Classification of Power Quality Disturbances in Emerging Power System using S-transform and Support Vector Machine

Abstract

Classification of Power Quality Disturbances (PQDs) becomes a key issue for end-users in order to enhance the Power Quality (PQ). S-transform (ST) based Support Vector Machine (SVM) classifier is presented in this paper for classification of PQDs in emerging power systems. S-transform is a new and powerful signal processing technique used for feature extraction in the PQDs classification algorithms. In this article, an energy threshold-based algorithm for the classification of PQDs is also presented. In this research work three PQDs namely voltage sag, swell, and interruption are considered under study. The data required for the analysis of PQDs is generated by simulating the emerging power system with these PQDs in a MATLAB Simulink environment. By using S-transform, the non-stationary voltage signal was transformed to extract the appropriate features needed for the classification of PQDs. For automatic classification of PQDs, the features extracted using S-transform are used for training and testing the SVM classifier. The proposed approach uses less features hence the classifier requires less memory and learning time. The simulation results reveal that S-Transform in combination with SVM is an effective way of classifying different PQDs in emerging power system.