Computer Integrated Manufacturing System ›› 2022, Vol. 28 ›› Issue (6): 1814-1822.DOI: 10.13196/j.cims.2022.06.019
Previous Articles Next Articles
Online:
Published:
Supported by:
郭润琪1,2,盛步云1,2+,陆辛成1,2
基金资助:
Abstract: The adaptive distribution of measurement points is the key to machining quality inspection by the digital measurement of free-form surface.To deal with the difficulties in the adaptive planning of measuring point for free-form surfaces,a study on the self-adaptive positioning of curved surfaces was carried out.A sample set was generated by a set method for determining the measurement point density under a given condition,and a model for predicting the number of measurement points to reduce the number of measurement points was constructed by training a neural network combination model with the sample set.According to the given measurement point density threshold,clustering centers were obtained adaptively,and the surface was divided into sub-surfaces with approximately equal measurement point density by using area growth algorithm.For each sub-surface,according to the measurement point density,local uniform distribution of points was proposed,and an adaptive algorithm for measuring points based on surface division was proposed.The experimental results showed that the proposed algorithm should reduce the number of measurement points in comparison to that of the iterative reconstruction algorithm,and the result of the algorithm described here should be closer to the sample group under the same measurement points in comparison to that of the uniform algorithm.Dividing the surface from a complex,irregular model into several sub-surfaces could reduce measurement points,and the planned measurement points could better reflect the machining quality of freeform surfaces.
Key words: free-form surface, adaptive sampling, measured point density, surface division
摘要: 测点自适应规划是复杂曲面类零件在加工质量数字化测量环节的关键。针对自由曲面自适应测点规划问题,开展曲面自适应布点研究。设置给定条件下的测点密度判定方法,生成样本集,再训练神经网络组合模型,构建测点密度预测模型,用于精简测点数量;按照给定的测点密度阈值自适应获取聚类中心,并采用区域生长算法将曲面划分为测点密度大致相等的子曲面;以子曲面为基础,根据其测点密度进行局部均匀布点,提出一种基于曲面划分的自适应测点规划算法。实验结果表明,在该算法的误差结果接近样本组的情况下,所需测点数量较迭代重构算法明显减少;在相同测点数量下,该算法的误差比均匀布点算法更接近样本组。根据测点密度,将曲面由一个复杂、非规则的模型分成若干测点密度相近的子曲面可以减少测点数量;所规划的测点能较好地描述自由曲面的加工质量。
关键词: 自由曲面, 自适应布点, 测点密度, 曲面划分
CLC Number:
TP391
郭润琪, 盛步云, 陆辛成. 基于三角网格模型的自由曲面自适应测点规划[J]. 计算机集成制造系统, 2022, 28(6): 1814-1822.
0 / Recommend
Add to citation manager EndNote|Ris|BibTeX
URL: http://www.cims-journal.cn/EN/10.13196/j.cims.2022.06.019
http://www.cims-journal.cn/EN/Y2022/V28/I6/1814