Probabilistic estimation model of power curve to enhance power output forecasting of wind generating resources

Wind-generating resources are variable and uncertain compared to traditional power generation resources. The accurate short-term forecasting of power outputs is essential to the extensive integration of wind generation into power grids. The variability of wind speed leads to uncertainty in wind power outputs. Consequently, forecasting errors increase the uncertainty of wind power forecasts. In this paper, we propose the probabilistic power curve estimation to enhance power output forecasting of wind generating resources. In order to enhance the wind power output forecasting, the probabilistic approach such as theoretical Weibull distribution parameters and Monte-Carlo simulation method is applied, the new multiple segments of the existing power curve are used for practical probabilistic power curve and spatial interpolation modeling based on Ordinary Kriging techniques is proposed for generating wind speed forecasting outputs. In addition, the new power slope estimation of the forecasting power outputs is proposed. To validate the proposed probabilistic power curve model, empirical data from the Jeju Island's wind farms are considered in South Korea. The proposed probabilistic power curve model will contribute to the accurate estimation of the relationships between measured wind speeds and electrical power outputs, thus quantifying the uncertainties in power energy conversion.