genetic algorithms, power flow convergence / Aplicação de algoritmos genéticos para convergência de fluxo de potência em sistemas de energia

AUTOR(ES)
DATA DE PUBLICAÇÃO

2009

RESUMO

Calculating power flow is a fundamental problem in the field of electrical power systems, therefore the development of computational methods to improve their convergence is a subject of research. Traditional numerical methods, such as the Newton-Raphson algorithm, are very efficient for solving this problem, but they are highly dependent on initial conditions and system loading. In some cases, they converge to a solution that is unfeasible, or in more critical cases, they do not present any solution at all. In other situations, convergence may be difficult due to the imbalance of active and reactive power generated, requiring a redispatch of active and reactive power. This work therefore proposes a system composed of the Newton-Raphson algorithm and a Genetic Algorithm (GA), the main purpose of the GA being to provide an initial solution to the Newton-Raphson algorithm, improving its convergence. In situations where this convergence does not occur, GA seeks, through the redispatch of generation, a viable operating point for resolving the power flow problem. The genetic algorithms developed, and the Newton-Raphson algorithm, were submitted to the IEEE test cases of fourteen and thirty bars, and to the test case of five bars reported in the literature. The results confirm the proposed solution

ASSUNTO(S)

algoritmos genéticos computabilidade e modelos de computacao convergência de fluxo de potência

ACESSO AO ARTIGO

Documentos Relacionados