Three-vector Model Predictive Control Method of DFIG with Inductance Parameter Identification

In the optimal duty cycle model predictive control(MPC) strategy of doubly-fed induction generator(DFIG), the voltage vector direction is fixed and the current has large ripples, and control performance depends on the accuracy of parameters. To address these issues, a three-vector-based MPC with online inductance parameter identification is proposed. Firstly, the expected voltage vector is equivalently synthesized by three basic voltage vectors in each sector, and the vector duration is calculated by the deadbeat control of the quadrature and direct axis current components. Then, the synthesized expected voltage vectors are substituted into the value function, and the optimal voltage vector is output by rolling optimization. Besides, to improve the parameter robustness of the system, the recursive least squares(RLS) method with forgetting factor is used to identify the inductance online, and the results are used in the calculation of control. Finally, experimental results show that the proposed method improves the dynamic and steady state performance and robustness of the parameters compared with the optimal duty cycle MPC.