Blackboard expert system of multi-expert cooperating diagnosis for transformers insulation

This paper uses a new knowledge base designed by blackboard prototype to develop an expert system (ES) for transformer fault diagnosis for the first time. According to object-oriented programming (OOP) principle, the knowledge base of this expert system is composed of many modularization components based on artificial neural network (ANN), fuzzy logic (FL), IEC DGA standards and expert experiences, which are core portions of fault diagnosis. The expert system succeeds in diagnosing the transformer´s insulation using multi-expert cooperating method. When the expert´s opinions are conflicting, original and feasible methods are brought forward to resolve this problem. The implementations are discussed in detail. It is confirmed that the ES is satisfactory in practice