Identification of partial discharge signals Marcus de Paula University of Wisconsin – Madison 12/13/2013 Background • Partial Discharges: • Localized dielectric breakdown of a small portion of a solid or fluid electrical insulation system under high voltage stress; • Can lead to loss of insulating capacity and electrical system failure. Background • Filtering problem: • Have the frequency spectrum close to the noise spectrum; • It requires more elaborate filtering method. Goal • Use the wavelet transform and a spatially-adaptive coefficient selection procedure to explore the localized processing capabilities of the WT as a way to improve the separation of coefficients related to the signal and noise. Goal • The process basically consists of 6 steps: • • • • • • 1. Decomposition of the signal into 6 levels using WT. 2. Extraction of each decomposition. 3. Construction of the Maxima Lines. 4. CLASSIFY lines associated with the signal or noise. 5. Delete rows associated with noise. 6. Rebuild signal using the remaining lines. Maxima Lines Training Data • Source: • Example: SVM classifier • Harmonic noise test: 97 3 10 90 • Classification rate = 93.5% • Confusion matrix = • Pulse noise test: 90 10 17 83 • Classification rate = 86.5% • Confusion matrix = • Real sample test: 97 3 13 87 • Classification rate = 92% • Confusion matrix = SVM classifier • Results: Future work • Use the MLP classifier; • Compare the results; • Analyze differences. References •  MOTA, H., Sistema de aquisição e tratamento de dados para monitoramento e diagnóstico de equipamentos elétricos pelo método das descargas parciais (Acquisition system and data processing for monitoring and diagnostic of electrical equipment by the method of partial discharges). Universidade Federal de Minas Gerais (UFMG), Electrical Engineering Graduate Program. Belo Horizonte, Minas Gerais, Brazil, March of 2001. •  MOTA, H., Processamento de sinais de descargas parciais em tempo real com base em wavelets e seleção de coeficientes adaptativa espacialmente (Signal processing of partial discharges in real time based on wavelets and selection of spatially adaptive coefficients). Universidade Federal de Minas Gerais (UFMG), Electrical Engineering Graduate Program. Belo Horizonte, Minas Gerais, Brazil, November of 2011.