Model-Free Predictive Control Based on RLS Algorithm for Grid-Forming Inverters With Virtual Synchronous Generator

As a key element to connect renewable energy sources, grid-forming inverters (GFIs) mostly use droop control as an outer loop, but it can cause a swift change in the system frequency when disturbances arise. In the role of the inner loop control, the traditional model predictive control (MPC) is widely employed in converters, but it is heavily dependent on model parameters. To overcome the above drawbacks, a model-free predictive control with virtual synchronous generator (VSG-MFPC) is proposed in this study. The scheme uses an autoregressive MFPC based on an auto-regressive with exogenous input (ARX) model as an inner-loop control structure for voltage prediction, combined with a simplified virtual synchronous generator (VSG) as an outer-loop control structure for power sharing and inertial support control systems. The implementation of MFPC improves the parameter robustness of the system. Building on this, the improvement schemes of algorithmic parameter initialization, delay compensation and current limiting capability are further investigated. Additionally, incorporating the VSG into the outer loop strengthens the stability of the system’s frequency and voltage while also enhancing the system’s dynamic characteristics. The aforementioned advantages, along with the efficacy and feasibility of the proposed control method, are validated through simulation and experimental results.