Hysteresis severely reduces positioning performance of a piezoelectric nanopositioning stage. Linear active disturbance rejection control (LADRC) is a practical solution to improve the positioning accuracy. However, the PD controller utilized in the LADRC is not effective enough to suppress the uncancelled total disturbance, and high-order pure integrators are difficult to be stabilized just by a PD controller. In this work, a robust U-model active disturbance rejection control (RUADRC) is proposed by incorporating the core idea of the U-model control and the Glover-McFarlane control. Then, the controlled plant can be dynamically transformed to a unit. Difficulties in stabilizing high-order pure integrators are decreased, the phase lag between the input and output of a controlled plant is reduced, and the closed-loop responses is sped up. In addition, the influence of both inaccurate total disturbance estimation and imperfect approximation is also minimized by the Glover-McFarlane control Closed-loop stability, steady-state tracking error, and the phase advantage of the RUADRC have been analysed. Theoretical results show that the RUADRC promises a timelier and more accurate positioning. Experimental results still confirm the advantages of the RUADRC over the LADRC on both reference tracking speed, accuracy and disturbance rejection ability.
Keywords: Hysteresis; LADRC; Nanopositioning; Piezoelectric actuator; U-model control.
Copyright © 2021 ISA. Published by Elsevier Ltd. All rights reserved.