Partial discharge pattern recognition of XLPE cable connector based on support vector machine

According to the characteristics of partial discharge of XLPE cable connector, a method based on support vector machine for Pattern Recognition is proposed in this paper. Six statistical operators, which include skewness, steepness, discharge factor, phase asymmetry, cross-correlation coefficient, and the modified cross-correlation coefficient, are considered as characteristic quantities. This paper describes a three-dimensional map structure of partial discharge, elaborates the extraction process of characteristic quantities, and analyses the basic principles based on support vector machines in detail. Finally, four kinds of typical partial discharge models are simulated in laboratory, and the experimental results show that the method is feasible and has significant effect.