Robust Predictive Repetitive Current Control for a Grid-Connected Inverter under Parametric Uncertainty

Model predictive control (MPC) is a well-established approach for applications using grid-connected inverters. However, because the MPC is a model-based method, its performance may degrade due to parametric errors and large perturbations in grid voltage or load current. Predictive repetitive control (PRC) combines the conventional MPC with repetitive control (RC) and presents an improved potential for dealing with parametric uncertainties in this application context that is not fully explored in the technical literature. This article proposes a robust PRC current control in the synchronous rotation frame for a grid-connected inverter under parametric uncertainty. The concept of robust stability (RS) margin, derived from the singular value decomposition (SVD), was used to examine how the PRC controller tuning affects the robustness under parametric uncertainty and grid voltage disturbance. Experimental results on a 1-kW workbench validated the robustness of the proposed PRC controller, maintaining satisfactory performance under nominal and severe parametric uncertainties, satisfying the IEEE Std. 1547.2-2008. In addition, it was demonstrated that the proposed PRC provides greater robustness to parametric uncertainties than the conventional MPC and the classical proportional-integral (PI) controllers.