Real-time power estimation through the stationary packet wavelet transform
power estimation, wavelet transform, power quality
The estimate of power quality indices promotes as the main goal the monitoring systems (measurement, control, protection, etc.) information about suitable levels of voltage and power quality. Therefore, this thesis proposes the application of the maximal overlap wavelet packet transform (MODWPT) in real-time to estimate RMS voltages and currents, primary power quantities (active, total apparent, nonactive power, and power factor), and distortion power. The MODWPT provides a uniform frequency band and possesses the time-invariance property, which is ideal for the real-time power estimation. The performance of the proposed wavelet-based power measurement was assessed and compared to the IEEE Standard 1459-2010 in different case studies including stationary synthetic waveforms and power quality disturbances in experimental setup. The proposed method presented good results in the estimation of RMS values and primary power quantities by using compact mother wavelets, whereas the estimation of distortion power was better with long mother wavelet.