A high-precision control method for nonlinear characteristics of piezoelectric actuators

This paper presents a high-precision control method for piezoelectric actuators (PEA) to improve displacement tracking accuracy in ultra-precision positioning applications. A discrete Duhem hysteresis model is established to describe the nonlinear input–output behavior of the actuator, and a modelling framework combining wavelet threshold denoising (WTD) with a recursive least-squares algorithm using a time-varying forgetting factor (RLS-TFF) is developed to improve parameter identification accuracy under measurement noise. Based on the identified model, a feedforward-feedback control scheme is constructed for hysteresis compensation and tracking enhancement. Experimental results show that the proposed modelling method achieves a root mean square error (RMSE) of 0.0446 μm in hysteresis identification. In addition, the proposed control scheme reduces the hysteresis error from 16.4% FSR to 0.02% FSR and improves tracking performance under both static and dynamic reference trajectories. These results demonstrate that the proposed method is effective for high-precision motion control of PEAs in applications requiring nanoscale positioning accuracy.