Comparação de métodos de detecção de distúrbios de qualidade da energia elétrica.

Abstract

The detection of disturbances of power quality is a very important task for system protection and maintenance of electric power supply. In this context, it must be done efficiently, in monitoring for signls of voltages and currents of the system, using signal processing methods for identifying these events. The monitoring proposed here starts with the decomposition of the signal generated in a bench simulation of disturbances in power quality and uses the Fourier Transform Window (TFW) and Wavelet Transform (WT). These methods use a sampling frequency of 30,720 Hz is chosen to be generated 512 points in each cycle of the analyzed signal whose fundamental frequency is 60 Hz Thus, the discretization can be made to the signal frequency of 15,360 Hz, which corresponds to 256º harmonic. Os methods make comparison of the sampled signal with the reference signal without disturbance. The result is compared with detection done at the bench to verify the error. The detection is to identify the event and pre-classifies it without the help of complementary techniques. Thus, this work presents a comparative performance of these methods by demonstrating that both have a low efficiency in detecting at least one disturbance in the power quality.