Statistical classification of contamination in glass insulators by reading its spectrum

Abstract

This paper proposes to use a statistical approach for classifying the contamination level of insulators used in high voltage transmission lines. This studied case considers the voltage applied on glass insulators used in a chain of six elements of a 69 kV power line. When submitted to high voltage, the insulators radiate radio frequency signals up to frequencies of some GHz. In this frequency range the signal can be detected using portable antennas, and then be applied to post processing. These received spectra are analyzed by statistical methods, mean and standard deviation, and the cross relation between them is used to classify the pollution level in insulators. To reduce complexity and cost a simple data processing using the mean and the standard deviation are used. Two types of glass insulators are considered: clean or polluted, and to cover the entire frequency range from 30 MHz to 1 GHz two types of antennas were needed. The results obtained with the mean and the standard deviation show that the first 100 MHz bandwidth can be used for statistical of the insulator pollution level.