Partial Discharge Localization using Neural Networks and Transformer Detailed Model

Partial discharge is a main source of insulation degradation in power transformers. Therefore accurately locating of partial discharge sources in transformers as the main equipment in power system is needed. This paper proposed two novel methods based on artificial neural networks for partial discharge localization in the power transformers. For this purpose detailed model of transformer and three capacitor model of partial discharge is used. Then impulse test is applied to transformer terminals and current in neutral point is measured for training and test of artificial neural networks. As actual current signals include noise components، the noisy component is added to measured current signals and performance of proposed neural networks for partial discharge localization is shown and results are compared.