Using Discrete Wavelet Transform to evaluate power quality at highly distorted three-phase systems

Waveforms acquired from the excitation of the main generator from a power plant are analyzed with the Discrete Wavelet Transform (DWT). The proper Wavelet mother order and the proper decomposition tree were determined considering the energy deviations between subsequent decompositions and respectively the maximum approximation error generated by the signal re-composition process. Power quality indices were determined using DWT individually for all phases and globally, for the entire three-phase system. A comparative analysis considering the same indices (calculated using Fast Fourier Transform - FFT) was performed and small differences were noticed between corresponding indices. The conclusion is that DWT can be applied considering both steady regimes and respectively non-steady regimes for which the FFT is unusable.