Paper Contents
Abstract
Transformer protection remains a critical concern in power systems, particularly due to the challenge of accurately and swiftly distinguishing magnetizing inrush currents from internal fault currents. To address this issue, artificial neural networks (ANNs) have been proposed and proven effective, offering a cost-efficient, reliable, and noninvasive solution for transformer monitoring and fault detection. This paper presents an innovative algorithm that utilizes statistical parameters derived from the detailed D1-level wavelet coefficients of the signal as inputs to the ANN. This method provides a novel, real-time approach for accurately differentiating between magnetizing inrush currents and inter-turn faults. Furthermore, the proposed ANN-based model extends its capability to identifying the fault locationdetermining whether the inter-turn fault resides in the primary or secondary winding of the transformer.
Copyright
Copyright © 2025 Vikramsingh R. Parihar. This is an open access article distributed under the Creative Commons Attribution License.
