Model Predictive Control Using the Singular Perturbation Theory for Permanent-Magnet Synchronous Machines

This study presents a permanent-magnet synchronous machine (PMSM) model using the singular perturbation theory to obtain the reduced-order model. The fast and slow subsystems of the PMSM drive system are decoupled. The electrical machine's two-time scale property is fully utilized for model predictive control (MPC). An enhanced MPC control strategy is designed to provide dead-beat speed control and improved predictive current control for the external and internal loops. The singular perturbation theory of the PMSM is investigated and assigned to boundary-layer and quasi-steady-state models. The proposed algorithm, built on a dual-core DSP, achieves quick transient dynamics and steady performance. The experimental results confirmed that the proposed method is robust to varying electrical machine parameters and disturbance torque uncertainty.