This study investigates model predictive speed control for permanent magnet synchronous motors (PMSMs) under repetitive disturbances. A novel single-loop dual-layer MPC framework incorporating an iterative learning observer (ILO) is proposed. First, a discrete-time ILO is designed to estimate repetitive disturbances. By fully utilizing the repetitive characteristic, the ILO enables accurate estimation of harmonic disturbances while simultaneously providing a preview of their future behaviors. Second, the state reference trajectory and the control reference trajectory of the PMSM are derived by incorporating both the reference sequence and the previewed disturbance sequence. Finally, a single-loop MPC is developed to track the reference trajectories. Departing from existing MPC schemes, the proposed framework demonstrates three distinctive advantages: (1) exact tracking of references under repetitive disturbances, (2) preservation of structural simplicity, and (3) no requirement of explicit exogenous signal modeling. The stability of the proposed controller is rigorously analyzed. Experimental results demonstrate that the proposed method outperforms existing disturbance-observer-based MPC designs in steady-state speed fluctuation suppression capability.

Permanent magnet synchronous motor (PMSM),Model predictive control (MPC),harmonic disturbances,iterative learning observer (ILO)