Benchmarking Traditional Machine Learning and Deep Learning Models for Fault Detection in Power Transformers

TL;DR

This study compares traditional machine learning and deep learning approaches for fault detection in power transformers using a multi-sensor dataset collected at 15-minute intervals over 10 months. Labels are derived from fault indicators and balanced with SMOTE, enabling robust evaluation via accuracy, precision, recall, F1, and ROC/AUC metrics. Random Forest and XGBoost achieve top traditional ML performance (about 86%), while a 1D-CNN leads among DL models (about 86%), with ROC AUC values around 0.93 for strong performers. The results demonstrate that both ML and DL can effectively classify transformer faults on tabular sensor data, with implications for real-time, data-driven maintenance and potential future enhancements through temporal-spatial modeling and domain-informed features.

Abstract

Accurate diagnosis of power transformer faults is essential for ensuring the stability and safety of electrical power systems. This study presents a comparative analysis of conventional machine learning (ML) algorithms and deep learning (DL) algorithms for fault classification of power transformers. Using a condition-monitored dataset spanning 10 months, various gas concentration features were normalized and used to train five ML classifiers: Support Vector Machine (SVM), k-Nearest Neighbors (KNN), Random Forest (RF), XGBoost, and Artificial Neural Network (ANN). In addition, four DL models were evaluated: Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU), One-Dimensional Convolutional Neural Network (1D-CNN), and TabNet. Experimental results show that both ML and DL approaches performed comparably. The RF model achieved the highest ML accuracy at 86.82%, while the 1D-CNN model attained a close 86.30%.

Figures (1)

• Figure 1: Receiver Operating Characteristic (ROC) curves