Performance enhancement of direct torque-controlled permanent magnet synchronous motor drives for electrified-transportation applications

Abstract

The high-power density, low maintenance expense, and high efficiency of the PMSM drives have contributed to their widespread use in electrified transportation systems. Direct torque control (DTC) is widely adopted in PMSM drives to meet the torque control requirements of these systems. It uses two hysteresis regulators and a switching table to directly manipulate the switching states of the power converter, providing a fast dynamic response, simple structure, and high robustness. However, it is hard to provide adequate steady-state torque performance under different operating speeds because the hysteresis regulator cannot recognize the torque error magnitude, and the switching table cannot ensure the optimal switching transition.
Therefore, this dissertation has investigated the impact of the converter voltage vectors on the torque and flux performances under different operating conditions to identify the best option at each state in the switching table. Based on this investigation, the torque performance improvement is guaranteed through three main stages. First, a flexible switching table (FST)-based DTC strategy has been proposed, the structure of which changes according to the operating condition to minimize the torque ripple without the need for the hysteresis regulator. Then, a multiobjective duty-ratio regulator is employed to achieve further torque ripple reduction and eliminate the steady-state torque error based on a virtual-reference signal. This improved-torque performance has not been provided only by applying double voltage vectors at each control interval but also by giving the highest priority to the voltage vectors that provide the lowest torque deviation during steady-state operation. Finally, a model predictive control (MPC)-based solution with low complexity is presented to guarantee the optimal voltage vector selection. In this solution, the unnecessary predictions are avoided, reducing the computational efforts, and the control objectives are considered separately, eliminating the need for a weighting factor. The feasibility and effectiveness of the proposed DTC strategies have been verified through comparative evaluations with existing schemes in the literature using a finite-element (FE)-based simulation system and an experimental setup of PMSM drive.

Date of Award	Mar 2023
Original language	English
Awarding Institution	University of Nottingham
Supervisor	Chunyang Gu (Supervisor), Serhiy Bozhko (Supervisor) & Chris Gerada (Supervisor)