基于自适应极值的白光干涉信号解调算法研究
White Light Interference Signal Demodulation Algorithm Based on Adaptive Extremum
祖洪飞 1陈宁 2郭钢祥 2朱靖 3张翔 4陈章位4
作者信息
- 1. 中国矿业大学 机电工程学院,江苏 徐州 221116;浙江理工大学 机械工程学院,浙江 杭州 310018;浙江谱麦科技有限公司,浙江 宁波 315048
- 2. 浙江省计量科学研究院,浙江 杭州 310018
- 3. 浙江理工大学 机械工程学院,浙江 杭州 310018
- 4. 浙江谱麦科技有限公司,浙江 宁波 315048
- 折叠
摘要
针对传统时域解调算法等光程点定位稳定性差、抗干扰能力弱,而频域解调算法效率低且部分方法受扫描采样步长限制等问题,课题组提出一种基于自适应极值的白光干涉解调算法.从干涉信号统计学规律出发,通过对干涉信号包络曲线加权平均,将它修正为高斯型包络曲线;然后,通过对包络曲线高斯拟合插值得到等光程点,进而得到测量方向的深度信息以及三维形貌;最后,以高度为20 μm标准台阶样块为测量对象进行多次实验.结果表明:与传统算法相比,课题组所提出算法的测量重复性和精度均有显著提升.该算法为白光干涉中信号解调提供了一种重复性更好、误差更小的方法.
Abstract
In order to solve the problems of poor stability and weak anti-interference ability of optical path point positioning using traditional time-domain demodulation algorithms,as well as low efficiency of frequency-domain demodulation algorithms and some methods being limited by scanning sampling step size,a white light interference demodulation algorithm was proposed based on adaptive extremum.It was corrected to a Gaussian envelope curve starting from the statistical laws of interference signals,by weighted averaging the envelope curve of the interference signal.Then,equal optical path points were obtained by Gaussian fitting and interpolation of the envelope curve,thereby obtained depth information of the measurement direction and three-dimensional morphology.Finally,multiple experiments were conducted on the 20 μm standard step sample block.The results showed that the measurement repeatability and accuracy of the proposed algorithm are significantly improved compared to the traditional algorithms.This algorithm provides a better method of repeatability and accuracy in the demodulation of white light interference signal.
关键词
白光干涉测量系统/解调算法/自适应极值/高斯包络/干涉信号解调Key words
white light interferometric measurement system/demodulation algorithm/adaptive extremum/Gaussian envelope/interferometric signal demodulation引用本文复制引用
基金项目
国家重点研发计划(2020YFC1522703)
宁波市重点研发计划暨"揭榜挂帅"项目(2023Z135)
宁波市重点研发计划暨"揭榜挂帅"项目(2023Z131)
浙江省市场监督管理局重大科研项目(20210107)
浙江省科技厅重大专项(2020C01028)
出版年
2024
国家科技期刊平台
NSTL
维普
万方数据