Short-term Combined Prediction Model Of Electric Vehicle Load Based On Data Mining Technology

In recent years,in the process of continuous development of the global economy,the phenomenon of greenhouse effect,energy shortage and environmental destruction frequently occur.Under this background,green transportation has become the development direction.In addition,the electric vehicle has advantages of energy saving,low emissions,But at the same time it has the characteristics of volatility and uncertainty.Therefore,predicting the load of electric vehicles can guide the charging of electric vehicles,thereby reducing the peak-to-valley difference of the power grid,improving the economics of power grid operation,and fully utilizing the power grid to bring economic benefits to enterprises.Therefore,the study of electric vehicle load forecasting method is important.In this paper,the following research is carried out:Firstly,the development history and development status of electric vehicles at home and abroad is studied,and the common algorithms of load forecasting are summarized,as well as their respective advantages and disadvantages.Secondly,based on the historical data of electric vehicle load,the research on the improvement of load variation characteristics is studied,including its own volatility,periodicity,and analysis of factors related to load changes,including temperature and rainfall.Therefore,the important factors of the load change of the electric vehicle are fully explained.Then,based on the data mining technology,the fuzzy clustering method is used to calculate the similarity between the forecasting date and the historical date,and the best similar day is obtained by orderly arrangement,thus providing a scientific method for screening the validity of historical data.Finally,a weighted-based combined forecasting method is proposed to combine the data features of different methods to effectively improve the accuracy and stability of load forecasting.And according to the characteristics of the combined prediction,the method for determining the weight is improved.The prediction of the actual load of electric vehicles in Shenyang proves the scientific and effective method.