Computationally Efficient Deadbeat Direct Torque Control Considering Speed Dynamics for a Surface-Mounted PMSM Drive

In this article, a computationally efficient deadbeat (DB) direct torque and flux control is investigated for surface-mounted permanent magnet synchronous motor (SPMSM) drives. Unlike conventional DB direct torque control (DB-DTC), the proposed DB-DTC technique simultaneously manipulates the electromagnetic torque and stator flux along with rotor speed in a combined DB controller structure. Moreover, a simple and computationally efficient DB-DTC structure is achieved through a novel DB controller design in the stationary reference frame and it ensures an improved transient performance within finite time steps. The feedforward terms are properly designed to mitigate the effects of system uncertainties. The stability of the proposed DB-DTC has been proven and discussed in detail through Lyapunov theory and eigenvalue analysis. The designed DB-DTC strategy has been simulated via MATLAB/Simulink software and practically evaluated on a laboratory SPMSM drive with TI digital signal processor TMS320F28335. Extensive comparative evaluation with conventional proportionalintegral (PI)-DTC and DB-DTC corroborate an improved control performance in speed/torque dynamic response (fast rise time) as well as reductions in torque and flux ripples under tough practical conditions (e.g., speed and torque step-changes, and speed reversal test cases) with significantly reduced computational complexity.

Automated summary

This article investigates a computationally efficient deadbeat direct torque and flux control technique for surface-mounted permanent magnet synchronous motor drives. The proposed control strategy manipulates electromagnetic torque and stator flux while ensuring improved transient performance and stability within finite time steps. Extensive simulations and practical evaluations demonstrate enhanced control performance and reductions in torque and flux ripples under tough practical conditions with reduced computational complexity.