Accurate wind power prediction is significant for secure and stable operation of power system

and the lag is the main reason of wind power prediction error

especially when wind speed changes rapidly

the lag will result in big error. Considering strong relationship between wind speed and wind power

an error correction method based on LR（Local Ramp） of wind speed is proposed to improve the lag

and the predicted wind speed after correction and historical wind power are taken as input for wind power prediction. The parameters of LSSVM（Least Square Support Vector Machine） are optimized by using GWO（Grey Wolf Optimization） algorithm to improve the accuracy of wind power prediction. Case results show that the proposed method can effectively improve the accuracy of wind power prediction.