Prediction of Polymeric Insulator Leakage Current under Water Droplets Using Neural Network

In this paper, the parameters of electric field and leakage current (LC) amplitude in polymeric insulator are calculated using finite element method (FEM) and predicted in the neural network. The variable angle and electrical conductivity of water droplets are changed.  Increased electric field due to the presence of water droplets on the insulator surface causes the insulator breakdown probability. In this paper, for predicting the on the polymeric insulator leakage currents a neural network model was proposed, which could arrange for references to prevent contamination flashover. Through the study of a LC data obtained via the FEM simulation, the characteristics of the LC combined with BP neural network. Simulation results illustrate that enhancement in leakage current and electric field values are created when the values of the electrical conductivity and drop contact angle is increased.