Identification of Voltage Sags in Distribution System Using Wavelet Transform and SVM

Voltage sags have become an important power quality issue in power system over the past several years. Voltage sags are short duration reductions in RMS voltage, mainly caused by short circuits, transformer energizing, and starting on large motors. Aim of this paper is to show that there are certain types of voltage sags that should be considered in the analysis of monitoring data as showed the analysis of a large number of recordings. In the paper voltage sags were analyzed based on a PSB simulation model. Then presents a method to identify the voltage sag in distribution system using a novel combination of wavelet transform and support vector machines. The simulated waves were discomposed into 7 levers using wavelet transform, afterwards, the energy features, which was extracted from the wavelet coefficients under different levers, were employed as the inputs of the support vector machines to identify and classify the voltage sags. The results of simulation and case study show the proposed method is simple and validity.