A Neuro-Fuzzy Technique for Discrimination between Internal Faults and Magnetizing Inrush Currents in Transformers

This paper presents the application of fuzzy-neuro method to investigate transformer inrush current. Recently, the frequency environment of power systems has been made more complicated and the magnitude of second harmonic in inrush current has been decreased because of the improvement of cast steel. Therefore, traditional approaches will likely mal-operate in the case of magnetizing inrush with low second component and internal faults with high second harmonic. The proposed scheme enhances the inrush detection sensitivity of conventional techniques by using a fuzzy-neuro approach. Details of the design procedure and the results of performance studies with the proposed detector are given in the paper. Performance studies’ results show that the proposed algorithm is fast and accurate.