A novel forecasting model of contaminated insulator flashover voltage

To solve the problem of the selecting of the external insulation under complex circumstance conditions, a flashover voltage forecasting model of contaminated insulators based on double ANNs is proposed in the paper. The equivalent salt deposit density (ESDD) is the key of flashover voltage on contaminated insulator, and circumstance conditions are also great influence on it. The equivalent salt deposit density (ESDD) value of insulator can be treated as a nonlinear time series and be forecasted by the nonlinear time series ANNs model. The flashover voltage forecasting model consists of two artificial neural networks. The first is used to forecast the equivalent salt deposit density (ESDD) time series and the second is employed to calculate the withstand voltage of insulator. A series of artificial pollution tests show that the results of the forecasting model is acceptable in engineering application.