Time-Frequency Characterization of Dynamic Phase Synchrony in Power Grids

Dynamic Phase Synchrony (DPS), a key concept for understanding time-varying interactions and phase synchronization in oscillating systems such as the brain networks, power grids, and behavioral synchrony in social networks, is explored in this paper. The Hilbert-Huang transform and Nonlinear Mode Decomposition, two advanced signal processing techniques, are employed to characterize the spatial distribution of instantaneous modal phases. Classification techniques based on tensor analysis are also used to classify dynamic modal phase synchrony using multichannel signals. The proposed methods are validated using transient stability data from two test systems and actual system event measurements. The analytical results demonstrate the value of phase-based characterization in capturing and interpreting dynamic system behavior. Furthermore, factors affecting the quality of modal estimates are critically examined.