Discrete wavelet transform based classifier for PQ disturbance detection

Abstract

In this paper the Discrete Wavelet Transform (DWT) concept using multi-resolution signal decomposition (MSD) technique with Daubechies family <i>db4 </i>second order vanishing moment mother wavelet for detection of power quality (PQ) disturbances has been proposed and presented. The PQ disturbances are also classified by extracting energy distribution features from multiresolution analysis (MRA) curve based on detailed energy distribution pattern (DEDP). The information obtained from detail squared wavelet transform coefficients (SWTC) of the disturbance signal at each wavelet decomposition level is utilized for plotting MRA curve. The proposed method algorithm has been verified by simulating various PQ disturbances and results are analyzed using Mathworks Matlab/Simulink. The classification and detection sensitivity of the proposed method is also verified and presented when signal to noise ratio (SNR) of 20 dB is superimposed during the duration of PQ disturbances. The proposed technique appears to be effective and accurate tool for the detection, localization and classification of PQ disturbances.