A novel approach for classifying the intensity of audible discharge noise caused by contaminated insulator strings

This paper presents a novel approach for classifying the annoyance levels of audible discharge noise caused by contaminated insulator strings. In the course of the study, the nature of audible discharge noise was studied at first with SSG (sound spectrograph) analysis for the sound generated in a plain anechoic room, which revealed the temporal change and the frequency range of strong spectrum. Exploiting the characteristics of the noise, the audible discharge noise was successfully detected with the developed instrument for the study to the next step. The measured sound data were examined for various annoyance levels. It was found that the shape of the distribution of the amplitude histogram differs according to the annoyance level, which was explained by the difference of discharge phenomena with the observations on the sites. Therefore, the statistical properties of Skewness and Kurtosis have been adopted to classify the annoyance levels. The values of these properties and the suitable time span of the calculation for the classifying have been discussed.