A universal multiple-vectors-based model predictive direct power control for doubly fed induction generators

Abstract

Model predictive direct power control (MPDPC) has been widely concerned in the control of doubly fed induction generator (DFIG) systems owing to its simple concept and good multivariate control ability. The optimal voltage vector is selected based on the principle of power error minimization. Nevertheless, the conventional MPDPC still presents relatively high power ripples and the switching frequency is variable. To cope with the problems above, recently multiple-vector-based MPDPC is proposed to improve the steady state performance, however, at the cost of increased control complexity. This paper proposes a universal multiple-vectors-based MPDPC. It firstly obtains the desired rotor voltage vector reference based on the principle of deadbeat power control, and then calculates the three-phase duties using carrier-based pulse width modulation (PWM) with injection of zero sequence component. The vector selection and duty cycle calculation in the proposed MPDPC are obtained according to the three-phase duties, which is much simpler than conventional multiple-vectors-based MPDPC. Furthermore, this method reveals the inherent relationship between multiple-vectors-based MPDPC and deadbeat control. The effectiveness of the proposed method is validated by the presented simulation results from a 15 kW DFIG system.