Prediction of Ferroresonant Modes using Artificial Neural Network (ANN)

In this paper, an investigation into the unpredictability at the onset of ferroresonant modes, attributed to its nonlinear characteristic, is conducted using Artificial Neural Network (ANN). The accuracy performance of the neural network for the training target dataset is evaluated using the regression plot `with R (training) = 0.9997. The regression plot for the prediction accuracy of the sample dataset on the onset of ferroresonant modes is R (target) = 0.97765, which is considered accurate when predicting outside the range of the target data set of capacitance.