Based on the measured data of an electric vehicle company, this paper analyzes the tracking data of multi-type electric vehicles. Firstly, this paper sets conditions to preprocess the data, deletes the abnormal data and adopts Lagrange interpolation to supplement some data; then, the K-S test and J-B test methods are used to verify whether the electric vehicle cluster data conform to the regular distribution. Secondly, Monte Carlo simulation is used to predict the charging load of electric vehicles, and the charging load curve of multiple types of electric vehicles is formed. Finally, the daily electric load is used to analyze the impact of electric vehicle loads on the power grid. The results show that electric private cars and electric taxis have a large load dispatching potential, but the phenomenon of "peak plus peak" will occur under disorderly charging.