Otimização metaheurística de linhas de transmissão pela avaliação do vetor de poynting utilizando o método dos elementos de contorno

Abstract

With the legal opening for private investments in the Brazilian energy sector in the 90's, in particular the systematic implementation of auctions for defining power transmission concessions, several studies were made with the aim of developing transmission line projects in order to compete in these auctions. Major transmission companies such as Furnas, Eletronorte and CHESF promoted consortium research with line builders and hardware and insulators manufacturers, in order to develop projects with low cost/MW transmitted. In this scenario, many projects of high surge-impedance loading lines were developed and implemented, and as a result of these projects, publications were made describing the results. However, despite the quantity, quality and resourcefulness of published information, the descriptions of the optimization process of electrical line parameters, especially regarding characteristic impedance, were not given prominence, being often suppressed from the articles. Thus, this study proposes the investigation of a method for optimizing the transmission capacity of high surge-impedance loading lines, varying the power cables in the bundle, indirectly reducing its characteristic impedance through an original approach that uses basically the Poynting's vector as objective function of the Particle Swarm Optimization method.The electric and magnetic fields needed for defining the Poynting vector were calculated using the Boundary Element Method. The method was validated through the optimization of bundle configuration, using as a model the characteristics of the 500 kV line North / South Interconnection III - Segment 2 and the 500 kV line Presidente Dutra / Teresina / Sobral / Fortaleza. It was found transmission capacity gains of 7% and 22% respectively, by reducing the characteristic impedance, which was calculated after the optimization using the EMTP/ATP Line Constants Program, comparing it with the original values.