Previsões de curto prazo da produção de energia elétrica de um parque eólico

• Ano de defesa: 2013
• Autor(a) principal: Silva, Carlos Eduardo Gama da
• Orientador(a): Silva, Milthon Serna
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Programa de Pós-Graduação: Pós-Graduação em Engenharia Elétrica
• Palavras-chave em Português: Engenharia elétrica; Energia eólica; Energia - Fontes alternativas; Recursos energéticos
• Palavras-chave em Inglês: Electric engineering; Power resources; Renewable energy sources; Wind power
• Área do conhecimento CNPq: CNPQ::ENGENHARIAS::ENGENHARIA ELETRICA
• Link de acesso: https://ri.ufs.br/handle/riufs/5033

Due to the uncertainties and unforeseen variations of wind conditions, the intermittent nature of wind power generation implies a new challenge for the national electric sector agents especially to the Electricity Regulatory Agencies in view of proliferation and major investments planned for the current decade, in Brazil. For wind systems in operation, it is required prior knowledge of energy generation designed for a short term of up to one week. A good prediction of short term production from a wind farm enables the definition of operational conditions borderline, which will provide subsidies for their integration into the power grid and giving mainly information to the National System Operator - NSO. The forecasting process applied in this work is a statistical approach involving five steps: the first one refers to the daily data collection of real wind speed; secondly is made a refinement of statistical data, decomposing the series of wind speed in two components (an approximation component and a set of detail components to optimize the accuracy of short term forecasts); then, in these time series, it is applied the statistical model ARIMA (Box and Jenkins) and adaptations, considering the short term wind speed prediction; after that, taking into consideration the power curves of the generators, it is defined the energy production forecast. Finally, a statistical analysis is made of the predictions performed, aiming to assess the quality of the projections of the implemented model