Proteção de transformadores de potência eclassificação de transitórios elétricos por meio da transformada wavelet discreta

Abstract

This dissertation presents the development of an algorithm for classifying electrical transients in power transformers for protection. Initially the main transient involving power transformers were analyzed, with a special focus on the magnetizing current that occur during energizing equipment, known as inrush currents. Among the studied transient, current inrush are those that present greater difficulty in identification and discrimination by the differential protection, due to its high peak magnitude. In a second moment, was developed a test system implemented in ATP / EMTP software for the simulation of different operating situations of the power transformer, current data being obtained from TCs for analysis. After known the problem and implement a test for acquisition system, we developed an algorithm based on the last samples of the detail coefficients, called Details Signals, from the decomposition of the data of differential phase current using the Discrete Wavelet Transform. The methodology has the objective improve the system for protection of the transformer, becoming one efficient alternative for the identification of transients.