Model predictive direct torque control of permanent magnet synchronous motors with extended set of voltage space vectors

Abstract

The finite-control-set model predictive direct torque control (FCS-MPDTC) is a novel control scheme for permanent magnet synchronous motors (PMSMs). A key feature of FCS-MPDTC is that the eight possible voltage space vectors or switching combinations of the power converters are directly taken into account as the control input of the system. A cost function is used in FCS-MPDTC to evaluate each possible voltage space vector and the one with minimum cost is applied to the power converter. Due to the considerable torque and flux ripples, to improve the performance of FCS-MPDTC, this study presents an extended set of 20 modulated voltage space vectors with fixed duty ratio. For further improvement, a larger set size can be chosen, but this requires a larger computing power. To mitigate the computational burden caused by increased number of voltage space vectors, a pre-selective scheme is designed for the proposed FCS-MPDTC to filter out the impractical voltage vectors instead of evaluating all 20 voltage space vectors. The drive system efficiencies of conventional direct torque control, conventional FCS-MPDTC and proposed method are investigated. The theory and simulation are validated by experimental results on a PMSM prototype.