Evaluation and Prediction of Contamination Level in Insulators Based on the Leakage Current Characteristics Using Neural Network

摘要

This effort demonstrate the serious issues of coastal, industrial and polar region pollutions on insulator surface and their Leakage Current (LC) and flashover voltage on insulator. The disc type porcelain insulator is tested under normal and abnormal conditions for the surface defect near HV electrode and far from HV electrode, with different pollutants such as marine, industrial and polar. The cavity size, location and pollution level decides the flashover performance of the insulator. The test is performed using the standard IEC60507 at artificial test chamber and leakage current is continuously recorded. Finally the recorded leakage current is given as input to Back Propagation Neural Network (BPNN) to predict the level of contamination severity and the test results are compared with normal and abnormal condition.