Wavelet Aided KNN Based Power System Disturbances Classification

In this work, few most commonly occurring Power System Disturbances (PSD) have been considered for their detection and classification with the help of K Nearest Neighbor (KNN) based Machine Learning (ML) technique. Experimentally created disturbance signals have been processed using Discrete Wavelet Transform (DWT) to obtain time and frequency domain information before they have been fed to KNN based classifiers. The switching, line faults and load change have been chosen as disturbing creating events. The performance of the proposed method has been compared with linear discriminant and linear support vector machine based methods. The accuracy level of the proposed method has been shown to be 97.1%.