Estimation of salt contamination level on the high voltage insulators surfaces during rainy season using artificial neural network

Power system reliability depends mainly on the environmental and weather conditions which cause flashover on polluted insulators. Such flashover leads to system outages. Near the coastal areas the salt contamination can be rapidly built on the surface of the insulators and form a conducting layer by absorbing wet from the mist. This layer in some cases leads to flashover. Study of contamination of insulator under marine pollution is the aim of this research and the effects of various meteorological factors on the pollution severity have been investigated thoroughly. In the present paper an attempt has been made to estimate the pollution severity under various weather conditions during the rainy season using an artificial neural network. The outlier selection problem has been considered in this work to eliminate few highly dispersed experimental data. The relationship between ESDD with temperature, humidity, pressure, rainfall and wind velocity has been developed using ANN as a function estimator.