An improved Hilbert Vibration Decomposition method for analysis of low frequency oscillations

In this paper, a masking-based Hilbert Vibration Decomposition (HVD) method is proposed to analyze nonlinear and non-stationary power system data. An adaptive masking signal method derived from the data itself is used to improve the decomposition ability of the HVD method. Techniques to compute the masking signal are described and precise criteria to increase the modal resolution of the HVD method are derived. Simulation results using both synthetic and simulated data show that the technique can be used to characterize complex oscillatory processes in power systems.