Performance of Robust Model Predictive Control in Grid-Connected Three-Phase Inverters with Unknown Inductive Filter Parameters

This paper investigates the performance of a robust Model Predictive Control (RMPC) strategy applied to a three-phase inverter connected to the grid through an inductive filter with unknown but bounded resistance and inductance parameters. Using a Simulink/Matlab simulation environment and the Mosek/Yalmip solver, the RMPC’s performance was evaluated in both transient and steady-state conditions. Various filter configurations were tested to assess the controller’s robustness under different parameter values. The RMPC optimization problem, incorporating Linear Matrix Inequalities (LMI) constraints, was solved to minimize a cost function upper bound, ensuring stable and efficient current control. The results demonstrate that the RMPC effectively manages the current across all tested filter configurations, thus validating the controller’s robustness and effectiveness in power electronics applications with uncertain parameters.