Abstract: Wind speed has the characteristics of large stochastic volatility, which affects the wind power and the stability of the grid connected with it. Good predictions of wind speed and wind power are the key to solve the integration problem of wind power with grid. A comparison between grey models and Markov chain models for predicting wind speed is made. The gray-Markov chain prediction model and the wind speed prediction error are obtained by analyzing the error transfer series of the fitted values with grey models and establishing Markov state transition probability matrices. The expected values of the transfer matrices are used to correct the traditional Markov chain, and the improved gray-Markov chain model can thus be obtained to predict the wind power directly. Comparing with the real power curve, the improved grey-Markov model is better in prediction accuracy.

Key words: wind speed forecasting, wind power prediction, grey model, Markov chain