Window Optimisation of Power Quality Signal Detection using Gabor Transform

Abstract

This paper presents power quality analysis on different signal characteristics, namely instantaneous sag, momentary sag, temporary sag, instantaneous swell, momentary swell, and temporary swell. Power quality signals were analyzed using linear time-frequency distribution (TFD) namely short-time Fourier transform (STFT) and proposed Gabor transform (GT), and the best technique for power quality detection was determined based on the performance analysis of varied window length. Optimum window length for different signal characteristics which are effective and reliable for developing real-time monitoring system was employed using MATLAB. From a time-frequency representation (TFR) results based on STFT and GT, parameters such as instantaneous RMS voltage, VRMS(t) and instantaneous total waveform distortion, VTWD(t) had been extracted. In finding the best technique for power quality, the TFDs had been compared in terms of accuracy, memory and computational complexity of the analysis. Based on the performance analysis conducted, GT was able to compute with high accuracy with 94% averagely as well as low memory size by 6% compared to STFT. Hence, GT is considered as the best TFD, and recommended for low cost PQ monitoring system.