中国机械工程 ›› 2010, Vol. 21 ›› Issue (12): 1397-1400.

• 机械基础工程 • 上一篇    下一篇

基于支持向量机的数控机床空间误差辨识与补偿
 

吴雄彪1;姚鑫骅2;何振亚2;傅建中2
  

  1. 1.金华职业技术学院,金华,321017
    2.浙江大学,杭州,310027
  • 出版日期:2010-06-25 发布日期:2010-07-01
  • 基金资助:
    国家科技重大专项(2009ZX04014-026)
    National Science and Technology Major Project ( No. 2009ZX04014-026)

Identification and Compensation for Volumetric Errors of CNC Machine Tools Based on SVM

Wu Xiongbiao1;Yao Xinhua2;He Zhenya2;Fu Jianzhong2   

  1.  
    1.Jinhua College of Profession and Technology,Jinhua,Zhejiang,321017
    2.Zhejiang University,Hangzhou,310027
  • Online:2010-06-25 Published:2010-07-01
  • Supported by:
    National Science and Technology Major Project ( No. 2009ZX04014-026)

摘要:

为了精确地建立数控机床空间误差预测模型,提出基于支持向量机的空间误差辨识与补偿方法。通过训练学习,优化了最小二乘支持向量机参数,采用激光矢量分步对角线法在一台数控铣床上进行了空间误差辨识和补偿实验研究。结果表明:支持向量机预测模型能有效辨识数控机床的空间误差。与神经网络预测模型进行了比较,表明支持向量机空间误差预测模型补偿精度高、建模速度快,通过其补偿可有效提高数控机床的加工精度。

关键词:

Abstract:

A novel method of identification and compensation volumetric errors for machine tools based on SVM was presented herein. Training and learning the error samples and optimizing the SVM parameters, the volumetric error models were obtained for compensation. By laser sequential step diagonal measurement, the volumetric errors were identified and compensated. The experimental results illustrate that the SVM model is adaptive to forecast volumetric errors of machine tools.
Comparing the neutral network model, the SVM model of volumetric error is more precise and quick. Using the SVM volumetric error model, it is succeeded to enhance the precision for CNC machine tools.

Key words: CNC machine tool, volumetric error, support vector machine(SVM), error compensation

中图分类号: