A Novel Fuzzy Based Technique For Transformer Health Index Computation

Abstract

Transformers are lavish machines installed in a power system that is inherently designed to have high efficiency and reliability. But transformers contain inflammable oil in the vicinity of high voltage elements, which if mishandled, may result in transformer failures and catastrophic explosions. Thus, to enhance the safety and reliability of a transformer, its health assessment is imperative. For this purpose, Health index is an apt indicator. But traditional health index computational techniques have certain limitations which are surmounted using the approach mentioned in this research paper. The computational algorithm used to develop a fuzzy model in this paper eliminates need of a diagnostic expert and have the capability to assess arbitrarily complex diagnostic test results. A decision logic is established to select the non-routine diagnostic tests to be performed by assessing routine test results. Particular attention is given to take into account the effect of factors like operation and maintenance history, expected life, etc. for computation of health index. This paper will be helpful for researchers and electrical utilities involved with health assessment and optimum maintenance of transformers.