Modelagem caixa cinza de transformadores de potência com o emprego de algoritmos de otimização

Abstract

The proper functioning of the power transformer is essential for the efficient delivery of electricity to consumers. Aiming at the high performance of the transformer, studies and analyzes are being developed on the failures of these devices that can interfere with their performance. To assist in this understanding, different types of modeling of power transformers are used: white box, black box and gray box. The white box and black box models have limitations, as they are not able to carry out a detailed analysis of the behavior inside a transformer that does not offer its geometric data. Thus, gray box modeling has become an interesting option for modeling the transformer and generating an approximate detailed circuit. The present work employs optimization methods to obtain the values of the R-L-C parameters of the gray box model of a power transformer winding. Three optimization methods were implemented, namely: Cuckoo Search (CS), Gray Wolves (GWO) and Whales (WOA) optimization algorithms. Also, there was a variation in the calculation procedures of the objective function to increase the accuracy of the proposed gray box modeling. The proposed modeling low levels of difference between the simulated and experimental frequency response of the winding, thus, it stands out as a good option to model a power transformer winding.