A Support Vector Machine (SVM) Based Voltage Stability Classifier

This paper proposes a new concept of support vector machine based voltage stability classifier using local measurement data at a substation. The excellent performance of SVM for binary classification matches the real-time identification of long-term voltage stability from time-series data by local measurements. The methodology can provide a warning signal during system voltage instability that aims to leave the time to perform immediate corrective actions to stop system degradation by the effect of major disturbances.