China Mechanical Engineering ›› 2011, Vol. 22 ›› Issue (8): 971-975.

Previous Articles     Next Articles

Study on Parameter Design Based on Orthogonal Test and Support Vector Machine

Sun Lin;Yang Shiyuan
  

  1. Hefei University of Technology,Hefei,230009
  • Online:2011-04-25 Published:2011-05-04
  • Supported by:
    National Natural Science Foundation of China(No. 70672096)

基于正交试验和支持向量机的参数设计方法研究

孙林;杨世元
  

  1. 合肥工业大学,合肥,230009
  • 基金资助:
    国家自然科学基金资助项目(70672096);合肥工业大学科学研究发展基金资助项目(2010HGXJ0076)合肥工业大学博士专项科研基金资助项目(2010HGBZ0302)
    National Natural Science Foundation of China(No. 70672096)

Abstract:

In response to the defects of traditional parameter design method,the paper put forward an optimization design based on SVM and orthogonal
test.SVM intelligent model was established and the limited orthogonal test swatches
 were studied, the precise relationship among each parameter and quality features of products as well as signal-to-noise ratio can be obtained,so as to guide the optimization design of technological parameter.

Key words: parameter design, support vector machine(SVM), orthogonal test, quality feature, signal-to-noise ratio

摘要:

针对传统参数设计方法存在的问题和不足,提出了一种基于支持向量机(SVM)和正交试验的参数设计方法。建立了SVM智能模型,并对有限正交试验样本进行了学习,得到了产品质量特征值及信噪比与各参数之间的精确关系,所得结果可用于指导工艺参数的优化设计。

关键词:

CLC Number: