摘要
复杂工件轮廓在机测量中,由于单个几何特征数据量小,转接处易受噪声干扰,单纯使用最小二乘法拟合评定具有一定的局限性。基于数控加工中心在机测量系统,首先提出了特征辨识和滤波算法,将复杂工件轮廓测量数据分解成水平直线段、倾斜直线段和圆弧段等典型几何特征;然后开发了相应的特征评定算法,并计算出几何特征的标注尺寸;最后进行了实例分析。结果表明:在工件随机放置的状态下,特征辨识和滤波算法能准确有效地提取出几何特征点,实现典型几何特征的分解,特征评定算法精度高,能满足智能制造在机测量的需求。
In the process of on-line measurement for complex contour,the least-square method is limited due to the lack of feature points and disturbance of noise.The algorithm of feature identification and filtering was presented based on the on line measurement system of a CNC machining center firstly.The measurement data of complex contour was decomposed into classical geometric features such as horizontal line,inclined line and circle by this algorithm.Then the corresponding feature evaluation algorithm was developed,and the dimension of every feature was calculated.Finally the analysis of two cases were performed.The results indicate that the geometric feature points can be extracted by the algorithm of feature identification and filtering effectively when the workpiece is placed randomly,and the classical geometric features were obtained.They also indicate that the accuracy of the feature evaluation algorithm is high enough to meet the need of on-line measurement for intelligent manufacturing.
作者
丁文政
王明涛
陈勇
卞荣
DING Wen-zheng;WANG Ming-tao;CHEN Yong;BIAN Rong(Industrial Center,Nanjing Institute of Technology,Nanjing 211167,China)
出处
《组合机床与自动化加工技术》
北大核心
2021年第12期71-74,78,共5页
Modular Machine Tool & Automatic Manufacturing Technique
基金
国家自然科学基金(51805242)
江苏省产学研合作项目(BY20200169)
江苏高校“青蓝工程”中青年学术带头人资助项目和南京工程学院校级科研基金(JCYJ201843)。
关键词
在机测量
特征滤波
特征评定
最小二乘法
on-line measurement
feature filtering
feature evaluation
least square metho
