Redefined Power Quality Indices for Stationary and Nonstationary Power Quality Disturbances

Abstract

The recommended power quality (PQ) indices for electric power systems are defined based on the discrete Fourier transform (DFT). DFT is mostly evaluated applying fast Fourier transform (FFT) algorithm which requires signals to be periodic in nature and therefore provides accurate results in case of stationary PQ disturbances only. In case of nonstationary disturbances, signals become aperiodic in nature, and FFT produces erroneous results due to spectral leakage phenomena. In addition, FFT is incapable of providing any time information, which is a significant property when dealing with time-varying PQ disturbances. FFT can provide only amplitude-frequency spectra of the signals. Thereby this paper presents a framework of redefining stationary and nonstationary PQ disturbances employing time-frequency distribution (TFD) method. Reduced interference distribution (RID), which is a member of Cohen’s class TFD, is studied here to define and formulate PQ indices for electric power systems. RID represents the most suitable properties in case of time-varying PQ disturbances comparing with other TFD methods. The results of four synthetic examples considering stationary and nonstationary PQ disturbances imply that the redefined RID-based PQ indices are closer to the actual values, and provide more accurate results than the existing transient PQ indices and traditional FFT-based method.