A Modified Model-Free Predictive Current Control Method Based on an Extended Finite Control Set for DFIGs Applied to a Nonideal Grid

Abstract

Conventional model-free predictive current control (MFPCC) for doubly fed induction generator (DFIG) systems presents high steady-power ripples and current harmonics, because it uses only eight basic voltage vectors of a two-level converter. Furthermore, the stator and rotor currents of DFIGs are severely distorted under a nonideal grid if no special measures are taken. In this article, a modified MFPCC (MMFPCC) is proposed to address the problems above. On the one hand, an extended finite control set, including 20- rather than 8-voltage vectors as in the conventional finite control set, is used to improve the steady-state performance. On the other hand, the stator current reference is calculated so that sinusoidal and balanced stator currents are obtained even under a nonideal grid. The proposed MMFPCC is compared to conventional MPC and MFPCC, and the experimental results confirm its effectiveness.