Selecting the best wavelet function for power quality disturbances identification patterns

This paper proposes a technique to select a wavelet function that shows good characteristics for the identification of power quality disturbances. It considers the low frequency disturbances such as flicker and harmonics as well as high frequency disturbances such as transient and voltage sags. Due to time-frequency localization properties, the discrete wavelet transform permits signal decomposition in different energy levels, which are used to characterize disturbances that contain information on the frequency domain. Four wavelet families were studied in which biorthogonal showed excellent performance.