• 论文 •    

基于支持向量机工具的性能劣化建模方法

邓超,吴军,万紫娟   

  1. 华中科技大学 数字制造装备与技术国家重点实验室,湖北武汉430074
  • 出版日期:2009-04-15 发布日期:2009-04-25

Performance degradation model based on support vector machine

DENG Chao,WU Jun,WAN Zi-juan   

  1. State Key Lab of Digital Manufacturing Equipment & Technology,Huazhong University of Science and Technology, Wuhan 430074, China
  • Online:2009-04-15 Published:2009-04-25

摘要: 针对基于故障数据的数控装备可靠性研究中的小样本问题,提出了建立基于支持向量机的性能劣化模型。在研究支持向量机的建模理论和参数优化方法的基础上,将最小二乘法支持向量机工具LSSVM.M应用于性能退化数据处理,提出一种改进的参数选择方法,以提高拟合和预测准确性。通过实例,验证了该方法的可行性,并建立了数控机床加工精度的性能劣化模型,为可靠性评估奠定了基础。

关键词: 支持向量机, 性能劣化模型, 退化数据, 可靠性评估

Abstract: In order to solve the small samples problem in reliability assessment based on failure data of Numerical Control (NC) equipments, the method of degradation modeling based on Support Vector Machine (SVM) was proposed. After studying on modeling theories of SVM and the parameter optimization of SVM, the Least Squares Support Vector Machines tool (LSSVM.M) was used to process the performance degradation data. An improved parameter selection method was presented, which improved the veracity of simulation and forecasting. After validating by examples, the performance degradation model of NC machining accuracy was established, which laid foundation for reliability assessment.

Key words: support vector machine, performance degradation model, degradation data, reliability assessment, numerical control system

中图分类号: