Expert system for condition monitoring of power transformer using fuzzy logic

Abstract

The status of the health of a power transformer is determined by continuously monitoring the condition of its solid and liquid insulations. Appropriate condition monitoring assists in improving the service life of the transformers by initiating suitable actions based on their current health condition. In the present paper, a fuzzy logic model which utilizes the data gathered from various diagnostic tests has been proposed to determine the overall health index of power transformers. The proposed method also determines the individual health index of transformer oil and paper insulations. Furthermore, the model identifies the incipient faults present within the transformers and handles all situations corresponding to single or multiple faults. The proposed fuzzy model overcomes the shortcomings of conventional fault diagnostic methods which are based on dissolved gas analysis. Thirty transformer oil samples of Indian Railways collected from different traction sub-stations have been tested to prove the efficacy and reliability of the technique.