A fuzzy ultrasonic system for estimating degradation of insulating oil

Abstract

This paper proposes an ultrasonic estimation system for the degradation of the insulating oil of electric power supply aided by fuzzy inference. In general, it is known that the viscosity of insulating oil depends on the acidity. We employ the viscosity as characteristic value for estimating the degradation. First, we show a strong positive correlation between the acidity and the viscosity. Second, we construct a fuzzy estimation system. Third, we infer the acidity of insulating oil by using the system. Our experimental results show that the fuzzy system can identify the acidity with the accuracy of 74% on 50 samples. Thus, this system can successfully estimate the degradation of the oil.