An Adaptive Basis Technique Applied to Mode Extraction of Non-Stationary Low Frequency Oscillations

Extraction and visualization of the temporal pattern of modal oscillations have been quite informative to the operators in the prediction, estimation, and model reduction of dynamical power systems patterns and parameters. This manuscript advocates several signal-processing techniques which rely on the adaptive basis for mode extraction of Low-Frequency Oscillations (LFOs) pertaining to the non-stationary frequency deviation signals. The paper provides a comparative assessment of the Variational Mode Decomposition (VMD), as the non-stationary signal processing technique, to Empirical Mode Decomposition (EMD) and Frequency partition-based Mode Extraction (FPBME) algorithms. Hilbert Spectral Analysis (HSA) is applied to the modal oscillations obtained from the above non-stationary techniques to visualize and characterize Instantaneous Energy (IE) and Instantaneous Frequency (IF) vs. Time plots. VMD the technique discussed in the paper has been applied to both conventional Kundur two-area system and mixed generation of Wind energy incorporated into Kundur two-area system with signals obtained from the PSSE environment.