McMaster University
Model Predictive Control for Dual Three-phase Interior Permanent Magnet Synchronous Motor Drives
Abstract
dc:description.abstractThe increasing demand for electrified transportation requires high-power, efficient, and reliable traction drive systems. Dual three-phase (DTP) interior permanent magnet synchronous machines (IPMSMs) offer important advantages for these applications, but also introduce challenges in control accuracy and current harmonic suppression. This thesis develops a comprehensive technical path that incorporates two improvements to predictive current control for DTP-PMSM drives. First, a finite-control-set model predictive control (FCS-MPC) strategy with an enhanced sector selection method is proposed for high-saliency DTP-PMSM drives. The proposed method improves tracking accuracy in the d-q subspace by reliably identifying the optimal virtual vectors, while a quasi-proportional-resonant (QPR) controller suppresses current harmonics in the x-y subspace. Simulation and experimental results show that the proposed method achieves fast dynamic response, reduced current ripple, and effective harmonic suppression, outperforming existing virtual-vector-based FCS-MPC and field-oriented proportional-integral (PI) control methods. Second, this thesis addresses the limitations of conventional ultra-local model-free predictive current control. Although such methods improve robustness to parameter mismatches, their observers must reconstruct lumped dynamics containing fast current-dependent components, leading to a trade-off between dynamic response and noise amplification. To overcome this limitation, a physics-informed model (PIM) is proposed for model-free predictive current control of DTP-PMSM drives. The PIM reformulates the current dynamics as slowly varying unknown coefficients combined with fast known state variables, providing a structurally parameter-insensitive formulation for observer-based model-free control. The proposed PIM is implemented in a model-free deadbeat (MFDB) predictive current controller with an extended state observer (ESO) and evaluated through simulation and experiments on a 9-kW DTP-IPMSM drive. The results demonstrate fast current and torque responses under low observer bandwidth, improved robustness to parameter mismatch, and reduced noise amplification and current ripple compared with conventional ultra-local model-free and model-based deadbeat control methods.
Degree
thesis:*- Department dc:contributor.department
- Electrical and Computer Engineering
- Year dc:date.issued
- 2026
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Yang, Jingru
- Advisor dc:contributor.advisor
-
- Nahid-Mobarakeh, Babak
Rights
- Language dc:language.iso
- en
Identifiers
dc:identifier.*- Identifier URI
- https://doi.org/10.71548/25663