A CURRENT-TORQUE DUAL-OBJECTIVE MODEL PREDICTIVE CONTROL STRATEGY FOR ENHANCED PMSM DRIVE PERFORMANCE

Abstract

In this study, a current-torque dual-objective model predictive control (CTDO-MPC) method is proposed to improve steady-state and transient performance. Unlike conventional CO-MPC, which mainly minimizes stator current errors, the proposed method integrates current-tracking error and electromagnetic torque deviation into a unified objective function. To balance the current and torque terms' sensitivity within the same objective function, a weighting factor is analytically determined under equal voltage variation conditions. All feasible inverter voltage vectors are evaluated at each sampling instant, and the vector that minimizes the objective function is applied in the next control interval. The controller is validated in MATLAB/Simulink with a 1.5 kW PMSM drive and compared with CO-MPC. Simulation results demonstrate that the proposed strategy reduces current total harmonic distortion from 5.1% to 3.6% at half-load and 2.5% to 1.8% at full-load. In addition, the RMS torque ripple is reduced from 0.31 Nm to 0.24 Nm at half-load and 0.29 Nm to 0.22 Nm at full-load. These results indicate improved current quality, reduced torque ripple, and enhanced PMSM drive performance.