This　paper　proposes　a　transformer　fault　diagnosis　method　based　on　DC-SMOTE-SSA-SVM,　aimed　at　improving　the　accuracy　of　transformer　fault　diagnosis　in　power　systems.　The　paper　first　highlights　the　importance　of　power　transformers　and　the　challenges　associated　with　fault　diagnosis,　particularly　in　cases　of　imbalanced　data　samples.　To　address　this　issue,　the　DC-SMOTE　algorithm　is　employed　to　augment　fault　samples　derived　from　dissolved　gases　in　oil,　thereby　enhancing　the　classification　accuracy　of　the　model.　Subsequently,　a　fault　diagnosis　model　is　developed　by　integrating　the　global　optimization　capability　of　the　Sparrow　Search　Algorithm　(SSA)　with　the　classification　ability　of　Support　Vector　Machine　(SVM).　Through　case　studies,　the　effectiveness　of　the　proposed　method　in　improving　fault　diagnosis　accuracy　is　validated.　Experimental　results　demonstrate　that,　compared　to　the　original　dataset　and　the　dataset　processed　with　SMOTE,　the　DC-SMOTE-SSA-SVM　model　significantly　improves　both　accuracy　and　AUC,　exhibiting　strong　overall　performance　and　high　reliability,　thus　meeting　the　practical　requirements　for　operation　and　maintenance.　©　2025　IEEE.