Summary
The growing complexity of modern power systems, fuelled by the integration of renewable energy sources and the demand for greater reliability, underscores the necessity of sophisticated diagnostic tools for ensuring the operational integrity of essential equipment, including high-voltage circuit breakers. This study developed an artificial intelligence (AI) model to diagnose faults in high-voltage circuit breakers within 230 kV transmission systems.
Read more Read lessThe model is designed to identify whether the failed component corresponds to the circuit breaker itself or another element of the power system. This diagnostic capability supports maintenance teams by prioritizing inspections and reducing uncertainty during post-event analysis. Ultimately, the model aims to optimize the identification of fault patterns and trends in electrical components associated with the switchgear, and facilitate the analysis of operational events.
Machine learning techniques, such as the Extra Tree Classifier, Decision Trees, and Random
Forest, were used to analyze historical maintenance fault data. To address the problem of data imbalance, the SMOTE oversampling technique was applied. The Random Forest model achieved 99.7% accuracy in diagnosing breaker faults, optimizing response times by 15%, and reducing false positives. This system can be integrated into predictive maintenance strategies in the electrical sector to allow for the early detection of faults and improve operational safety.
Additional informations
| Publication type | Session Materials |
|---|---|
| Reference | A3_12035_2026 |
| Publication year | |
| Publisher | CIGRE |
| Country | Colombia |
| Study committees | |
| File size | 425 KB |
| Price for non member | 30 € |
| Price for member | 30 € |
Authors
GALINDO VARGAS Adriana Carolina - Intercolombia; FANDIÑO OLAYA Gonzalo - HMV
