Name: | Description: | Size: | Format: | |
---|---|---|---|---|
194.38 KB | Adobe PDF |
Abstract(s)
A energia eléctrica é um bem essencial para a maioria das sociedades. O seu fornecimento tem sido encarado como um serviço público, da responsabilidade dos governos, através de empresas monopolistas, públicas e privadas.
O Mercado Ibérico de Electricidade (MIBEL) surge com o objectivo da integração e cooperação do sector eléctrico Português e Espanhol, no qual é possível negociar preços e volumes de energia. Actualmente, as entidades podem negociar através de um mercado bolsista ou num mercado de contratos bilaterais.
Uma análise dos mercados de electricidade existentes mostra que estes estão longe de estarem liberalizados. As tarifas não reflectem o efeito da competitividade. Além disso, o recurso a contratos bilaterais limita frequentemente os clientes a um único fornecedor de energia eléctrica.
Nos últimos anos, têm surgido uma série de ferramentas computacionais que permitem simular, parte ou a totalidade, dos mercados de electricidade. Contudo, apesar das suas potencialidades, muitos simuladores carecem de flexibilidade e generalidade.
Nesta perspectiva, esta dissertação tem como principal objectivo o desenvolvimento de um simulador de mercados de energia eléctrica que possibilite lidar com as dificuldades inerentes a este novo modelo de mercado, recorrendo a agentes computacionais autónomos.
A dissertação descreve o desenho e a implementação de um simulador simplificado para negociação de contratos bilaterais em mercados de energia, com particular incidência para o desenho das estratégias a utilizar pelas partes negociais. Além disso, efectua-se a descrição de um caso prático, com dados do MIBEL.
Descrevem-se também várias simulações computacionais, envolvendo retalhistas e consumidores de energia eléctrica, que utilizam diferentes estratégias negociais. Efectua-se a análise detalhada dos resultados obtidos. De forma sucinta, os resultados permitem concluir que as melhores estratégias para cada entidade, no caso prático estudado, são: a estratégia de concessões fixas, para o retalhista, e a estratégia de concessões baseada no volume de energia, para o consumidor.
Electrical energy is a key asset of modern societies. However, its supply has been regarded as a public service, of government‘s responsibility, through monopolistic utility companies both public and private. The Iberian Market of Electricity (MIBEL) has arisenwith the objective of integrating the Iberian electricity sector, where market participants can negotiate mutually beneficial agreements. Specifically, participants can submit production and consumption bids to a stock market (Pool) or negotiate agreements on delivery and receipt of power (bilateral contracts). At present, an analysis of most electricity markets (EMs) yields the observation that they are still far from deregulated. In particular, tariffs do not reflect the pressure of competition. Multi-agent systems represent a relatively new and rapidly expanding area of research and development. A multi-agent approach in which software agents are capable of flexible action to meet their design objectives is an ideal fit to the naturally distributed domain of a deregulated energy market. Accordingly, this project addresses the challenge of using software agents to help manage the complexity of EMs. In the past years, a number of multi-agent energy management tools have emerged. However, despite the power and elegance of existing EM simulators, they often lack generality and flexibility, mainly because they are limited to particular market models, specific types of market participants, and particular features of market players. The thesis describes the design and implementation of a simplified EM simulator for negotiating the terms of bilateral contracts, placing emphasis on negotiation strategies. Additionally, it describes a case study aiming at evaluating the different strategies and the EM simulator generally. The experimental results showed that the best strategies for both agents are the following: i) retailer fixed concession strategy, and ii) consumer strategy based on energy volume.
Electrical energy is a key asset of modern societies. However, its supply has been regarded as a public service, of government‘s responsibility, through monopolistic utility companies both public and private. The Iberian Market of Electricity (MIBEL) has arisenwith the objective of integrating the Iberian electricity sector, where market participants can negotiate mutually beneficial agreements. Specifically, participants can submit production and consumption bids to a stock market (Pool) or negotiate agreements on delivery and receipt of power (bilateral contracts). At present, an analysis of most electricity markets (EMs) yields the observation that they are still far from deregulated. In particular, tariffs do not reflect the pressure of competition. Multi-agent systems represent a relatively new and rapidly expanding area of research and development. A multi-agent approach in which software agents are capable of flexible action to meet their design objectives is an ideal fit to the naturally distributed domain of a deregulated energy market. Accordingly, this project addresses the challenge of using software agents to help manage the complexity of EMs. In the past years, a number of multi-agent energy management tools have emerged. However, despite the power and elegance of existing EM simulators, they often lack generality and flexibility, mainly because they are limited to particular market models, specific types of market participants, and particular features of market players. The thesis describes the design and implementation of a simplified EM simulator for negotiating the terms of bilateral contracts, placing emphasis on negotiation strategies. Additionally, it describes a case study aiming at evaluating the different strategies and the EM simulator generally. The experimental results showed that the best strategies for both agents are the following: i) retailer fixed concession strategy, and ii) consumer strategy based on energy volume.
Description
Keywords
Sector eléctrico Sistemas multi-agente Sistemas multi-agente Contratos bilaterais Simulação MIBEL Electricity sector Multi-agent systems Negotiation strategies Bilateral contracts Simulation
Citation
Rodrigues, Tiago Miguel Reverendo - Estratégias para negociação de contratos bilaterais em mercados multi-agente de energia eléctrica. Lisboa: Instituto Superior de Engenharia de Lisboa, 2012. Dissertação de mestrado.