Condition Based Assessment and Diagnostics of Transformer in Smart Grid Network Using Adaptive Neuro Fuzzy Inference System Framework

In smart grid infrastructure, power transformers are one of the most integral assets that must be constantly monitored during their operational life. Several diagnostic methods, such as interfacial tension (IFT), moisture presence, degree of polymerization (DP), harmonics, temperature rise furans compound have detrimental effect on transformer insulation. This research article introduces a novel adaptive neural fuzzy inference system (ANFIS) model to analyse the insulation degradation for oil and paper based on the datasets of moisture contents, IFT, harmonics and temperature rise within the insulation. The proposed study findings are verified using real data gathered from various transformers used in utilities, industry and literatures including the impacts of electric vehicles (EV) and Distributed energy resources (DER). Time varying harmonics and temperature effect is important aspect of this research. The novel proposed ANFIS model is more than 90% efficient and errors of less than 1% for training and testing data sets utilized for study.