Partial discharge pattern recognition and noise reduction using expert system

Abstract

newline Partial Discharge (PD) is an undesirable phenomenon that reduces the life expectancy of power system apparatus and causes permanent damage to them. PD usually happens at defects such as voids, cavities, cracks, joints, and delaminations. In general, PD does not cause an instant breakdown. The damage from the discharge can be estimated depending on the type of discharge, like internal or surface discharge, termination discharge, corona, electrical treeing, etc. The significance of partial discharges on insulation life has long been recognized. Denoising of PD Signals and identification of PD Sources are the two broad research areas in the analysis of PD. Measurement and identification of PD signals are essential for the safe operation and condition monitoring of power apparatus. However, high-level noises in the signals limit the accuracy of diagnoses in such measurements. Hence, denoising of PD signals is usually the first issue to be accomplished during PD analysis and diagnosis