Direct torque control of induction machines based on predictive control

To reduce the electromagnetic torque ripple of induction machines under conventional direct torque control, a improved direct torque control based on predictive algorithm is proposed. The control strategy predicts the change of electromagnetic torque and stator flux magnitude for each output voltage vector of inverter in several steps using a discrete-time model of an induction machine. An objective function consisted of predictions of electromagnetic torque, stator flux magnitude and its dynamic errors is proposed as an optimal control rule. The stability of the system was analyzed through the Pole-Zero map of closed-loop transfer function. Based on that, the optimal solution is obtained through minimizing the objective function. A simple algorithm to obtain the optimal solution is proposed for the voltage vector of inverter being finite. The new control strategy possess fast dynamic response and reduce the electromagnetic torque ripple effectively. Simulation study verify the feasibility of the proposed method.