Efficient Direct Model Predictive Control for Doubly-Fed Induction Generators

In reality, heavy calculation burden is an obstacle for applying direct model predictive control (DMPC) schemes in industrial applications. To resolve this problem, this paper proposes a computationally efficient DMPC scheme for doubly-fed induction generators (DFIGs) based on variable-speed wind turbine systems (WTSs). The procedure of selecting the optimal switching vector is enhanced by calculating the reference voltage vector (VV) directly from the reference current. Then, the sector where this reference VV is located is identified. Finally, the cost function is evaluated only for three times to get the best switching action. Consequently, the necessity to test all candidates VVs will be avoided, which reduces the calculation burden of the conventional DMPC method. Control performance of the proposed efficient DMPC method is compared with the conventional DMPC by simulation results for all operation conditions. Furthermore, the performance of the proposed efficient DMPC is investigated under variations of the DFIG parameters and in case of unbalanced grid conditions.