Grey Theory Based Neuro-Fuzzy Approach for State Assessment of Power Transformer Using Dissolved Gas Analysis

Abstract

Power transformers are key component because power system operation depends on it. The reliability of Power Transformer is essential and hence the monitoring of such equipment is necessary at substation level. The assessment techniques of transformer include various methods. Dissolved Gas Analysis (DGA) is a universally accepted and highly recommended technique for fault diagnosis of Power Transformer. There are several DGA methods for detection of faults to measure the concentration in particle per million in oil sample. Gas concentrations indicate fault type and accordingly health of Power Transformer can be judge. This paper introduces Grey Theory approach in which analysis is carried out based on partial information to help in standardizing DGA interpretation techniques and to identify transformer state assessment through it. Grey Theory is perfectly matched to the said problem as the DGA samples are less. The keystone of Grey Theory is to find target heart degree and from that making a decision. The key gases used for Grey analysis. Grey model built in soft computing tool Adaptive Neuro-Fuzzy Inference system (ANFIS). The results of soft computing tools compared with Grey Model output. The soft computing technique shows certain degree of success to validate benchmarking of Grey Model.