Automatic Power Quality Events Recognition Using Modes Decomposition Based Online P-Norm Adaptive Extreme Learning Machine

Abstract

This article presents an automatic recognition of power quality events (PQEs) by integrating variational mode decomposition (VMD) with Hilbert transform (HT) and the proposed online P-norm adaptive extreme learning machine (OPAELM). The robust parameters estimation capability from the highly nonstationary PQE patterns is presented using VMDHT method and a novel mode selection scheme is introduced based on the correlation coefficient. Three most efficient power quality indices are extracted and fed as an input to train and test the OPAELM classifier with a few existing advanced classifiers. The distinctive modes extraction, low computational burden, robust antinoise performance, short event recognition time, and outstanding recognition capability are the prime superiority expediencies of the VMDHT-OPAELM method. Finally, the proposed method is developed in Xilinx integrated synthesis environment (ISE) Design Suite 14.5 configured with MATLAB/Simulink software environment and implemented in a high-speed field-programmable gate array digital circuitry hardware platform to validate the cogency in real time.