Type recognition of partial discharge source based on PCA and GWO-SVM

As an important indication and manifestation of insulation failures in high-voltage equipment, the type identification of partial discharge (PD) is important for the assessment of the insulation state of electrical equipment. In order to identify the type of partial discharge source accurately, this paper presents a method for partial discharge source type identification based on principal component analysis (PCA) and support vector machine (SVM) with grey wolf optimization algorithm (GWO). The PCA method is used to select the three features out of 10 that best represent the original data, namely the phase, E/I
peak
and Ipeak/Q. On this basis, the SVM kernel function
$g$
and the non-negative penalty factor
$c$
are optimized by GWO to establish a support vector machine classification model based on the grey wolf optimization algorithm (GWO-SVM). The results show that PCA is able to extract the main features well and that GWO-SVM can identify partial discharge source more accurately than the genetic algorithm-based support vector machine (GA-SVM), multilayer perceptron (MLP) and k-nearest neighbor (kNN) classifier.