Estimador inteligente do momento ótimo de limpeza de isoladores de vidro a partir da medição da corrente de fuga

Abstract

In the thesis, an intelligent estimator of the optimal cleaning time for glass insulators is proposed based on the measurement of the leakage current. For this purpose, ambient pollution tests were conducted outdoors to record the pollution levels on the surface of glass insulators and investigate the evolution of these levels over time. The pollution levels were systematically measured, along with the temporal data of each measurement. To simulate the identified ambient pollution levels, artificial pollution electrical tests were performed in a high voltage laboratory. This was done by depositing pollution solutions on the insulators, with each solution differentiated by the levels of Equivalent Salt Deposit Density (ESDD) and Non-Soluble Deposit Density (NSDD). In the laboratory tests, ESDD, NSDD, leakage current, and surface discharge signals were recorded, along with humidity and temperature information. Furthermore, during the artificial pollution tests, ESDD and NSDD values associated with the occurrence of surface discharges were determined to serve as pollution level thresholds. Using the data derived from leakage current, pollution levels (ESDD and NSDD), and time, it was possible to estimate when the pollution level would become high enough to cause surface discharges on the glass insulators. Time series forecasting models based on artificial intelligence were developed to forecast pollution levels (ESDD and NSDD values) and assist in estimating the optimal time for cleaning maintenance of the insulators for 69 kV systems.