Power coupling modeling and analysis method of VSG control-based system for quantifying dynamic interaction

• The established [P Q]-[ω V] model, taking power, frequency and voltage amplitude as variables, is able to directly characterize the power coupling feature, and is easy to be extended to multi-converter systems. • A quantitative analysis method is proposed based on the relative gain matrix theory to quantify the degree of power coupling and the interaction among multiple VSGs. • This paper investigates the influence of the key control parameters and grid impedance on the power coupling characteristics of VSG, and the mechanism of the interaction effect among multiple VSG units. • A modified decentralized damping control method is proposed that additional active power negative feedback and reactive power positive feedback terms is introduced in the frequency and voltage control loops, respectively. It provide more control degree of freedom, and help power decoupling and suppress oscillation. In multi-VSG system, there are coupling between the power control loops and among the multiple VSGs, which may deteriorate the control performance and even pose risks to the system stable operation. However, there have been few studies on quantitative index and analysis of power coupling and the dynamic coupling among multiple VSGs. Thus, this paper proposes power coupling modeling and analysis method of VSG control-based system for quantifying dynamic coupling. The established [P Q]-[ω V] model, taking power, frequency and voltage amplitude as variables, is more convenient to characterize power coupling feature. Besides, a quantitative analysis based on relative gain matrix theory is presented to quantify the degree of power coupling and the coupling among multiple VSGs. Further, a modified decentralized damping control method is proposed to help power decoupling and suppress oscillation, which provide more control degree of freedom. Finally, the correctness of the proposed modeling and quantitative analysis method is validated by the control-hardware-in-loop experiment results.