The large-scale development of electric vehicle and the continuous construction of charging facilities seriously threaten the stability of the power system. However, there is still a lack of simple and effective methods for electric vehicle ownership and load prediction. Therefore, the prediction model of electric vehicle ownership based on the comprehensive prediction is established. Three prediction models of grey prediction, back propagation (BP) neural network and long-short term memory (LSTM) network are used to predict the electric vehicle ownership, and the prediction results of the single prediction model are obtained. The entropy weight method is used to assign weight to the prediction results of the single prediction model, and the comprehensive prediction results are calculated. The electric vehicle load prediction model based on Monte Carlo algorithm is established. On the basis of ownership prediction, the characteristic parameters of electric vehicle batteries and user travel habits are simulated to predict the disordered charging behavior of electric vehicle and to form the daily load curve of electric vehicle. Finally, the effectiveness of the proposed model is verified by the data of electric vehicle ownership and charging load in a city. The example analysis shows that the proposed comprehensive prediction model has higher prediction accuracy than the single prediction model, and the load prediction results show that the grid connection of large-scale electric vehicle will bring new challenges to the power system.