Flashover Prediction of Polymer Insulators Based on Dynamic Modeling of Pollution Layer Resistance Using ANN

In this article, a novel dynamic approach is proposed to analyze pollution layer resistance of polymer insulators. One of the main challenges in dynamic modeling of pollution layer resistance is the effects of pollution layer conduction and pollution layer thickness on discharge length due to change in pollution layer thickness and wetting rate in real outdoor conditions. The represented approach investigates the effects of such parameters on variation of discharge length. Then, an artificial intelligence (AI) algorithm based on artificial neural network (ANN) is developed to predict flashover voltage (FOV) of polluted insulators according to the prediction of critical discharge length. Calculations are done based on experimental data of leakage current (LC) and voltage of insulators in different pollution conditions. Obtained results by the proposed approach show a close correlation with experimental results. In addition, various validation tests are performed to investigate the effects of hydrophobicity class (HC) of polymeric surface and pollution type on the quality of the represented analysis. Results of the validation also show a close correlation between calculations and real data.