Recognition of complex and multiple power quality disturbances using wavelet packet-based fast kurtogram and ruled decision tree algorithm

This paper introduces an algorithm based on wavelet packet supported fast kurtogram and decision rules for the identification and classification of complex power quality (PQ) disturbances. Features are extracted from the signals using fast kurtogram, envelope of filtered voltage signal and amplitude spectrum of squared envelop. Proposed algorithm can be implemented for the recognition of the complex PQ disturbances, which include the combination of voltage sag and harmonics, voltage momentary interruption (MI) and oscillatory transient (OT), voltage MI and harmonics, voltage sag and impulsive transient (IT), voltage sag, OT, IT and harmonics. Proposed work has been performed using the MATLAB software. Performance of the algorithm is compared with performance of algorithm supported by discrete wavelet transform (DWT) and fuzzy C-means clustering (FCM).