Previsão do preço spot do mercado majoritário boliviano utilizando redes neurais artificiais

• Ano de defesa: 2015
• Autor(a) principal: Padilla Alvarez, Ángela Patricia [UNESP]
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Estadual Paulista (Unesp)
• Palavras-chave em Português: Redes neurais (Computação); Sistemas de energia eletrica; Preços; Neural networks (Computer science)
• Link de acesso: http://hdl.handle.net/11449/145480; http://www.athena.biblioteca.unesp.br/exlibris/bd/cathedra/19-09-2016/000870791.pdf

With the improvement of the Bolivian electricity market, it becomes important for all the participants of the electricity sector the prediction of the energy price in a short time horizon, giving with this the opportunity to the generating companies to optimize their supply strategies and maximize their profit. In this work, the problem of prediction will be approached using artificial intelligence techniques, specifically the use of an artificial neural network to price prediction for a 24 hours horizon, using a nonlinear autoregressive network with exogenous inputs (NARX) and a pattern recognition network in order to identify the best results of the prediction