FPGA-in-the-Loop implementation of enhanced Lyapunov-based modulated predictive control approach for LCL-filtered grid-connected inverters

Finite control set model predictive control (FCS-MPC) is a technique that has the ability to optimize multiple constraints and variables with a superior performance than linear control strategies. However, it is limited by a high computational burden, variable switching frequency, and the lack of a formal stability proof. This work offers a comparative study of four control methods, two conventional and two proposed, for an LCL-filtered grid-tied inverter. The investigated predictive control techniques use two-step prediction strategy on the discrete model of the power system to compensate for computational delays introduced by online optimization. The proposed methods ensure the stability of the controlled system, since they are applying Lyapunov stability criterion on the objective function defined to optimize the control problem. Moreover, one of the proposed control techniques integrates a modulation stage to reach a fixed switching frequency and low steady-state error. By combining these enhancements, the proposed strategies effectively address the limitations of classical predictive control methods, offering a simplified control structure and significantly reduced computation time. A detailed comparison of the control methods is developed through an FPGA-in-the-Loop (FiL) co-simulation implementation using MATLAB/Simulink® and Xilinx ISE under System Generator. The evaluation of the control techniques is validated using real controller Virtex 6 FPGA kit in real-time for multiple cases such as steady-state, dynamic response, weak grid conditions, and parameter mismatch. The co-simulation results demonstrate that the suggested control strategies offer improved steady-state performance, enhanced robustness against the stiff grid conditions and parameters variations, and notable reduction in the computational effort.