A computationally efficient FS-PTC for IM with minimum voltage vectors

In finite-state predictive torque control (FS-PTC) scheme, torque and stator flux of induction motor (IM) are predicted for all admissible voltage vectors of a voltage source inverter (VSI). Then the predicted torque and flux are evaluated by a predefined cost function. The voltage vector which produces minimum torque and flux ripples is selected for the next control period. Most of the time of the control period is spent on prediction and actuation, and longer execution time of a control algorithm limits the sampling frequency. This is one of the main causes of degradation in torque and flux performance in FS-PTC. The prediction and actuation time is greatly increased with the number of admissible switching states of the inverter. This paper proposes reduced number of switching states for prediction and actuation using stator flux position and sign of the torque deviation between reference torque and actual torque. Experimental results show that the performances of the proposed and the conventional control schemes are similar, while the execution time of the proposed control algorithm is reduced.