Diagnostics of Current Technical State of Transformer Equipment Using the Analytic Hierarchy Process

The new method of diagnostics of current technical state for power equipment at substations of electrical networks and industrial plants is proposed. It is based on the analytic hierarchy process and fuzzy logic. It was shown that it is possible to make a plausible prediction on the failures and fault reasons for power equipment and to minimize the time for decision making and fault reason elimination using pairwise comparisons of expert evaluations. To perform such analysis, it is necessary to have high-qualified experts and symptoms of defective state of power equipment. The comparison of these symptoms is realized using nine intensities of importance of the fundamental scale of Saaty. Fuzzy evaluations are based on the attempts to digitize verbal information (i.e. to represent information as linguistic variables by words or phrases in the form of numbers). Symptom ranking for possible faults is performed according to expert preferences that allows deriving the most considerable symptoms for each suggested defect and making a decision on a further operation of power equipment. The reliability of the proposed method is validated by calculations that demonstrate the adequacy of the model applied to transformers with high temperature superconducting windings (HTS transformers). It was shown that a conclusion on the further operation of a superconducting transformer or its removal for repair could be made if some fuzzy symptoms are observed with their evaluation using the scale of intensities of importance in the analytic hierarchy process. The new mathematical model developed with the use of fuzzy relations of transformer faults includes the elements of prediction for possible failures of power equipment in power supply systems and can be implemented into the systems of short-term forecasting and on-line diagnostics.