Recognition of single-stage and multiple power quality events using Hilbert–Huang transform and probabilistic neural network

Abstract

This article presents the Hilbert–Huang transform based algorithm for recognizing the single-stage and multiple power quality events comprising voltage sag, swell, notches, spikes, harmonics, transients, and flicker. The change of the state during the course of the single-stage power quality event is called a multiple power quality event. Hilbert–Huang transform is an adaptive signal processing technique that combines empirical mode decomposition and Hilbert spectral analysis and makes it an attractive tool for analysis of power quality events. A synthetic database of power quality events is generated in MATLAB (The MathWorks, Natick, Massachusetts, USA) as per Standard IEEE-1995. The significant features are extracted from the instantaneous amplitude, phase, and frequency contours of intrinsic mode functions of each disturbance, and the events are classified using the probabilistic neural network technique.