Condition Assessment of Transformer Health by using Intelligent Technique

Abstract

The condition of Transformer insulation can be effectively calculated by using the concept of health index. The overall health of transformer can be determined by performing different diagnostic test in the transformer's insulation. The outcome of these tests is used to assess the health and for the estimation of remaining life of transformer. This shows that, the deterioration of transformer insulation takes place with the assistance of several transformer parameters. In this paper, different parameters such as Degree of Polymerization (DOP), Dissolved Gas Analysis (DGA), Insulation Resistance (IR), Dissipation Factor (DF), Frequency Response Analysis (FRA), Breakdown Voltage (BDV), Power Factor (PF), Moisture and Hotspot Temperature are used to develop Fuzzy model and Adaptive neuro fuzzy model for analyzing respective critical conditions of the transformer. The above-mentioned parameters are integrated into one by using Fuzzy logic in Fuzzy Inference System (FIS) and the same is done in case of Adaptive Neuro Fuzzy Logic in Adaptive Neuro Fuzzy Inference System (ANFIS) in MATLAB and it will decide the overall criticality of transformer insulation from which the overall Health index (HI) of Transformer can be found out. These two models give the provision to evaluate the condition of insulation and its criticality level. After calculating the health index by both the models, the results of Fuzzy model and Adaptive neuro fuzzy model are observed and from the observation, it can be said that adaptive neuro fuzzy model has numerous advantages over fuzzy model.