Innovative diagnosis of transformer winding defects using fuzzy and neutrosophic cross entropy measures

Abstract

Power transformers are critical components in electrical power systems, and their reliable operation is essential for the stability of power grids. This research introduces an innovative methodology for the diagnosis and taxonomy of transformer winding defects, leveraging fuzzy and neutrosophic cross entropy measures. Traditional transfer function (TF) analysis, while widely used, often depends on expert interpretation and is limited in detecting minor and incipient faults. To address these challenges, we propose the integration of fuzzy cross entropy measure (FCEM) and neutrosophic cross entropy measure (NCEM) with TF analysis. The methodology encompasses several critical steps: measuring the frequency response of transformer windings under various faults, normalizing the transfer function responses, extracting lower and upper bounds, and constructing fuzzy and neutrosophic sets of faulty and healthy conditions. Subsequently, cross entropy values between healthy and faulty conditions are computed to identify and classify defects. The proposed approach is validated through a real case study, demonstrating its effectiveness in automating fault detection and reducing reliance on expert knowledge. The results indicate that the highest cross entropy measure values accurately reflect the presence of transformer winding defects across different frequency bands. This innovative approach not only enhances the accuracy of fault detection but also greatly advances intelligent fault diagnosis in power transformers, offering a more reliable and user-friendly solution for maintaining the integrity of power systems.