Generalized Predictive Control for $LC$-Filtered Voltage-Source Inverters With Enhanced Predictive Horizon

Generalized predictive control (GPC) for power converter applications provides the benefits of constant switching frequency, explicit stability check, and high control bandwidth. However, featuring the second-order characteristic in the
$LC$
-filtered voltage source inverter, the assessment and investigation of GPC as the inner control loop have not been fully fulfilled, especially under the grid-connected mode. Besides, the inherent ability of GPC enables the prediction of future system behavior within longer horizons, which, however, has been rarely investigated and, thus, it is still unclear about the closed-loop property brought by the enhanced predictive horizons. To this end, this work first clarifies the contribution of adopting GPC by comparing it to the capacitor-voltage control+inductor-current feedback active damping with optimal damping coefficient design, where the virtual synchronous generator is used to implement the case study. Through the simulation and experimental results, the GPC expresses a smoother and faster dynamic process and enhanced transient frequency nadir compared to the existing multiloop strategy. The robustness and stability are analyzed through the closed-loop eigenvalue, and it is also revealed that the GPC achieves stronger robustness against the variation of the control and plant parameters simultaneously by increasing the prediction horizon length.