Electromechanical Modes Identification Based on an Iterative Eigenvalue Decomposition of the Hankel Matrix

This paper proposes a novel strategy to precisely extract modal patterns from non-stationary multi-component signals associated with electromechanical oscillations in large-scale power systems. The strategy is composed of two stages: (
i
) a time-frequency representation (TFR) method; and (
ii
) an energy-based operator. The former is equipped with a multivariate and iterative eigenvalue decomposition of the Hankel matrix (IEVDHM) that captures the swing dynamics as a mono-component signal criterion is fulfilled, meanwhile the latter instantaneously estimates the modal information (damping and frequency) through the discrete energy separation algorithm (DESA) that implements the discrete-time energy operators derived from the Teager-Kaiser energy operators (TKEO). The attained results and their comparisons with state-of-the-art techniques confirm the effectiveness and performance of the proposed strategy to demodulate synthetic, simulated and real oscillating signals, even under high noisy conditions, and to be a useful tool for off-line contingency analysis thanks to the capability of differentiating concurrent modes with close frequencies.