Speed-Sensorless Dual Reference Frame Predictive Torque Control for Induction Machines

In finite control set predictive torque control (FCS-PTC), mismatched parameters will lead to the electromagnetic torque prediction error. To solve the aforementioned problem, this paper proposes feedback-correction-based dual reference frame predictive torque control for induction machines, abbreviated as DRF-PTC. First, a DRF flux observer is proposed, which abandons a traditional model reference adaptive observer with rotor speed as adaptive variable. The proposed flux observer decouples the flux observation and the speed estimation and improves the estimation accuracy of flux linkage and rotor speed. Second, a DRF-PTC is proposed to eliminate the effect of speed estimation error on torque prediction. For the influence of mismatched parameters, a correction term is incorporated in double reference frame torque prediction equations, in which the correction term is derived from the correction term of the proposed flux observer. Finally, the proposed DRF-PTC in this paper is verified on the experimental platform of an induction machine. Experimental results show that the proposed DRF-PTC has excellent performance in speed estimation and electromagnetic torque prediction, compared with the traditional FCS-PTC.