| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 3.18 MB | Adobe PDF |
Authors
Advisor(s)
Abstract(s)
Durante os últimos tempos, o setor da eletricidade tem sofrido importantes reestruturações em diversos países, nomeadamente em Portugal. Este sector, que anteriormente se caracterizava por um funcionamento centralizado, depara-se agora como o paradigma dos Mercados de Eletricidade, mercados descentralizados.
A complexidade característica do problema de planeamento da exploração de centrais hídricas deve-se à relação não linear entre a potência, o caudal e a altura da queda, à interligação hidráulica entre os aproveitamentos dispostos no mesmo curso de água e às incertezas associadas à previsão das afluências e aos preços de mercado e à possibilidade de efetuar bombagem, processo este que se revela fundamental e imprescindível para uma correta gestão dos recursos hídricos.
Devido a complexidade do problema e ao considerável benefício económico associado à sua exploração, surge a necessidade de desenvolver metodologias que otimizem o planeamento da gestão das centrais hidroelétricas. Dada a necessidade das empresas detentoras de ativos de produção hídrica encontrarem o método mais adequado, vários trabalhos têm sido desenvolvidos. Com efeito, várias técnicas de otimização têm sido aplicadas a este problema, que se distinguem pelas simplificações introduzidas, porém sem nunca se ter encontrado uma clara vantagem para alguma delas.
Para resolver o problema, neste trabalho, foi abordado um modelo de otimização baseado nos Algoritmos Genéticos, de forma a estimar a remuneração das centrais hídricas, através da otimização da produção de energia elétrica em ambiente de mercado. Este método permite abordar o problema relativo a diferentes períodos horários e ao estado de operação (turbinamento/bombagem) tendo em conta as principais características de funcionamento das centrais hidroelétricas.
Através dos resultados obtidos, verifica-se que na resolução de problemas de otimização de centrais hidroelétricas baseado nas metodologias dos AGs, se trata de uma ferramenta computacional eficaz e com boas perspetivas de desenvolvimento futuro.
During recent times, the electricity sector has been suffering significant restructuring in several countries, including Portugal. This sector, which previously was characterised by a centralized operation, faces now as the paradigm of the Electricity markets, decentralized markets. The complexity feature of planning problem of exploration of water plants is due to the nonlinear relationship between power, flow rate and the height of the fall, the hydraulic interconnection between the potentials in same watercourse and uncertainty associated with the estimate of inflows and market prices and the possibility of effecting pumping, process this is fundamental and essential for a correct management of water resources. Because of the complexity of the problem and the considerable economic benefit associated with their exploration, arises the need to develop methodologies to optimize the planning of the management of hydroelectric plants. Given the need of companies holding assets of hydro production find the most suitable method, several works have been developed. Indeed, various optimization techniques have been applied to this problem, which are distinguished by the simplifications introduced, however without ever having found a clear advantage for any of them. To solve the problem, in this work, was approached a optimization model based on genetic algorithms in order to estimate the hydro power plants remuneration, by optimizing the production of electricity in market environment. This method makes it possible to address the problem related to different slots and the condition of operation (turbinate/pumping) having regard to the main features of operation of hydroelectric plants. Through the results obtained, we can see that in solving problems of optimization of hydro power plants based on methodologies of the AGs, it comes to a computational tool efficiently and with good perspectives of future development.
During recent times, the electricity sector has been suffering significant restructuring in several countries, including Portugal. This sector, which previously was characterised by a centralized operation, faces now as the paradigm of the Electricity markets, decentralized markets. The complexity feature of planning problem of exploration of water plants is due to the nonlinear relationship between power, flow rate and the height of the fall, the hydraulic interconnection between the potentials in same watercourse and uncertainty associated with the estimate of inflows and market prices and the possibility of effecting pumping, process this is fundamental and essential for a correct management of water resources. Because of the complexity of the problem and the considerable economic benefit associated with their exploration, arises the need to develop methodologies to optimize the planning of the management of hydroelectric plants. Given the need of companies holding assets of hydro production find the most suitable method, several works have been developed. Indeed, various optimization techniques have been applied to this problem, which are distinguished by the simplifications introduced, however without ever having found a clear advantage for any of them. To solve the problem, in this work, was approached a optimization model based on genetic algorithms in order to estimate the hydro power plants remuneration, by optimizing the production of electricity in market environment. This method makes it possible to address the problem related to different slots and the condition of operation (turbinate/pumping) having regard to the main features of operation of hydroelectric plants. Through the results obtained, we can see that in solving problems of optimization of hydro power plants based on methodologies of the AGs, it comes to a computational tool efficiently and with good perspectives of future development.
Description
Keywords
Energia hídrica Otimização de centrais hidroelétricas Algoritmos genéticos