DTCWT-SVM Based Identification of Single and Multiple Power Quality Disturbances

Abstract

This paper presents the use of a modified Digital Signal Processing (DSP) technique to identify and classify various single stage and multiple power quality (PQ) disturbances such as transient, flicker, interruption, swell, sag, swell with harmonics, flicker with harmonics, sag with harmonics, interruption with harmonics, and transient with harmonies accurately. A PQ monitoring algorithm based on Dual Tree Complex Wavelet Transform (DTCWT) as a DSP technique and support vector machine (SVM) as an AI tool has been proposed for this purpose. Based on IEEE-1159 standard, a synthetic data bank of these disturbances has been generated for the simulation studies using MATLAB which plays a vital role in defining the detection capability of DSP technique and generalization capability of AI technique. Using DTCWT, features of various PQ disturbances are extracted and given it as an input to SVM classifier for the classification. The extracted feature dataset has been trained and tested using SVM classifier, as a result of this accurate classification results are obtained from the proposed algorithm. Additionally, the accuracy of the proposed methodology is compared with other classifiers for identification of the PQ disturbances.