Previsão de Demanda de Energia Elétrica em Microgrid Considerando Níveis Menos Agregados por Meio da Aplicação de Rede Neural Artificial GRNN Combinada com o Método Estatístico SARIMA

The growth in electricity consumption the world forces countries to have a well-structured planning relation forecasting demand for their most diverse sectors. Several techniques are used predict electrical loads, such as artificial intelligence models, statistical models and hybrid models. This work aims present model based on combination of method, SARIMA, an neural network, GRNN, improve accuracy forecasts consumption. data set this belongs group buildings located Itaipu Technological Park was acquired through electronic meters installed together with transformers that serve each these buildings, performing collection every 15 minutes. After processing refining database, were applied, one using forecast horizon 1, 3 5 days, first technique being between GRNN SARIMA other methods themselves separately, allowing comparison results. results obtained proposed combined are, general, more accurate when compared individually, they combine advantages end up smoothing negative characteristics other, thus causing balance reflects forecast. generated.