Enhancing Diagnostic Accuracy of Transformer Faults Using Teaching-Learning-Based Optimization

The early detection of the transformer faults with high accuracy rates guarantees continuous operation power system networks. Dissolved gas analysis (DGA) is a technique that used to detect or diagnose based on dissolved gases due electrical and thermal stresses influencing insulating oil. Many attempts are accomplished discover an appropriate correctly fault types, such as Duval Triangle method, Rogers' ratios IEC standard 60599. In addition, several artificial intelligence, classification, optimization techniques merged previous methods enhance their diagnostic accuracy. this article, novel approach proposed introducing new concentration percentages limits gases' help separate conflict between diverse faults. To do so, model established which simultaneously optimizes both so maximize agreement respect actual ones achieving Accordingly, efficient teaching-learning (TLBO) developed accurately solve considering training datasets (Egyptian chemical laboratory literature). TLBO algorithm enhances at significant level, higher than some other DGA were presented in literature. robustness optimization-based confirmed against uncertainty measurement where its not affected by rates. prove efficacy approach, it compared five existing approaches using out-of-sample dataset superior rate reached for different types.