Effect of Power Factor of a Synchronous Machine on Eccentricity Faults Classification Accuracies

Abstract

The research work studies the effect of changing power factor of a Salient Pole Synchronous Machine (SPSM) on eccentricity fault classification accuracies of machine learning and deep learning models. The SPSM was subjected to static eccentricity (SE) and dynamic eccentricity (DE) with a severity of forty percent. Data was collected at different operating conditions, such as lagging, leading, and unity power factor. The data was used to train an Artificial Neural Network (ANN) and a one-dimensional Convolutional Neural Network (1D CNN) for eccentricity fault classification. Results show that the SPSM’s changing power factor significantly affected the classification accuracy of both neural networks.