基于MTS的设备状态特征识别与优选

王宁; 孙树栋; 李淑敏; Saygin Can

基于MTS的设备状态特征识别与优选

1.
西北工业大学机电学院现代设计与集成制造教育部重点实验室, 西安 710072;
2.
德州大学圣安东尼奥分校制造系统与自动化实验室, 德克萨斯圣安东尼奥 78249-0670

基金项目:

民用飞机关键技术预先研究项目

航空科学基金项目(2009ZE53052)

陕西省科技计划项目(2010K8-11)资助

详细信息

作者简介:
王宁(1982-)，博士研究生，研究方向为故障预测和维修决策，wangningnpu@gmail.com;孙树栋(联系人)，教授，博士生导师，sdsun@nwpu.edu.cn

王宁(1982-)，博士研究生，研究方向为故障预测和维修决策，wangningnpu@gmail.com;孙树栋(联系人)，教授，博士生导师，sdsun@nwpu.edu.cn

计量
- 文章访问数: 178
- HTML全文浏览量: 26
- PDF下载量: 5
- 被引次数: 0
出版历程
- 收稿日期: 2011-03-11
- 刊出日期: 2015-06-10

Mahalanobis-taguchi System Based Characteristic Variable Identification and Selection Method for Equipment States

1.
Ministry of Education Key Laboratory of Contemporary Design and Integrated Manufacturing Technology, School of Mechantronics, Northwestern Polytechnical University, Xi'an 710072;
2.
Manufacturing Systems and Automation Laboratory, The University of Texas at San Antonio, San Antonio, Texas 78249-0670

摘要

摘要: 针对传统数据挖掘方法中对于估计设备状态假设条件过多的问题,提出了基于马氏田口法(mahalanobis-taguchi system,MTS)的设备状态特征变量识别与优选方法。将马氏距离作为描述空间的单位量,基于正常样本建立特征变量基准空间;通过特征变量正常样本与异常样本马氏距离的对比对基准空间的有效性进行验证;利用正交表和信噪比分析对特征变量进行优选,剔除对状态识别结果影响不显著的特征变量;以微型数控铣床主轴上的滚动轴承为例搭建实验平台,证明通过该方法所选特征变量对滚轴运行状态具有较强的表征能力。
- 设备状态 /
- 特征变量优选 /
- 马氏田口法 /
- 滚动轴承
Abstract: To solve the problems of overwhelming assumptions used in equipment state estimation,this paper proposes a novel kind of characteristic variable identification and selection method based on Mahalanobis-Taguchi system(MTS).First,the Mahalanobis distance(MD) is applied to describe the space distance,and the standard space is established based on the normal samples.Then,the MD of exceptional samples are calculated and compared with the normal samples to verify the correctness of the standard space.The orthogonal array and the signal to noise ratio are introduced for characteristic variable selection.An experimental platform is constructed based on the Rolling-element bearing of a spindle headstock of the micro computer numerical control(CNC) milling machine.The experimental results demonstrate that the prominent characteristic variables for Rolling-element bearing state can be effectively identified based on MTS.
- equipment state /
- characteristic variable selection /
- Mahalanobis-Taguchi system /
- rolling element bearings

HTML全文

参考文献(14)

[1]	Faiz R B,Eran A.Edirisinghe decision making for predictivemaintenance in asset information management[J].Interdiscipli-nary Journal of Information,Knowledge,and Manage-ment,2009,(4):23～36
[2]	Wang P,Vachtsevanos G.Fault Prognostics Using DynamicWavelet Neural Networks[M].AI in Equipment Service,2000
[3]	杨帆等.主元分析方法在航空发动机故障检测与诊断中的应用[J].机械科学与技术,2008,27(3):330～333
[4]	Qiu H,Lee J.Feature fusion and degradation using self-organi-zing map[A].Proceedings Machine Learning and Applica-tions[C],2004:107～114
[5]	Chinnam R B,Baruah P.A neuro-fuzzy approach for estimatingmean residual life in condition-based maintenance systems[J].International Journal of Materials and Product Technology,2004,20(1):166～179
[6]	Taguchi G,Chowdhury W Y.The Mahalanobis-taguchi System[M].New York:the McGraw-Hill Companies,2001
[7]	钟晓芳,韩之俊.评价计测仪器精度的一种新方法[J].计量与测试技术,2004,(6):27～28
[8]	王雪,李勇.零件形状误差的MTS测量识别方法[J].电测与仪表,2004,41(461):7～10
[9]	Taguchi G,et al.The Mahalanobis Taguchi System[M].New York:McGraw Hill,2001
[10]	Taguchi G,Jugulum R.The Mahalanobis-Taguchi Strategy-aPattern Technology System[M].New York:John Wiley andSons,2002
[11]	Ocak H,et al.A new bearing fault detection and diagnosis schemebased on hidden markov modeling of vibration signals[J].Acous-tics,Speech,and Signal Processing,Proceedings ICASSP,2001,(5):3141～3144
[12]	Li C J,Wu S M.On-line detection of localized defects in bear-ings by pattern recognition analysis[J].ASME Journal ofEngineering for Industry,1989,111(4):331～336
[13]	Harris T A.Rolling Bearing Analysis Fourth Edition[M].New York,NY 10158:Wiley-Interscience,2001
[14]	McFadden P D,Smith J D.Model for the vibration produced by asingle point defect in a rolling element bearing[J].Journal ofSound and Vibration,1984,96(1):69～82