The integration of new-energy power electronic devices into the grid is prone to induce subsynchronous/supersynchronous oscillations, posing a threat to grid stability. Conventional oscillation identification methods are often affected by mode mixing and noise interference during parameter extraction, which compromises analysis accuracy. To address these issues, this paper proposes a variational mode decomposition (VMD) method based on a trigonometric topology aggregation optimization (TTAO) algorithm. The TTAO algorithm adaptively optimizes the key VMD parameters—the number of modes K and the penalty factor α—thereby improving the accuracy and robustness of signal decomposition. Furthermore, the total least squares-estimating signal parameter via rotational invariance techniques (TLS-ESPRIT) is employed to denoise and reconstruct the decomposed signals, enabling accurate identification of oscillation components. The proposed method is validated through composite signal tests, ADPSS simulations, and field measurements from a renewable energy plant, with comparisons made against the CEEMD algorithm and an improved Prony algorithm under various noise conditions. The results demonstrate that the proposed approach can effectively separate and accurately extract the coupled subsynchronous/supersynchronous oscillation modal parameters, while significantly suppressing noise interference. The identification accuracy reaches the order of 10−4, and the error remains below 2% even under complex noise scenarios. The method exhibits good engineering applicability and provides a new perspective for tracing, monitoring, and suppressing oscillations in wind power systems.
Li et al. (Thu,) studied this question.