Implementation of Real-Time Power Swing Detection Based on Support Vector Machine of Distributed Generation System Using LabVIEW MyRIO

This paper presents a method for power swing and fault diagnosis of distributed generation (DG) system based on real-time and Support Vector Machine (SVM) classifier using LabVIEW NI myRIO. The method adopts real-time SVM classifier to identify the power swing and fault occurring in power system. The data input to SVM using Wavelet Packet Transform (WPT) extract current signals that calculate Rate of Change of Relative Wavelet Energy and symmetrical components. The process of training the SVM using a K-folded cross validation process for determining the appropriate values of parameter sigma σ in RBF kernel parameters can minimize the classification error. The experimental results, the LabVIEW NI myRIO processor can be detecting the power fluctuations that occur in the power system. The proposed method can be successfully real-time to detect power swing and provide power swing blocking signal, but unblocking when fault occurring during power swing.