Otimização no cálculo da parte ativa do transformador de distribuição trifásico com uso de algoritmos genéticos visando melhor eficiência e menor custo

Abstract

The present work is a study of a methodology that seeks to be more economical and efficient compared to traditional methods, the design of three phase distribution transformers. To this end, we use a tool of Computational Intelligence, based on previous calculations made in the traditional way by manufacturing industries and books that address this matter. It is evident here, another way to design the active part of the transformer, the dimension of the core and coil, for this, used Genetic Algorithms (GA) which will assist in the calculation and selection of parameters, providing the best choice of design of the transformer, minimizing, on average, 5% in the cost and losses in 2.5% as input parameters the following standard. With changes in the fitness function of GA, one can simply adapt to new requirements and standards. As a result of using the genetic algorithm can find a design solution that is more efficient than that used by the manufacturers. The losses calculated values meet established by NBR-5440