Detection and classification of multi-complex power quality events in a smart grid using Hilbert–Huang transform and support vector machine

Abstract

This article examines the potential ability of a chosen Hilbert–Huang transform (HHT) in detecting and identifying a multi-complex electric power quality events (PQE) signal in smart grid power systems under various noise situations. HHT is an active signal processing technique, comprising the units of empirical mode decomposition (EMD) and Hilbert spectral analysis (HSA) to detect the non-stationary electric signals. The function of EMD is to detect and decompose the non-stationary electric signal in different scales of frequency modules ranging from maximum to minimum values, and thereby the attained signals are characterized as intrinsic mode functions (IMFs). HSA details the IMF signals individually to produce a unique Hilbert spectrum which carries the information of original signal. By observing the instant time-varying deviations of frequency and amplitude of the resultant signals, it is possible to categorize the disturbing signals from the original signal. The discussed slots of work is simulated under MATLAB environment, and the results report that the HHT successfully detects the single PQE, complex PQE and multi-complex PQE signals under 25 dB, 50 dB and without noise situations. The outcomes of HHT technique are compared with other transforms such as S-transform and wavelet transform to highlight superior qualities of HHT. The identified PQE signals from HHT are classified using support vector machine to improve its classification accuracy. It is wiser to disclose that the proposed system with inbuilt monitoring and identification of PQE signals will suit present smart grid system.