of the car,but not with the interior,detailed configuration and space.Finally,a time series research is made on the sales of new energy vehicles from March 2016 to December 2019.It was observed that vehicles sales increased year by year and were affected by seasons.Vehicles sales peaked at the end of the year and bottomed out at the start of the year.After determining the influencing factors of time series,SARIMA（1,0,0）（0,1,0）₁₂ model was established.The model is used to predict the future sales of new energy vehicles.It can be seen from the result that,except the influence of external factors such as policy factors,the real values are all within the 90%prediction range,and the prediction effect is better.The prediction model can be used as a reference for manufacturers.