摘要
机械零件的光学超精密检测可提升零件的经济效益,为提升机械零件的光学超精密检测精度,研究基于人工智能技术的光学超精密检测技术,在机械零件的光学超精密检测中赋予人工智能机械学习模式,将机械记录的每次人为移动机械零件的动作和位置信息、光照强度等信息作为输入量进行机械计算,并将输出信息和输入量保存在存储器中。采用光学白光干涉测量和共聚焦测量技术,检测一次机械零件微米级以上宏观运动或微米级以下微观运动,根据检测光路光电探测器接收的光强大小获取检测零件各点的高度数值,在此基础上通过执行机械学习模式,在存储器中检索获取机械零件宏观运动或微观运动的全部动作和位置信息、光照强度等信息,学习得到待检测机械零件各处相对焦平面距离,实现机械零件的全面光学超精密检测。实验结果表明,该种方法可无损进行机械零件光学超精密检测,且检测误差率和检测用时较短,可用于大范围机械零件光学超精密检测。
Optical ultra-precision detection of mechanical parts can enhance economic benefits of the parts,in order to promote the optical ultra-precision accuracy of mechanical parts,the optical ultra-precision detection technology based on the technology of artificial intelligence,measured in optical ultra-precision machine parts to give ai machine learning model,move the machine records every human action and position information of mechanical parts,light intensity and other information as mechanical input calculation,and the input and output information stored in memory.Optical white light interferometry and confocal measurement technology test a machine part micron level above macroscopic movement or micron level below the micro-movement.According to the detection of the light path,photodetector receives the light of the robust little access to each point of numerical test parts’height.Based on the this,by performing a machine learning model,in memory retrieval for all mechanical parts of macro or micro-motion movement and location information,the information such as light intensity,learned all mechanical parts to be detected in focus plane distance,achieve comprehensive optical ultra-precision detection of mechanical parts.The experimental results show that this method can be used for the optical ultra-precision detection of mechanical parts without damage,and the error rate and detection time are relatively short.
作者
王琦
谭娟
WANG Qi;TAN Juan(School of Computer and Software,Weifang University of Science and Technology,Shouguang 262700,China)
出处
《激光杂志》
CAS
北大核心
2021年第2期156-160,共5页
Laser Journal
基金
山东省高等学校科技计划项目(No.J09LG27)。
关键词
人工智能
光学
超精密
检测技术
机械学习
干涉测量
共聚焦
微米级
artificial intelligence
optical
ultra precision
detection technology
mechanical learning
interferometry
confocal
micron grade
