The fault diagnosis algorithm for transformer based on Extenics and rough set theory

As one of the most important equipments to guarantee power gird operating safely and stably, the fault diagnosis of transformer has most important sense. Since the general diagnosis methods such as DGA and attributes´ reduction of rough set haven´t enough precision for transformer´s fault diagnosis, Extenics and rough set theory are brought into diagnosing the fault of transformer. Using attribute predigesting method in rough set theory to classify the attribute term which needed by each fault diagnosis. Then building matter element model for transformer´s fault diagnosis. Using DGA testing datum to be attribute set and the transformer´s standard fault model to be the decision set for transformer´s fault diagnosis. Utilize association function from Extenics to count each fault degree. Define the fault accepting rule to get transformer´s fault. Use this method to diagnose one fault transformer and the diagnosis result matched case of fact fault. Apply this method to diagnose 76 DGA testing data and the right ratio of diagnostic result is better than IEC method.