Reconfiguração eficiente de redes primárias de distribuição de energia elétrica para redução de perdas técnicas

Abstract

The Reconfiguration of Primary Distribution Systems is the process of determining which equipments of a certain network will be closed and which ones will be opened to reduce the amount of power losses. Operational restrictions must be met during the entire process. This problem presents high complexity due to its combinatorial, discrete and non differentiable nature. According to the literature, no polynomial time algorithm for this purpose is known. Therefore, the application of evolutionary computation seems reasonable. This work introduces a specific genetic algorithm provided with an codification designed to solve the problem. Some difficulties reported in the literature as the maintenance of feasibility and search space reduction are discussed. An efficient power flow algorithm for radial systems is presented and extended to weakly meshed networks. Lastly, the developed methodology is compared with three others reported in the literature. The results suggest that the proposed method is among the best ones on dealing with the reconfiguration problem.