Abstract: Electric vehicle cluster connected to the grid for disorderly charging will increase the peak-valley difference and cause the problem of ‘peak on peak’. To solve this problem, a hierarchical control method for orderly charging of electric vehicles which considers new energy accommodation is proposed. Firstly, based on the theory of long short-term memory neural networks, the charging load forecasting model for electric vehicles is established; Secondly, the mathematical models of wind and photovoltaic power generation are established; Finally, the hierarchical orderly charging control strategy is proposed. The upper master control system formulates the electric vehicle charging power guidance curve. And the lower secondary control system reasonably arranges the electric vehicle charging plan and uses the probability transfer matrix algorithm to obtain the optimal charging control strategy. The simulation results show that the proposed orderly charging control strategy can effectively reduce the peak-valley difference of grid load.