The use of wavelets for the monitoring and diagnostic of surface state of HV polluted insulators

Novel colour image encoding system combined with ANN for discharges pattern recognition on polluted insulator model

“…Figs. [4][5][6] represent an example of the extracted images from video recorded when using moistened sand pollution of 0.02 g/cm 2 NSDD. For this, the applied voltage is increased (with ∼2 kV/s) from zero to flashover voltage.…”

“…They are intended to perform and operate effectively under the most severe climatic conditions. Consequently, monitoring the performance of these insulators under pollution is of the upmost importance to maintain safe and continuous operation of power on the network [1][2][3][4][5][6]. If insulators are not correctly monitored especially under severe pollution conditions, the flashover can occur through the following steps: accumulation of contamination layer, wetting of the insulator, increasing of the leakage current, the formation of dry band arcs and finally the extension of such arcs to cover the leakage path [1,[7][8][9].…”

“…Based on the interpretation of the signals of LC, discharges or applied voltage, several studies have been devoted to monitor and/or diagnosis polluted insulators. Chaou et al [6] also used the wavelets to monitor and diagnose the surface state of the HV insulator model under uniform and non-uniform pollution. For this, continuous wavelet transform (CWT) and discrete wavelet transform (DWT) have been used to detect electrical discharges on the applied voltage signal and the LC one, respectively.…”