Filtered Repetitive MPC Applied to DFIG Grid Side Converter Frequency Support in Microgrids During Island Operation

As the trend towards the decentralization of grids continues, new typologies such as microgrids have gained popularity in recent years. Microgrids offer greater reliability as they can operate both connected to the main grid and in islanded mode. However, the transition to island mode can be even more challenging when a microgrid has a high penetration of intermittent sources like Doubly-Fed Induction Generators (DFIGs). In this context, designing new controller algorithms is essential to improve frequency support in microgrids with DFIGs. Therefore, this paper proposes a novel Filtered Repetitive Model Predictive Control (FR-MPC) applied to the frequency support in a microgrid. The control structure is applied to the Grid Side Converter (GSC) of a DFIG, while the DC bus voltage is supplied by a Battery Energy Storage System (BESS). The FR-MPC prediction model uses a state-space model with an embedded repetitive action, which has the capacity to mitigate harmonics of the microgrid. However, the inclusion of the repetitive loop degrades the closed-loop robustness. To overcome this limitation, the prediction model features a robustness low pass filter. Furthermore, this paper presents an analysis of the frequency response and pole map of the key components of the controller: the robustness filter, the repetitive loop, and the control law. Finally, simulation results demonstrate that the proposed controller outperforms both the pure Repetitive MPC (R-MPC) and the Generalized Predictive Control (GPC) in terms of performance.