Generation of wind power time series to fit time-domain characteristics

Abstract

The generation of wind power time series is important for electric power system planning and decision making. A method to generate a synthetic series of wind power outputs is proposed, considering the state transition, the duration time and the variation features of wind power. The simulation results using the proposed method for 25 wind farms at six different locations in different countries show that the wind power time series generated by the proposed method are able to reflect more comprehensive wind power characteristics than that generated by the conventional Markov chain Monte Carlo (MCMC) method.