A Method for Identifying Key Parameters of Frequency Modulation in GFM Converters Based on LS-GA

Abstract

With the high proportion of renewable energy connected to the grid through power electronic interfaces, the frequency support of the system gradually shifts from being dominated by traditional synchronous machine inertia to being dominated by control strategies. Its dynamic inertia exhibits obvious scene dependence and time-varying characteristics, which brings higher difficulty to the identification of key frequency regulation parameters. Especially when the grid-forming (GFM) converter participates in frequency regulation, the superposition of different control links will result in strong coupling and segmented changes in power frequency response. Traditional equivalent inertia evaluation methods based on fixed model parameters are difficult to directly apply. To address this issue, this paper develops a parameter identification method that combines least squares (LS) fitting with a genetic algorithm (GA). Firstly, the influence of key frequency-regulating parameters on various frequency dynamic response indicators is analyzed. Then, a collaborative estimation method is proposed, utilizing the LS method and GA for sliding-window fitting to estimate the inertia and damping of GFM converters. Simultaneously, considering the challenge of verifying the validity of estimation results, a dual-model comparative verification method based on the P-ω admittance method is presented. The proposed parameter identification method is applicable to various GFM converter carriers. Finally, based on the proposed method, key frequency-regulating parameters of a receiving-end GFM converter in an example system connecting photovoltaics via VSC-HVDC are estimated and verified under different operating conditions. This enables intuitive verification of estimation result accuracy, providing an essential foundation for overall system estimation and security and stability analysis.