Performance analysis of wavelet based denoise system for power quality disturbances

Abstract

The performance analysis of wavelet based denoise system for power quality disturbances including sag, swell, flicker, harmonic, transient and notch is presented in this paper. The denoise algorithm uses Daubechie family wavelet and a universal soft threshold technique applied on the detailed wavelet coefficients based on its standard deviation and data length. This eliminates and reduces the noise content in the signal. Lastly, the inverse discrete wavelet transform is done to reconstruct and recover the denoised PQ disturbance signal. All the disturbance signals are contaminated with Additive White Gaussian Noise of initial noise SNR of 30dB. The wavelet based denoise system in this paper shows excellent performance for sag, swell and flicker, good performance for harmonic, moderate performance for transient and poor performance for notch. The performance analysis presented in this paper provides an understanding of how the power quality disturbance frequency contents and its magnitude effect the performance of the wavelet based denoise system. © 2009 IEEE.