China Mechanical Engineering ›› 2023, Vol. 34 ›› Issue (08): 923-930.DOI: 10.3969/j.issn.1004-132X.2023.08.006

Previous Articles     Next Articles

An ER-BFGS Evaluation Method for Datum-free Profile of Complex Surfaces

FU Gaocai1;SHENG Buyun1,2;WAN Run3;YIN Xiyan2;SHENG Ganlin4   

  1. 1.School of Mechanical and Electrical Engineering,Wuhan University of Technology,Wuhan,430070
    2.School of Mechanical Engineering,Hubei University of Technology,Wuhan,430068
    3.New H3C Technology Co.,Ltd.,Hangzhou,310000
    4.Hexagon Manufacturing Intelligence(Qingdao)Co.,Ltd.,Qingdao,Shandong,266114
  • Online:2023-04-25 Published:2023-05-16

一种复杂曲面无基准轮廓度的ER-BFGS评定方法

付高财1;盛步云1,2;万润3;殷希彦2;盛甘霖4   

  1. 1武汉理工大学机电工程学院,武汉,430070
    2.湖北工业大学机械工程学院,武汉,430068
    3.新华三技术有限公司,杭州,310000
    4.海克斯康制造智能技术(青岛)有限公司,青岛,266114
  • 通讯作者: 盛步云(通信作者), 男, 1964年生, 教授、博士研究生导师。研究方向为数字集成制造、智能工厂。发表论文120余篇。E-mail:shengby@whut. edu. cn。
  • 作者简介:付高财, 男, 1989年生, 实验员。研究方向为精密测量、数字制造。发表论文10余篇。E-mail:whutjdfgc@163. com。
  • 基金资助:
    湖北省科技重大项目(2021AAA007)

Abstract:  To solve the problem that a large number of measurement points caused the multiplication of time spent in calculating surface profiles, a new surface profile evaluation method was proposed based on ER and BFGS algorithm. Based on the minimum distance function from point to surface, the minimax problem of the profile evaluation model for the minimum region was transformed into an unconstrained and differentiable optimization problem through ER principle, and solved by the fast convergent BFGS algorithm, which realized the rapid profile evaluation of complex surfaces without datum. The experiments show that the calculation time of this method may decrease as 5%~19% compared with the sequential quadratic programming method, and effectively improve the efficiency of on-line measurement. 

Key words: on-line measurement, entropy regularization(ER), BFGS(Broyden-Fletcher-Goldfarb-Shanno)algorithm, surface profile

摘要: 针对大量测点导致曲面轮廓度计算耗时倍增的问题,提出一种基于熵正则化和BFGS算法的曲面轮廓度评定方法。该方法在点到曲面的最小距离函数的基础上,通过熵正则化原理将轮廓度最小区域评定模型的极大极小问题转化为无约束可微优化问题,并利用快速收敛的BFGS算法进行求解,实现了复杂曲面无基准轮廓度的快速评定。实验表明该方法在计算耗时方面比序列二次规划方法缩短约5%~19%,能有效提高在机测量效率。

关键词: 在机测量, 熵正则化, BFGS算法, 面轮廓度

CLC Number: