A transformer fault diagnosis method based on small sample unbalanced data set
By constructing a transformer fault diagnosis model based on a small-sample unbalanced dataset, and utilizing a diversified hierarchical bootstrap resampling and expert system-enhanced Bagging-stacking fusion mechanism, the problems of complex fault mode recognition and accurate diagnosis of specific categories in transformer fault diagnosis are solved, achieving high-precision fault identification and stable operation of the power system.
Patent Information
- Authority / Receiving Office
- CN Β· China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- SHENYANG AGRI UNIV
- Filing Date
- 2025-10-17
- Publication Date
- 2026-06-09
AI Technical Summary
Existing technologies for transformer fault diagnosis in the context of small-sample imbalanced datasets suffer from problems such as complex fault mode recognition, inaccurate diagnosis of specific categories, and insufficient model generalization ability. This results in the inability to accurately diagnose transformer faults, threatening the safe and stable operation of the power system.
A transformer fault diagnosis model based on a small sample imbalanced dataset is constructed. By acquiring dissolved gas analysis data of transformers, and combining diversified hierarchical Bootstrap resampling, fast feature selection, base learner combination, category expert system and probabilistic fusion logic module, an integrated model is built for diagnosis.
It improves the predictive accuracy and model generalization ability of transformer fault diagnosis, enhances the level of intelligent transformation of the power system, and ensures the safe and stable operation of the power grid.
Smart Images

Figure CN121412830B_ABST
Abstract
Citation Information
Patent Citations
Oil-immersed transformer fault diagnosis method based on data balance and GBDT
CN118114155A
Transformer bushing fault diagnosis method based on data balance
CN118861821A