Power Quality Event Classification Using Convolutional Neural Networks on Images

Abstract

There is an increasing need to detect powerquality events due to the surge of such disturbances and the losses they can cause. Detection and classification of such events allow for an immediate response. In this paper, a novel approach to this problem has been detailed, wherein Convolutional Neural Networks have been used to classify power quality events using image classification. These Convolutional Neural Network models use scalograms as input images to carry out the power quality classification task. Scalograms are generated by feature extraction using the wavelet transform. The model is then trained and tested on the same. The model performance has then been evaluated, wherein it was shown to perform well on the validation data set. © 2019 IEEE.