August 20, 2025MachinesOpen Access

Fault Classification in Power Transformers Using Dissolved Gas Analysis and Optimized Machine Learning Algorithms

View Full Paper

Ask AI

Bookmark

Optimized k-Nearest Neighbor achieved 92.5% accuracy in classifying faults, outpacing traditional methods significantly.
Conventional methods yielded classification accuracies of only 35.8% for Duval Triangle and 30.8% for Duval Pentagon, showing their limitations.
The framework incorporates Chi-squared feature selection and Random Search for hyperparameter optimization to boost performance.
Enhanced diagnostic accuracy supports more reliable maintenance strategies for critical power transformer assets.

Discussion

No takes yet. Share an insight, caveat, or question.

Machine learning improves fault classification accuracy in power transformers using DGA data, indicating enhanced diagnostics.

Dladla et al. (2025) studied this question.

View Full Paper

Ask AI

Bookmark