Metrics
1 Record Views
Thesis
An Intelligent Differential Protection Scheme Based Fault Classification in Transformers Using Machine Learning
University of West Florida Libraries
Master of Science (MS), University of West Florida
2026
Abstract
This research investigates the application of machine learning (ML) techniques to enhance differential protection for power transformers by accurately classifying both the type and location of faults in a transmission line. Conventional protection schemes face challenges in providing fast and precise fault diagnosis, which is critical for maintaining grid stability and minimizing outages. To address this, a detailed power system model featuring a dual-slope differential protection scheme was developed in PSCAD. A comprehensive dataset was systematically generated by simulating a wide range of fault scenarios, including various fault types, locations, and loading conditions. The differential current signal data from these simulations were used to train and evaluate multiple supervised ML classifiers within the MATLAB Statistics and Machine Learning Toolbox. Two distinct classification tasks were addressed: identifying the fault type and determining the fault location. Model performance was rigorously assessed using five-fold cross-validation and an independent test set to ensure robust and generalizable results.
The findings demonstrate that ML classifiers can effectively distinguish fault characteristics from differential current data. For fault type classification, an Ensemble Subspace K-Nearest Neighbors (KNN) model achieved the highest test accuracy of 97.14%. For the task of fault location, a Cubic Support Vector Machine (SVM) model proved most effective, achieving a superior test accuracy of 91.43%. The study concludes that the integration of ML provides a significant enhancement to traditional transformer protection, with specific algorithms showing exceptional performance for different aspects of fault diagnosis. This work validates a powerful framework for developing and evaluating intelligent protection schemes, confirming the potential of data-driven methods to improve the reliability and accuracy of power system protection.
Details
- Title
- An Intelligent Differential Protection Scheme Based Fault Classification in Transformers Using Machine Learning
- Resource Type
- Thesis
- Contributors
- Bhuvaneswari Ramachandran (Committee Chair)Ezzat Bakhoum (Committee Member)Bassam Shaer (Committee Member)
- Publisher
- University of West Florida Libraries
- Format
- pdf
- Number of pages
- 59
- Copyright
- Permission granted to the University of West Florida Libraries by the author to digitize and/or display this information for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires the permission of the copyright holder.
- Identifiers
- 99381757109106600
- Academic Unit
- Dr. Muhammad Harunur Rashid Department of Electrical and Computer Engineering
- Language
- English
- Awarding Institution
- University of West Florida; Master of Science (MS)
- Theses and Dissertations
- Master of Science (MS), University of West Florida