Noise Elimination in Partial Discharge On-Line Monitoring for Transformer Based on Wavelet Transform

The study of noise elimination in partial discharge (PD) on-line monitoring for transformer was carried out to early detect the latent failt of power transformer insulation. PD signal usually mixed up all kinds of noise, it was difficult to extract faint PD impulse signal from strong background noise with Fourier transform. Adaptive filtering algorithm based on wavelet packet transform was presented to suppress the periodic noise and the complex wavelet transform was used to eliminate white noise from PD signal. Results show that the proposed method is an effective optimal denoising method for power transformer PD On-line monitoring.