A High-Gain Loop-Shaping Method for Precision Motion Control
Yixuan Meng,Linlin Li,Xiangyuan Wang,Minyu Pan,Limin Zhu
DOI: https://doi.org/10.1109/tase.2023.3337199
2024-01-01
Abstract:This article presents a novel high-gain loop-shaping (HGLS) method for precision motion devices by introducing high gains into the control loop. The high gains can be generated via the inclusion of a low pass filter in a feedback manner, which could significantly improve the tracking accuracy for general trajectories with a simple structure. This control law also endows the superiority of robustness against model variations and external disturbances. To evaluate the performance of the proposed HGLS method, tracking and disturbance rejection experiments are conducted on a custom-designed piezo-actuated nanopositioner. The experimental results demonstrate the advancement of HGLS in terms of precision motion control, where the root-mean-squared error is reduced from 7.8 nm to 1.8 nm as compared with the high-gain proportional-integral controller with the same phase margin when tracking a random Non-Uniform Rational B-splines curve with large external disturbances. With its remarkable advantages of high tracking accuracy and simple structure, this method offers a practical solution for industrial automation applications for enhancing the control performance. Note to Practitioners —Advanced control methods are of great importance in meeting the tremendous requirements of automation in mechanical and electronic systems. Nonetheless, it is still challenging for a controller to simultaneously realize high control accuracy and strong robustness for general trajectories while retaining a simple structure. In this paper, a novel HGLS method is proposed. For this method, high control accuracy and strong robustness can be achieved by introducing high gains into the control loop, which can be generated by the inclusion of a simple low pass inside the closed-loop plant. A trade-off between the control gain and stability margin can be balanced only by tuning the parameters of the employed low-pass filter, and thus the implementation of the HGLS is rather simple. This method has no constraints on the form or order of the plant models and holds robustness against model variations. As compared with the high-gain proportionalintegral controller with the same phase margin, the control gain of the HGLS is much higher within the effective control bandwidth. Therefore, outstanding control performance can be achieved for general trajectories within this bandwidth. Due to the strong universality and simple structure of the HGLS, it is promising to be implemented in various industrial automation equipment, such as machine tools, multi-axis systems, robotics, scanners, etc.