Model Predictive Torque Control with Synchronized Sampling Frequency for High Frequency Induction Machine Drives

Model Predictive Control (MPC) is a promising technique for control of electric drives, as it enables to optimize multiple parameters and offers operation with non-linear systems. Implementation of MPC for high speed/high-pole drives with high rated fundamental frequency - from few hundred up to thousand Hz - poses many challanges due to the low G sampling to fundamental frequency ratio around the rated speed of the machine. The paper proposes a novel MPC scheme, where the sampling frequency is sychronized to the fundamental frequency to improve the performance. The algorithm can be implemented on a digital device by the recursive equations given in the paper. As it is shown, the method can reduce the number of switchings, achieve better harmonic performance and reduce the torque ripple at the same sampling frequency without synchronization at low G values.