Analysis Of Influencing Factors Of China’s New Energy Vehicle Ownership And Research On Its Forecasting Methods

With the rapid development of automobile industry,my country’s carbon emissions are also increasing,and energy dependence far exceeds the international police line.The popularization of new energy vehicles can promote the transportation industry to reduce the demand for oil and improve the efficiency of energy use.In recent years,the improvement of charging facilities and the implementation of subsidy policies have made new energy vehicles appear to replace fuel vehicles.However,due to the short development history of new energy vehicles,the low level of refinement of data,and poor data availability,scholars mostly rely on subjective judgments when conducting research on the development trend of the new energy vehicle industry,and seldom build on prediction models,etc.Conduct research from the perspective of empirical analysis.In order to accurately grasp market conditions and long-term development trends,and provide a basis for planning and formulating corporate activities such as production and financing,it is necessary to build a new energy vehicle monthly sales forecast model.In order to help government policy makers detect environmental pressures and provide a theoretical basis for urban transportation planning,it is also necessary to build a forecast model for the annual number of new energy vehicles.For these reasons,this article first analyzes the development status of the new energy automobile industry,and deeply explores the obstacles in the current stage of the development of new energy automobiles;Secondly,in order to realize the forecast research on the monthly sales of new energy vehicles,this paper carefully selects the input data and preprocesses the data,constructs a new energy vehicle monthly sales forecast model based on LSSVM,and finally completes the simulation of the model based on the Matlab2018b environment analysis.The results show that the built model can meet the forecast accuracy requirements of new energy vehicles in the early stage of development.But considering that the model still has shortcomings in parameter selection,this paper chooses PSO algorithm to optimize the two parameters,and constructs a new energy vehicle monthly sales forecast model based on PSO-LSSVM.Then,the superiority of PSO LSSVM model is proved by error comparison analysis.Thirdly,this paper theoretically analyzes the influencing factors of new energy vehicles from the perspectives of economy,policy,technology,population,resources and environment,and finally extracts 26 characteristic indexes to construct the characteristic index system of influencing factors of new energy vehicle ownership.In order to eliminate the correlation between the indicators and simplify the model,this paper selected the grey correlation analysis method to screen and sort the pre-selected indicators,and then picked out 8 factors with correlation degree greater than 0.85 to build the new energy vehicle ownership prediction index system.Finally,this paper establishes GRA-GM(1,8)model based on grey system theory.After that,the prediction accuracy of the model is tested by posterior difference test and relative error test.The results show that the prediction model performs well and has a good application prospect.In summary,the two prediction models established in this paper can meet the requirements of new energy vehicles’ prediction accuracy in the early stages of development,and the models are versatile and dynamic.Subsequent,the input vector of the prediction model can be continuously updated to achieve rolling prediction,improve the prediction accuracy,and save the complicated repeated calculation process for subsequent model applications.