Grey Box Modeling Method of Grid-Connected Inverters Based on Parameter Sensitivity

As the core device of renewable energy systems, the inverter's accurate modeling is the basis for optimizing control strategies and analyzing system performance. Due to the manufacturers' commercial confidentiality, available inverters are typically grey box models with known main circuit structures and unknown control parameters, which hinders the accurate modeling of the inverter. To address this issue, a grey box modeling method of grid-connected inverters based on parameter sensitivity is proposed. First, the inverter mechanism model is analyzed. Then, the AC-side impedance characteristics and parameter sensitivity of the inverter model are analyzed to determine identification frequency and impedance elements. On this basis, the identification scheme is designed, where the PSO algorithm is used to identify parameters. Finally, case studies show that the integral coefficient at 30Hz and proportional coefficient at 80Hz exhibit average errors of 4.38% and 1.78%, respectively. The degree of fit between the external characteristics waveform of the grey box model and the actual waveform, as determined by the final identification results, exceeds 99%. The proposed grey box modeling method is efficient in identifying control parameters, demonstrates good adaptability, and achieves high accuracy under complex conditions.