Partial discharge recognition based on SF 6 decomposition products and support vector machine

Partial discharge (PD) is usually aroused before the failure of gas-insulated switchgear (GIS) caused by insulation defects, which results in the decomposition of SF 6 used as insulation gas. Concentrations of decomposition products of SF 6 under different kinds of PDs are disparate. Thus, SF 6 decomposition products can be used for PD recognition. In this study, a gas chamber and four defect models were designed to simulate four kinds of typical PDs in GIS. SF 6 decomposition experiments were conducted under the four kinds of PDs. Four kinds of decomposition products, that is, SO 2 F 2 , SOF 2 , CO 2 and CF 4 , were selected as feature components. Their concentrations were detected under each experiment. Three concentration ratios, that is, c (SO 2 F 2 )/ c (SOF 2 ), c (CF 4 )/ c (CO 2 ) and c (CO 2 +CF 4 )/ c (SOF 2 +SO 2 F 2 ), were proposed as feature parameters for PD recognition. Their physical significances were also analysed. Then, a support vector machine (SVM) was employed as classifier to recognise the four kinds of PDs. The parameters of the SVM were optimised using particle swarm optimisation algorithm. Results show that the recognition method based on SF 6 decomposition products and SVM performs well in PD recognition.

[1]  B. Florkowska,et al.  Partial discharge mechanism in a non-uniform electric field at higher pressure , 2011 .

[2]  L. Hao,et al.  Partial discharge source discrimination using a support vector machine , 2010, IEEE Transactions on Dielectrics and Electrical Insulation.

[3]  L. G. Christophorou,et al.  Neutral Decomposition Products in Spark Breakdown of SF6 , 1986, IEEE Transactions on Electrical Insulation.

[4]  R. J. Brunt.  Physics and chemistry of partial discharge and corona. Recent advances and future challenges , 1994 .

[5]  J. Casanovas,et al.  SF6 decomposition under power arcs : chemical aspects , 1991 .

[6]  Martin D. Judd,et al.  Differentiating between partial discharge sources using envelope comparison of ultra-high-frequency signals , 2010 .

[7]  R. J. Van Brunt,et al.  Plasma chemical model for decomposition of SF6 in a negative glow corona discharge , 1994 .

[8]  James Kennedy,et al.  Particle swarm optimization , 2002, Proceedings of ICNN'95 - International Conference on Neural Networks.

[9]  Chih-Jen Lin,et al.  A comparison of methods for multiclass support vector machines , 2002, IEEE Trans. Neural Networks.

[10]  Vladimir Vapnik,et al.  Statistical learning theory , 1998 .

[11]  J. Suehiro,et al.  Analysis of PD-generated SF/sub 6/ decomposition gases adsorbed on carbon nanotubes , 2006, IEEE Transactions on Dielectrics and Electrical Insulation.

[12]  F. Chu,et al.  Degradation of Silica-Filled Epoxy Spacers By ARC Contaminated Gases in SF6-Insulated Equipment , 1985, IEEE Transactions on Power Apparatus and Systems.

[13]  Ming Tang,et al.  Study on mathematical model for VHF partial discharge of typical insulated defects in GIS , 2007, IEEE Transactions on Dielectrics and Electrical Insulation.

[14]  A. Belarbi,et al.  Study of the decomposition of SF6 under dc negative polarity corona discharges (point‐to‐plane geometry): Influence of the metal constituting the plane electrode , 1992 .

[15]  Martin D. Judd,et al.  Partial discharge diagnostics for gas insulated substations , 1995 .

[16]  L.E. Lundgaard,et al.  Partial discharge. XIV. Acoustic partial discharge detection-practical application , 1992, IEEE Electrical Insulation Magazine.

[17]  Ju Tang,et al.  Suppressing white‐noise in partial discharge measurements—part 1: construction of complex Daubechies wavelet and complex threshold , 2010 .

[18]  R. J. Vanbrunt,et al.  Production Rates for Oxyfluorides SOF2, SO2F2, and SOF4 in SF6 Corona Discharges. , 1985, Journal of research of the National Bureau of Standards.