Aplicação da Decomposição Variacional de Modos na avaliação da presença de Distúrbios da Qualidade da Energia Elétrica

This paper discusses the use of Variational Mode Decomposition (VMD) and the Random Forest (RF) classification algorithm, applied to the classification of Electric Power Quality Disturbances (PQD). Electrical disturbances were generated without and with Gaussian noise. VMD has been used to decompose the signals into Intrinsic Mode Functions (IMFs). Characteristics of each IMF were obtained, then the RF model has been parameterized to classify the DQEEs based on these characteristics. The implemented method achieved classification accuracy of 99.49% for 50 db noise, demonstrating the versatility and great potential that this method provides.