Improved probabilistic wind power forecasts with an inverse power curve transformation and censored regression

Forecasting wind power is an important part of a successful integration of wind power into the power grid. Forecasts with lead times longer than 6 hours are generally made by using statistical methods to postprocess forecasts from numerical weather prediction systems. Two major problems that complicate this approach are the nonlinear relationship between wind speed and power production and the limited range of power production between zero and nominal power of the turbine. In practice, the nonlinearity is often tackled by using nonlinear nonparametric regression methods while the limited range is typically not addressed explicitly. However, such an approach ignores valuable and readily available information: the power curve of the turbine's manufacturer. Much of the nonlinearity can be directly accounted for by transforming the observed power production into wind speed via the inverse power curve so that simpler linear regression models can be used. Furthermore, the limited range of the transformed power production can be easily exploited by adopting censored regression models. In this study, we evaluate quantile forecasts from a range of methods: (a) using para-metric and nonparametric models, (b) with and without the proposed inverse power curve transformation, and (c) with and without censoring. The results show that with our inverse (power-to-wind) transformation, simpler linear regression models with censoring perform equally or better than nonlinear models with or without the frequently used wind-to-power transformation.