A Novel Defect Classification System of Cast-Resin Transformers by Neural Network under Acoustic Emission Signal

Degraded insulating property of electric equipments will lead to serious accident and great loss for the utilities and customers. Partial discharge detection is an efficient diagnosis method to prevent the failure of electric equipments arising from degrading insulation. However, universal offline partial discharge detection could be performed only during shutdown of equipments. By using the principle of Acoustic Emission (AE) and real-time online detection functions, this paper analyzed partial discharge pattern for cast-resin transformers and conducted high-voltage test of pre-faulty transformers. Furthermore, it collected partial discharge AE signals with selected features and identified their faulty types using artificial neural network. The research results show that the average identification rate can reach as high as 92.5%. Key-Words: Neural Network, Partial Discharge, Acoustic Emission, Transformer.