Short-Term Wind Power Forecasting Based on Gaussian Process Regression

The growing share of renewable energy sources with stochastic nature generation in the powers systems introduces new challenges for short-term and ultra-short-term power systems balances planning. A significant portion of such generation is consisting of wind power plants. Wind power forecasting is one of the most effective and least capital-intensive for wind farms integration to power systems. This paper presents a technique for short-term wind power plants generation forecasting using a regression model based on Gaussian processes.