Model Predictive Direct Torque Control of permanent magnet synchronous motors

Abstract

Model Predictive Direct Torque Control (MPDTC) is a recent computational control methodology that combines the merits of Model Predictive Control (MPC) with the ones of Direct Torque Control (DTC). Specifically, with respect to standard DTC, the converter's switching frequency and/or losses are considerably reduced, while at the same time the Total Harmonic Distortion (THD) levels of the phase currents and the torque are improved. Moreover, DTC's favorable dynamic and robustness properties are preserved. This paper presents an MPDTC scheme for a permanent magnet synchronous motor that achieves long prediction horizons in the range of up to 150 time-steps through the use of extrapolation and bounds. A discrete-time internal controller model of the drive system is derived from the physical equations. Simulation results for a three-level voltage source inverter indicate that such an MPDTC scheme, compared to an industry standard controller, is capable of reducing the switching losses and the switching frequency by up to 50%, and the torque THD by 25%, while leaving the current THD unchanged.