In order to improve the power network stability involving the wind power, a twostage price adjustment strategy is proposed for electric vehicles based on the wind power prediction. This strategy promotes the wind power accommodation by predicting wind power and then regulating the price of electric vehicles. In the prediction stage, the LSTM neural network with a memory ability of time series is utilized to predict the wind power. At the pricing stage, an optimization model of price adjustment is established with an objective function of the high similarity between the predicted wind power curve and the charging load curve, and the small charging cost. The price is set based on the forecast wind power and then it is utilized to adjust the load so that the charging load is close to the wind power over time. Finally, a simulation is included to verify the effectiveness of the strategy. The charging load before and after optimization is compared. It is shown that the latter is closer to the prediction of wind power.