Research on APF control strategy based on improved recursive least squares parameter on-line identification

With the large-scale integration of new energy sources, the issue of harmonics has become increasingly prominent. Active power filters, renowned for their excellent harmonic suppression capabilities, are widely utilized for harmonic control. However, in harsh operating environments, the parameters of these filters often suffer from mismatches, which can degrade their performance. To address the decline in harmonic control efficiency caused by parameter mismatch and related control errors, this paper proposes an online parameter identification method based on an adaptive forgetting factor recursive least squares (RLS) algorithm. By dynamically adjusting the parameters of the control system, the method enhances the precision of harmonic suppression even under parameter mismatch conditions. Furthermore, the adaptive adjustment of the forgetting factor balances the trade-off between identification accuracy and convergence speed, effectively preventing parameter saturation.