Antiwindup Composite Adaptive Speed Control for Permanent-Magnet Synchronous Motors

This paper presents an anti-windup composite adaptive speed controller (AW-CASC) for permanent-magnet synchronous motors (PMSMs) operating under parameter uncertainties, external load torque, and input saturation. The proposed AW-CASC integrates three components, an adaptive compensation control (ACC) term, an adaptive proportional-integral (API) controller, and an anti-windup (AW) mechanism, and is therefore referred to as AW[API-ACC]. The ACC compensates for unknown disturbances using an adaptive law that relies only on measurable feedback signals, ensuring robust operation of the composite controller without requiring disturbance observers. The API employs a nonlinear adaptive gain algorithm based on exponential and fractional functions of the velocity error, enabling fast transient response and improved tracking accuracy without the need for fuzzy logic or gain scheduling. To suppress overshoot caused by integrator windup under large step-speed commands, a lightweight AW strategy is integrated directly into the Lyapunov-based control framework. Stability analysis confirms that all closed-loop signals are uniformly ultimately bounded. The proposed AW[API-ACC] method is implemented and experimentally validated under various speed profiles, incorporating parameter uncertainties, external load torque, and current saturation conditions. Compared with a conventional PI controller with disturbance observer (PI-OB), an adaptive optimal control (AOC) strategy, and adaptive sliding mode control combined with an extended state observer (ASMC-ESO), the AW[API-ACC] method achieves up to a 61% reduction in RMS speed error (sinusoidal), 33% and 31% reductions in overshoot and maximum tracking error (ramp), a 70% faster recovery under load torque disturbances, and an 8.8% improvement in settling time (step). These results highlight the improved robustness and tracking performance of the proposed method.