Finite Control Set Model Predictive Control for LCL-Filtered Grid-Tied Inverter with Minimum Sensors

Recently, finite control set model predictive control (FCS-MPC) has been successfully applied in the grid-tied inverter with the LCL filter. However, to achieve active damping and grid synchronization, many sensors are required, increasing cost, and complexity. In addition, a considerable computational delay should be addressed when it is experimentally implemented, which may degrade the performance of the overall system. In order to reduce the number of sensors, eliminate the computational delay, and enhance the control reliability of the system, a novel FCS-MPC strategy with merely grid-injected current sensors is proposed, which contains four compositions: virtual flux observer, state observer, delay compensation, and FCS-MPC algorithm based on estimations. A 3-kW/3-phase/110-V experimental platform is established to validate that utilizing the proposed observations-based control method with only grid-injected current sensors is capable to obtain satisfactory performance of grid synchronization and high-quality grid-injected current both under balanced and unbalanced grid voltage condition.