A new feature extraction and pattern recognition of partial discharge in solid Material by Neural network

Abstract

Partial discharge measuring is the mean tool of diagnosis in High voltage systems, equipment and solid dielectrics. Void or any defect in solid dielectrics will produce the partial discharge and may cause permanent failure after some time. According to type of defect, it will produce different patterns of partial discharge. In this paper we will study patterns of partial discharge in solid dielectric that voids are artificially have made in this materials according to experimental measuring in High voltage laboratory. The patterns will distinct with neural network and results of different type of neural network will be discussed.