Transformer Insulation Fault Diagnosis Method Based on Rough Set and Fuzzy Set and Evidence Theory

To deal with incomplete and indeterminate information as well as fuzzy information in transformer fault diagnosis, the paper proposes a novel transformer insulation fault diagnosis method based on rough set and fuzzy set and evidence theory. The method firstly establishes rough model with two-universe of fault sets and symptom sets, and makes symptom sets fuzzy. Then based on the relationships between two universes, the largest fault set is sought based on the known fault symptom information. Finally, the best diagnosis result is achieved based on evidence reasoning. Actual example shows that the proposed method possesses good robustness, and is an effective transformer fault diagnosis method and processing technique for uncertain and incomplete information.