Classification of power system transients using discrete wavelet transform and random forest method

Power system transients are the main power quality problems. Classification of the power system transients is a common problem in power quality studies. Impulsive and oscillatory transients are caused by a variety of sources in power system including natural and power system related. Their effects on power systems are also differ. Therefore, power system transients should be classified as impulsive and oscillatory correctly. In power quality studies, usually, the transients are not classified into impulsive and oscillatory. In this study power system transients are classified into impulsive and oscillatory groups by using discrete wavelet transform and random forest method. The implementation was realized in LabVIEW. The proposed classification method was tested by real power quality data. The results show that this method can be used successfully for the classification of power system transients.