Power quality monitoring and compression using the discrete wavelet transform

Although several methods are used to measure various power system quantities such as voltage, current and frequency, it is still inadequate for post processing of power quality issues due to the possibility of missing significant events relevant to power quality. Therefore, IEC 61000-4-30 recommends storage of the power system raw waveform samples over a certain period. However, due to the large amount data, it is mandatory to compress the data for transmission and storage. In an earlier work, an approach for compression is based on separating the stationary and nonstationary components of the waveform. The stationary components are represented by Fourier coefficients and the nonstationary components are represented by discrete wavelet transform coefficients. However, the filter used has nonlinear phase response and data expansion problem. Therefore, the data is subjected to phase distortion. In this paper, it is proposed to use a biorthogonal 5/3 spline filter which provides linear phase response and solves the data expansion problem. Further, the filter reduces the operation count and yield a better compression ratio. Simulation results confirm the validity of the proposed approach.