Abstract
Sub-Synchronous Oscillations (SSOs) challenge the stability of modern power systems with high renewable and power-electronic penetration. This paper proposes a fully adaptive Continuous Wavelet Transform (CWT) framework that integrates Welch Power Spectral Density (WPSD)–guided scale selection with a two-stage Nonlinear Least-Squares (NLS) estimator for accurate frequency and damping identification. Three CWT-based strategies are developed: Adaptive CWT (ACWT) for computationally efficient mode extraction, Targeted CWT (TCWT) for enhanced separation of closely spaced modes, and Refined fixed-scale CWT (RCWT) for noise-robust tracking over a uniform frequency grid. Validation using synthetic signals and real-world Phasor Measurement Unit (PMU) data demonstrates that ACWT offers an effective trade-off between computational efficiency and estimation accuracy, TCWT provides superior localization for closely spaced modes, and RCWT delivers consistently robust performance under noisy conditions. Compared with benchmark methods—including Prony–AIC, VMD–Hilbert, EMD-Hilbert, and EWT–Yoshida—the proposed framework delivers improved accuracy, noise resilience, and real-time suitability for SSO monitoring in smart grids.
Funding Information
  • Open Access funding provided by Qatar National Library

This publication has 44 references indexed in Scilit: