Uma abordagem computacional para previsão de demanda de energia elétrica e apoio à tomada de decisão no mercado de curto prazo no Brasil

• Ano de defesa: 2017
• Autor(a) principal: Bezerra, Francisco Elânio
• Orientador(a): Pereira, Fabio Henrique
• Banca de defesa: Pereira, Fabio Henrique, Dias, Cleber Gustavo, Souza, Reinaldo Castro, Mendes, Jorge José de Magalhães
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Nove de Julho
• Programa de Pós-Graduação: Programa de Pós-Graduação de Mestrado e Doutorado em Engenharia de Produção
• Departamento: Engenharia
• País: Brasil
• Palavras-chave em Português: mercado de curto prazo; rede neural artificial; otimização de energia; leilões de energia; contratação de energia
• Palavras-chave em Inglês: short-term market; artificial neural network; energy optimization; energy auctions; contracting energy
• Área do conhecimento CNPq: ENGENHARIAS::ENGENHARIA DE PRODUCAO
• Link de acesso: http://bibliotecatede.uninove.br/handle/tede/1615

The technological advance, in the world, has brought about profound changes in the way the electric energy is generated, distributed and consumed. The increase in electricity consumption and the interruption of power supply in Brazil led to the creation of Decree 5.163/2004, proposing a new model for the sale of electricity in the National Interconnected System through auctions in the free contracting environments between buyers and Sellers, or regulated through auctions promoted by the Electric Energy Trading Chamber (CCEE), which accounts for the difference between contracting and energy consumption and through the settlement price of the differences and promotes the settlement of this energy short-term market .If you have more contracts than consumption, or more consumption than contracts, you will suffer penalties. With the change in the commercialization of energy, the generators and distributors suffer with forecast of consumption and with amount of energy that must contract in the auctions. In this scenario, several techniques such as genetic algorithm, multicriteria decision, fuzzy logic, artificial neural networks among others have been used to optimize the system of buying and selling energy in this new environment. Therefore, the proposal of this work is to develop an intelligent computational system, using historical data from a distributor to forecast demand by consumption class to support decision making in the short term market. The result of the work may provide conditions for a distributor to monitor energy demand by consumption class, provide possibilities for short-term market trading and minimize losses with subcontracting and over-contracting.