Detection of Transformer Faults: AI-Supported Machine Learning Application in Sweep Frequency Response Analysis

Hakan Çuhadaroğlu () and Yılmaz Uyaroğlu
Additional contact information
Hakan Çuhadaroğlu: Institute of Natural Sciences, Sakarya University, Sakarya 54050, Türkiye
Yılmaz Uyaroğlu: Department of Electrical and Electronics Engineering, Faculty of Engineering, Sakarya University, Sakarya 54050, Türkiye

Energies, 2025, vol. 18, issue 10, 1-20

Abstract: In this study, we discussed how the increasing demand for electrical energy results in higher loads on transformers, creating the need for more effective testing and maintenance methods. Accurate fault classification is essential for the reliable operation of transformers. In this context, Sweep Frequency Response Analysis (SFRA) has emerged as an effective method for detecting potential faults at an early stage by examining the frequency responses of transformers. In this study, we used artificial intelligence (AI) and machine learning (ML) techniques to analyze the data generated by SFRA tests. These tests typically produce large datasets, making manual analysis challenging and prone to human error. AI algorithms offer a solution to this issue by enabling fast and accurate data analysis. In this study, three different transformer conditions were analyzed: a healthy transformer, a transformer with core failure, and a transformer with winding slippage. Six different machine learning algorithms were applied to detect these conditions. Among them, the Gradient Boost Classifier showed the best performance in classifying faults. This algorithm accurately predicted the health status of transformers by learning from large datasets. One of the most important contributions of this study is the use of gradient boosting algorithms for the first time to analyze SFRA test results and facilitate preventive maintenance through the early detection of transformer failures. In conclusion, this study presents an innovative approach. The interpretation of offline SFRA results through various artificial intelligence-based analysis methods will contribute to achieving the ultimate goal of reliable online SFRA applications.

Keywords: sweep frequency response analysis (SFRA); machine learning; gradient boost classifier; artificial intelligence; online SFRA (search for similar items in EconPapers)
JEL-codes: Q Q0 Q4 Q40 Q41 Q42 Q43 Q47 Q48 Q49 (search for similar items in EconPapers)
Date: 2025
References: Add references at CitEc
Citations:

Downloads: (external link)
https://www.mdpi.com/1996-1073/18/10/2481/pdf (application/pdf)
https://www.mdpi.com/1996-1073/18/10/2481/ (text/html)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:gam:jeners:v:18:y:2025:i:10:p:2481-:d:1653883

Access Statistics for this article

Energies is currently edited by Ms. Agatha Cao

More articles in Energies from MDPI
Bibliographic data for series maintained by MDPI Indexing Manager ().