Abstract: One of the most important challenges of on-site partial discharge measurement is to capture the original signal in a noisy environment. The different sources of noise which includes thermal or resister noise added by the measuring circuit, signals due to electromagnetic interferences. Sophisticated and proper methods are required to detect and measured the partial discharge signals. Fortunately, due to the advancement in the analog to digital conversion techniques, signals can be easily captured with the help of interfacing devices. The signal can be extracted and converted into Excel sheets. This paper deals with de-noising of partial discharge signals during the breakdown phenomena in solid dielectrics. Several techniques are investigated and applied to the captured partial discharge signals. For the enhancement the insulation quality of the dielectric media, diagnosis plays a very important role in a power system. For diagnostic test such as detection, location and identification are used to find defects or faults and assess the ageing degree of the dielectric material. Wavelet Transform technique used to suppress a noise from partial discharge signal requires reasonable mother wavelet, amount of scales and thresholds to produce its best result. Results indicate that this technique not only attain better dancing effect, but also improve the sensitivity of partial discharge detection.

Keywords: Partial Discharge (PD), Discrete Wavelet Transform (DWT), Noise rejection, Signal denoising, Signal extraction.