Detecting transients in the voltage waveforms is one of the most important issues in power quality for both the preservation of equipment connected to the network, as well as the power supply. Among various techniques for detection of transients in electrical signals highlights, the wavelet transform is a method that can determine the instant when the disturbances occur and can detect different types in a same signal. The orthonormal wavelet filter banks yield a perfect reconstruction of a decomposed signal and if the regularity of the used wavelets is high, where the compression of the regular parts of the signal is good, and better evidencing of the singularities is at hand. In this paper, an orthonormal wavelet filter banks parameterization for detecting disturbances in the voltage waveforms is proposed. To evaluate the proposed methodology, some tests of detecting singularities in the voltage waveforms are presented.