Estratégias de previsão de geração de energia eólica utilizando redes neurais artificiais

• Autor(a) principal: Bulhões, Diego Bitencourt
• Data de Publicação: 2015
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Repositório Institucional da UFS
• Texto Completo: https://ri.ufs.br/handle/riufs/5015
• topic: Engenharia elétrica; Redes neurais (Computação); Energia eólica; Energia elétrica; Energia Fontes alternativas; CNPQ::ENGENHARIAS::ENGENHARIA ELETRICA

Nowadays, with the increase of electrical load demand, possible shortages of fossil fuels and strict restrictions to deployment of new hydroelectric, has seen a growing awareness for the use of environmentally clean energy source. Thus, the generation of energy through wind has been a widespread alternative in Brazil and the world. However, the accelerated expansion of wind power generation is a challenge because of the need to know the possible impacts in the operation, maintenance and interconnection with the existing energy system. The power generation forecast has been used in order to minimize this challenge, considering that the wind speed does not depend on human intervention. Therefore, in this work have been studied the potentiality of forecasting models based on artificial neural networks to forecast the wind generation in very short term horizon (a few minutes up to hours ahead) and short term horizon (a few hours up to a few days ahead).