Modal parameter identification of wide-band oscillations at the grid connection point in photovoltaic power stations based on improved VMD-TKEO

Aimed at the difficulty in identifying wide-band oscillation modal parameters at the point of common coupling (PCC) in photovoltaic power stations, this paper proposes a parameter identification method integrating improved Variational Mode Decomposition (VMD) and the Teager-Kaiser Energy Operator (TKEO). The method first utilizes the Beluga Whale Optimization (BWO) algorithm to adaptively optimize the critical VMD parameters (the number of modes K and the penalty factor α) with the minimum envelope entropy as the objective. Then, the signal is decomposed using the optimized BWO-VMD. Effective Intrinsic Mode Function (IMF) components are screened based on correlation coefficients and undergo secondary denoising via wavelet thresholding to enhance the signal-to-noise ratio. Finally, the TKEO is introduced to extract instantaneous features from the denoised IMF components, simultaneously obtaining the amplitude, frequency, and damping ratio of the oscillation modes. Compared to the EMD-TKEO and VMD-HHT methods, the proposed approach demonstrates significant advantages in modal resolution capability, noise immunity, and identification accuracy. The frequency identification error is below 0.04%, and the damping ratio identification error is less than 5%. It effectively suppresses modal confusion and enhances the accuracy and reliability of parameter identification.