中国机械工程 ›› 2026, Vol. 37 ›› Issue (9): 2092-2100.DOI: 10.3969/j.issn.1004-132X.2026.09.002

• 智能诊断与运维专栏 • 上一篇    

基于A-Zernike矩的精密运动平台定位误差校准方法

邓书朝1,2(), 胡奥康1, 童宝宏1,2(), 郑近德1, 王璐1   

  1. 1.安徽工业大学机械工程学院, 马鞍山, 243032
    2.特殊服役环境的智能装备制造国际科技合作基地, 马鞍山, 243032
  • 收稿日期:2025-09-11 出版日期:2026-09-25 发布日期:2026-10-21
  • 通讯作者: 童宝宏
  • 作者简介:邓书朝,男,1989年生,讲师。研究方向为机电装备运动控制。E-mail: dengsc@ahut.edu.cn
    童宝宏*(通信作者),男,1978年生,教授、博士研究生导师。研究方向为机电装备数字化设计。E-mail: bh_tong@163.com。
  • 基金资助:
    安徽高校协同创新项目(GXXT-2023-023);安徽高校协同创新项目(GXXT-2022-082);马鞍山市科技攻关项目(CSJ202303)

A-Zernike Moment Based Calibration Method for Positioning Errors of Precision Motion Stages

DENG Shuchao1,2(), HU Aokang1, TONG Baohong1,2(), ZHENG Jinde1, WANG Lu1   

  1. 1.School of Mechanical Engineering,Anhui University of Technology,Ma'anshan,Anhui,243032
    2.China International Science and Technology Cooperation Base on Intelligent Equipment Manufacturing in Special Service Environment,Ma'anshan,Anhui,243032
  • Received:2025-09-11 Online:2026-09-25 Published:2026-10-21
  • Contact: TONG Baohong

摘要:

针对光刻及类似精密装备中,微米级精密运动平台误差校准存在的精度不足、操作复杂且高成本方法普及难的问题,提出了一种基于A-Zernike矩(自适应梯度Zernike矩)的高性价比视觉校准方法。依据靶标形态特征筛选边缘初值,通过线性渐变模型、梯度采样与高斯拟合实现亚像素边缘的自适应提取,精确获取靶标中心位置。依据误差变化规律建立系统参数与靶标位置变化的关联模型,实现多参数联合标定。基于标定结果构建图像坐标与平台位移的转换关系,完成定位误差的解算与补偿。实验结果表明:A-Zernike矩边缘检测精度较传统方法的边缘检测精度提高了47.41%,校准后平台最大定位误差由15.601 μm降至0.783 μm。该方法显著提高了微米级运动平台的校准效率与可靠性。

关键词: 精密运动平台, 自适应梯度Zernike矩, 误差校准, 视觉测量, 联合标定

Abstract:

To address the issues of insufficient accuracy, operational complexity, and the high cost of popularizing precision calibration methods for micrometer-level precision motion stages in lithography and similar precision equipment, a cost-effective visual calibration method was proposed based on adaptive gradient Zernike moments(A-Zernike moments).First, initial edge points were screened based on the morphological features of the target. Adaptive extraction of sub-pixel edges was then achieved through a linear gradient model, gradient sampling, and Gaussian fitting, thereby accurately obtaining the target center position. Subsequently, by analyzing the variation patterns of errors, an association model was established between system parameters and changes in target position to realize multi-parameter integrated calibration. Finally, based on the calibration results, a conversion relationship was constructed between image coordinates and stage displacement to accomplish the calculation and compensation of positioning errors. Experimental results show that the edge detection accuracy of the A-Zernike moment method is improved by 47.41% compared with traditional methods, and the maximum positioning error of the stage is reduced from 15.601 μm to 0.783 μm after calibration. The proposed method significantly enhances the calibration efficiency and reliability of micrometer-level motion stages.

Key words: precision motion stage, adaptive gradient Zernike moment, error calibration, visual measurement, integrated calibration

中图分类号: