Performance analysis of thresholding techniques for denoising of simulated partial discharge signals corrupted by Gaussian white noise

Partial Discharge (PD) signal measurement is very significant tool in analyzing condition of the electrical insulation. The PD information is lost in the presence of various noises. The wavelet transform (WT) based denoising provides a better platform for pre and post processing of PD signal. The wavelet adaptive Thresholding de-noising techniques are well suited for reducing the noise. This paper adopts the various adaptive thresholding techniques such as VisuShrink, SureShrink, combination of the two called Heursure, minimax thresholding and BayesShrink, which are broadly classified as Global and Local thresholding methods. The algorithm presents the comparative analysis for the selection of optimal mother wavelet. Once the optimal mother wavelet is chosen, selection of the best thresholding rule is identified by comparing the values of signal to noise ratio (SNR), mean square error (MSE) and Peak Signal to Noise ratio (PSNR) of all the techniques. The algorithm also presents the comparison between Hard and Soft thresholding. It is shown that the soft thresholding is best suited to remove the noise compared to hard thresholding. The simulated Damped Exponential Pulse (DEP) and Damped Oscillatory Pulse (DOP) has been used. Three sets of PD data are considered to check the performance of the algorithm.