Fusion of Signal Processing Techniques to Design an Algorithm for Determination of Single Stage and Multiple Power Quality Events

Determination of the power quality events in the power system is a critical issue, which needs immediate attention for the improvement of service quality. This paper used the fusion of signal processing techniques to design an algorithm for the determination, classification, and localization of the power quality disturbance (PQD) in the time domain. Investigated PQDs included both the single-stage PQDs (SPQD) and multiple PQDs (MPQD). A combined power quality detection index (CPDI) has been designed considering the fusion of Stockwell transform (ST) and Hilbert transform (HT) to detect PQDs by extracting two signal characters using ST and one signal character using HT. Fusion of ST, HT, alienation coefficient (ACF), and Wigner distribution function (WDF) was used to design a power quality disturbance location index (PDLI) to localize PQDs in the time domain. Classification of SPQDs and MPQDs was performed using five signal features extracted using ST, ACF, and WDF. Five features have been computed from the voltage signal with a PQD to design decision rules for classifying the PQDs using a decision tree. The accuracy of PQD recognition has been tested on 125 datasets of every PQD, and it has been found to be greater than 99% in noise-free conditions and greater than 98% in noisy conditions, which is better compared to an algorithm reported in the literature that uses ST and HT. In this work, PQDs have been generated with the help of mathematical formulations in compliance with the IEEE-1159 standard. A detailed study has been carried out with the help of MATLAB software.