Transformer Fault Diagnosis Based on Ontology and Dissolved Gas Analysis

This paper proposes an ontology model for accurate and efficient transformer fault diagnosis using an explicit, formal and machine-readable format. The model makes use of ontology to represent formally faults and their features such as causes, symptoms, effects, which form a transformer fault diagnosis knowledge base. Moreover, the model can be employed to exchange and reason information for transformer fault diagnosis. In this study, a dissolved gas analysis method is encoded into an ontology-based knowledge base, and real fault samples are used to verify the developed model. The experiment results demonstrate that the proposed model can accurately diagnose various faults.