Wind and solar power is a renewable energy source, already in commercial scale, with exponential growth worldwide and, more recently, in Brazil. Cepel has been developing techniques for short-term prediction (up to 48 h ahead) of wind generation, based in physical models, which take into consideration climate information, time series models, artificial intelligence methods, as well as the combination of these techniques in order to obtain point or probabilistic forecasts of wind generation. Models for data gap filling and correction of outliers in wind records are also employed. Similar models are now starting to be applied in longer term horizons, and also for solar/photovoltaic sources.
Besides an accurate characterization of the probability distribution of these intermittent sources for power generation and ther correlations, Cepel has been also developing modeling approaches and solving strategies to represent their uncertainty in the power generation planning models of DEA department,at all planning horizon levels. Methodologies for the incorporation of small range storage devices, such as batteries, are also being developed, together with already existing approaches in our models to represent pumped storage generation.
Contact the responsible area via email: