High-Precision Antagonistic Hysteresis Modeling and Compensation for a Nanopositioning Stage With Antagonistic Piezo-Actuation

This article proposes a novel antagonistic Prandtl-Ishlinskii (API) model to accurately describe and compensate the complex hysteresis nonlinearity inherent in antagonistically actuated nanopositioning stages using piezo-actuation. Unlike conventional PI-based modeling strategies that independently treat each actuator, the proposed API model performs nonlinear modeling on the overall coupled output displacement of system. This unified modeling approach achieves higher accuracy with fewer parameters and supports both forward and inverse modeling in a compact structure. To further improve the system’s robustness against unmodeled dynamics and disturbances, an adaptive Kalman filter (AKF) is integrated for real-time parameter identification. Experimental validations confirm that the API model outperforms counterparts in hysteresis modeling with a minimum root mean square error (RMSE) of 129.0 nm. For hysteresis compensation in trajectory tracking, the inverse API model achieves a minimum RMSE of 213.8 nm. The AKF+API−1method enables accurate nonperiodic trajectory tracking. Even under disturbances, AKF+API−1method converges quickly within 200 ms, verifying its high-precision trajectory tracking performance.