An alternative approach for harmonic measurements in distribution systems with empirical wavelet transform

Abstract

Addition of a large number of nonlinear loads in the electric power grid has increased the volume of harmonics, which necessitates its accurate and reliable measurement for both control and metering purpose. Legacy techniques like FFT underperform with the present grid scenarios, wherein the frequency varies over time besides the variations of the harmonic combinations. This paper presents an alternative adaptive technique, Empirical Wavelet Transform (EWT), for individual frequency component extraction and measurement of nonstationary signals. The performance of EWT is evaluated through different test cases depicting typical distribution grid scenarios, where it is tuned to capture the frequency and magnitude variations of nonstationary signals, and the results are presented. Similarly, dynamic changes in frequency and magnitudes of the input signals are introduced and the time–frequency correlations are analyzed and presented.