Analysis of Low Frequency Oscillations using improved Hilbert-Huang Transform
Yang, D.C.,Rehtanz, C.,Li, Y.
DOI: https://doi.org/10.1109/POWERCON.2010.5666567
2010-01-01
Abstract:As a non-linear and time-varying tool, Hilbert-Huang Transform (HHT) has been widely used to analyze Low Frequency Oscillation (LFO) signals in power systems. It utilizes Empirical Mode Decomposition (EMD) to decompose the LFO signals into a collection of Intrinsic Mode Functions (IMFs), and then the instantaneous parameters including magnitude, frequency of every IMF can be calculated by applying the Hilbert Transform. However, HHT suffers from a number of shortcomings. In order to dispose the inherent problems of conventional HHT, an improved HHT is proposed based on Symmetrical Extrema Extension (SEE) method and frequency heterodyne technique. In this paper, SEE method is employed to expand the original signal during the processing of EMD and frequency heterodyne technique is used to overcome the mode-mixing phenomena. Next, the principle and influences of different shifting frequency factors are introduced. Based on these, the steps and flow chart of improved HHT are proposed. The results of testing signals and simulation model show that the improved HHT not only diminishes the influences of End Effect, but also expands the application of HHT. It is feasible and effective to overcome special mode-mixing problem.