摘要
针对深孔存在的轴线直线度误差,本文采用端点连线法、最小二乘法、极值搜索法、分割逼近法及遗传算法5种评定方法分别对某身管光电检测实验所得数据进行评定.结果表明:以最小包容区域条件为基础的后3种方法,其评定精度高于端点连线法和最小二乘法,可以较为准确地反映直线的直线度误差,尤其遗传算法作为全局优化搜索算法,其操作灵活实用、应用广泛且评定结果最好.此外本文还对深孔轴线直线度误差的修正提出了改进方法,具有较高的学术价值.
In view of the linear straightness error of deep hole,five methods is used in this paper which are endpoint connection,least squares,extreme search,segmentation approximation and genetic algorithm to evaluate the data obtained from the photoelectric detection experiment of a barrel.The results show that the latter three methods based on the minimum containment region are more accurate than the endpoint connection method and the least squares method,and can accurately reflect the linearity error of the straight line,Especially as a global optimization search algorithm,the genetic algorithm is flexible and practical to operate and widely used with the best assessment results.In addition,an improved method is also proposed for the correction of the straightness error of the deep-hole axis,which has significant academic value.
作者
李晓
包建东
LI Xiao;BAO Jiandong(School of Mechanical Engineering,Nanjing University of Science and Technology,Nanjing 210094;School of Automation Nanjing University of Science and Technology,Nanjing 210094,China)
出处
《测试技术学报》
2018年第5期411-415,共5页
Journal of Test and Measurement Technology
关键词
深孔直线度
分割逼近法
遗传算法
最小区域包容条件
全局优化搜索
deep hole straightness
segmentation approximation
genetic algorithm
the minimum containment region
global optimization search
