Wavelet Total Variation Method Applied to Partial Discharge Signal Denoising
Paulo Vitor do Carmo Batista, Hilton de Oliveira Mota, Gustavo M. Ferreira, Fernando T. de A. Silva, Flávio Henrique Vasconcelos · 2018
Signal processing techniques are used with the purpose of assisting in the process of equipment maintenance, avoiding failures and unwanted stops. One such fault is known as Partial Discharge (PD). The objective of this work is to evaluate the denoising performance of the Wavelet Total Variation method when applied to synthetic and laboratorial PD signals corrupted by Gaussian, harmonic and impulsive noises, as a way to improve their measurement conditions. The results obtained are promising and show that the method reconstructs the signal, maintaining just the discharges.