Multi-criterion fuzzy model for power transformer insulation monitoring

In the present paper, a new multi-criterion-based fuzzy logic (FL) model has been proposed to determine the health index and health status of power transformers. To facilitate an effective decision-making environment, the proposed model incorporates multi-criterion analysis and various diagnostic parameters concerning the transformer. Most significant parameters of the oil-paper insulation that represents the degradation as potential aging markers have been considered for developing the proposed model. In addition to these, other parameters including visual inspection, oil age, and the number of faults detected are also included in the loop. To validate the proposed model, monitoring data of 200 in-service identical transformers that belong to the local utility have been adopted. Comparison has been made between the output health indices of the present proposed method and the diagnostic expert model proposed previously by the author's group. The similarity index of the results in both models is found to be 94%. It is observed that existing fuzzy logic methods have shortcomings such as a large number of rules and complexity issues. Thus, routing to an inaccurate and non-reliable output. However, it is inferred that multi-criterion analysis along with FL may certainly overcome all such shortcomings. It is envisioned that the proposed model is accurate and reliable. Importantly, the proposed model is simple with less computational effort and is easily adoptable by the utilities and engineers.