Partial Discharge Pattern Analysis Based on 2-D Wavelet Transform

Partial discharge (PD) measurement and pattern recognition for insulation defect is an effective tool for condition assessment. The patterns obtained by PD detection system contain characteristic features of the insulation defect. This paper presents a PD experiment of semi-conductor layer tip defect in the joint, and applies a new feature method based on 2-D wavelet decomposition and pattern edge detection. Research shows that this method has robust capability to extract main information from complicated online PD patterns.