中国机械工程 ›› 2016, Vol. 27 ›› Issue (03): 285-289,322.

• 机械基础工程 • 上一篇    下一篇

数控机床温度敏感点变动性及其影响

苗恩铭;刘义;高增汉;刘辉   

  1. 合肥工业大学,合肥,230009
  • 出版日期:2016-02-10 发布日期:2016-02-03
  • 基金资助:
    国家自然科学基金资助重大项目(51490660,51490661);国家自然科学基金资助项目(51175142,E051102) 

Variability of Temperature-sensitive Points and Its Influences for CNC Machine Tools

Miao Enming;Liu Yi;Gao Zenghan;Liu Hui   

  1. Hefei University of Technology,Hefei,230009
  • Online:2016-02-10 Published:2016-02-03
  • Supported by:

摘要:

数控机床热误差补偿技术中的核心问题是建立能够反映机床温升与热误差之间的数学模型,其精度和稳健性则取决于模型自变量能否准确地反映机床温度场分布特性,即温度敏感点选择结果是否准确和稳定。通过对Leaderway-V450型数控加工中心主轴Z向的多批次空转数据进行分析发现,温度敏感点存在变动性特征,导致自变量间多重共线性程度发生变化,进而对模型的预测精度和稳健性产生严重影响。由于主成分回归算法具有消除自变量共线性影响作用,故提出采用该算法进行建模,并通过实际机床进行实践检验。结果表明,采用主成分回归算法建模,显著降低了温度敏感点变动性对模型预测精度的影响,能保证模型具有很好的预测精度和稳健性。

关键词: 数控机床, 温度敏感点, 多重共线性, 主成分回归

Abstract:

In thermal error compensation technology on CNC machine tools, the core issue was to establish mathematical model which might reflect the relationship between incremental temperature and thermal errors of the machine, and the accuracy and robustness of model depended on whether model's independent variables could reflect the temperature field distribution of CNC accurately, in other words, whether the temperature-sensitive points were accurate and stable. The variability of temperature-sensitive points was proved by analyzing batches of experimental data of air cutting experiments on Leaderway-V450 machine tool spindle Z direction, so it changed the degree of multi-collinearity among temperature variables, caused a serious impact on model's forecasting accuracy and robustness. Since PCR algorithm might eliminate the influences of multi-collinearity among variables, the modeling method was proposed based on PCR algorithm. And this method was used to practice tests through the experiments of actual machine. The results show, PCR model reduces the effects of changes in temperature-sensitive points on model's forecasting accuracy significantly, and the model has good forecasting accuracy and robustness.

Key words: CNC machine tool, temperature-sensitive point, multi-collinearity, principal component regression(PCR)

中图分类号: