PREDICTION ALGORITHM FOR TORQUE RIPPLE REDUCTION IN DTC-BASED DRIVES

In direct torque control, the selection of an applied voltage vector is based on the measured parameters at the beginning of the sampling period. The delay between the measurement and the application of the voltage vector is the origin of an extra torque ripple. In this paper, a new predictive controller is proposed for compensating this delay and reducing the torque ripple. The stator current is measured twice in each sampling period and its expected value at the end of the period is predicted according to a linear extrapolation algorithm. Therefore, none of the machine parameters are used in the prediction, and consequently, the algorithm is quite robust against machine parameter variations. The calculation of the electromagnetic torque is performed using the predicted value of the stator current. Therefore, the selection of the voltage vector is more realistic and prevents extra torque ripple. This controller is quite suitable for high power drives where the sampling frequency is low, and there is enough time for the extra measurements. Simulation and experimental results which confirm the ability of this method to considerably reduce the torque ripple are presented.