Planejamento da expansão de curto prazo de redes de distribuição considerando geração distribuída e confiabilidade
Carregando...
Arquivos
Data
2016-12-09
Autores
Paiva, Rodrigo Rodrigues da Cunha [UNESP]
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Estadual Paulista (Unesp)
Resumo
O planejamento de sistemas de distribuição consiste em encontrar uma configuração otimizada, a um baixo custo, que permita manter a qualidade e confiabilidade do fornecimento de energia. Fontes de Geração Distribuída (GD) quando inseridas na rede alteram as suas características físicas e operacionais e tornam o problema de planejamento de curto prazo mais complexo para ser resolvido. Portanto, é importante desenvolver ferramentas computacionais eficientes para reduzir custos e o tempo na tomada de decisão para o setor de planejamento das empresas, indicando onde, quando e quais os tipos de componentes devem ser instalados e/ou substituídos nas redes de distribuição na presença de fontes de GD. Neste trabalho propõe-se uma metodologia para o planejamento de curto prazo para alocação das fontes de GD do tipo eólica (considerando as incertezas presentes neste tipo de fonte primária de energia), alocação de bancos de capacitores, dispositivos de proteção e controle, bem como a possibilidade de recondutoramento e troca de estruturas da rede de distribuição, mantendo-se os índices de qualidade para o fornecimento de energia dentro dos padrões estabelecidos pela agência reguladora. O problema de planejamento de curto prazo é formulado como um modelo de programação inteira multiobjetivo, o qual consiste em minimizar os custos de investimento e perdas técnicas na rede (energia não suprida e perdas ôhmicas nos condutores) e está sujeito a restrições físicas, econômicas e operacionais. Este problema é resolvido através de um algoritmo genético multiobjetivo baseado no Non-dominated Sorting Genetic Algorithm (NSGA-II), e a metodologia proposta e implementada foi testada em um sistema de distribuição real de 135 barras, e os resultados obtidos mostram eficácia e robustez da metodologia proposta.
The planning of distribution systems consists of finding an optimized configuration, at a low cost, which allows to maintain the quality and reliability of the power supply. Distributed Generation (DG) sources, when inserted into the network, change their physical and operational characteristics and make the short-term planning problem more complex to be solved. Hence, it is important to develop efficient computational tools aiming to reduce costs and the impact of time limits on decision making for the utility planning sector, indicating where, when and what types of components should be installed and / or replaced in distribution networks in the presence of sources. Accordingly, this work proposes a new methodology for the short-term planning which considers the allocation of wind-type DG sources (considering the uncertainties present in this type of primary energy source capacitor banks, protection and control devices, as well as the possibility of reconnecting and exchanging structures of the distribution network, maintaining the quality indices for the supply of energy within the standards established by the regulatory agency. The short-term planning problem is formulated as an integer multiobjective programming model, which consists of minimizing the investment costs and technical losses in the network (energy not supplied and ohmic losses in the conductors) by subjecting to physical, economic and operational constraints. Thus, this problem is solved here by using a multiobjective genetic algorithm based on the Non-dominated Sorting Genetic Algorithm (NSGA-II). In order to testthe proposed methodology, a real distribution system, containing 135-buses, is implemented and analyzed. In short, the obtained results demonstrate its efficacy and robustness in light of the proposed methodology.
The planning of distribution systems consists of finding an optimized configuration, at a low cost, which allows to maintain the quality and reliability of the power supply. Distributed Generation (DG) sources, when inserted into the network, change their physical and operational characteristics and make the short-term planning problem more complex to be solved. Hence, it is important to develop efficient computational tools aiming to reduce costs and the impact of time limits on decision making for the utility planning sector, indicating where, when and what types of components should be installed and / or replaced in distribution networks in the presence of sources. Accordingly, this work proposes a new methodology for the short-term planning which considers the allocation of wind-type DG sources (considering the uncertainties present in this type of primary energy source capacitor banks, protection and control devices, as well as the possibility of reconnecting and exchanging structures of the distribution network, maintaining the quality indices for the supply of energy within the standards established by the regulatory agency. The short-term planning problem is formulated as an integer multiobjective programming model, which consists of minimizing the investment costs and technical losses in the network (energy not supplied and ohmic losses in the conductors) by subjecting to physical, economic and operational constraints. Thus, this problem is solved here by using a multiobjective genetic algorithm based on the Non-dominated Sorting Genetic Algorithm (NSGA-II). In order to testthe proposed methodology, a real distribution system, containing 135-buses, is implemented and analyzed. In short, the obtained results demonstrate its efficacy and robustness in light of the proposed methodology.
Descrição
Palavras-chave
Sistemas de distribuição, Otimização multiobjetivo, Geração distribuída, Confiabilidade, Planejamento, Distribution systems, Multi-objective optimization, Distributed generation, Reliability