杨海荣,薄翠梅,龚伟俊,张广明
YANG Hai-rong,BO Cui-mei,GONG Wei-jun,ZHANG Guang-ming(Nanjing University of Technology

• 1:南京工业大学

摘要(Abstract)：

提出了一种基于主元分析和支持向量多分类器的故障诊断方法。该方法首先对工业故障数据进行主元分析提取数据集特征并降低数据维数,再把故障特征数据通过支持向量多分类器进行模式分类,最后通过特征分类诊断故障。在DAMADICS阀门模型上进行了仿真,并利用Lublin Sugar Factory工业故障数据进行了验证。仿真结果表明该方法可以快速准确地检测与诊断故障。
The principal component analysis and support vector multi-classifier of the fault diagnosis method is introduced,first of all,the failure data has been extracted PCA data sets and reduce the characteristics of the data dimension,Second,the failure characteristics data has been classified by support vector classifier,final diagnosis failure by features.Some simulations were carried out on DAMADICS valve model and Lublin Sugar Factory failure data is used to further verify.The simulation results show that the method can detection and diagnosis failure fast and accurately.

关键词(KeyWords)： 主元分析;支持向量机;故障诊断;阀门故障
principal component analysis;support vector machines;fault diagnosis;valve failure

Abstract:

Keywords:

基金项目(Foundation): 国家自然科学基金项目(60804027);; 江苏省自然科学基金项目(BK2006176);; 江苏省工业装备数学制造及控制技术重点实验项目(BM2007201);; 江苏省高校自然科学基金项目(07KJB510042)

作者(Author): 杨海荣,薄翠梅,龚伟俊,张广明
YANG Hai-rong,BO Cui-mei,GONG Wei-jun,ZHANG Guang-ming(Nanjing University of Technology

参考文献(References)：

• [1]杨波,王金全,刘启国.阀门电动执行机构故障诊断研究[J].阀门,2007,(1):39-41.
• [2]杜聚武.调节阀的故障分析与处理[J].阀门,2005,(4):39-41.
• [3]胡军红,陈正刚,胡传明.阀门定位器的两种调校及故障分析[J].仪器仪表用户,2003,(5):70-71.
• [4]黄燕,周密,黄卫星,等.阀门故障分析及其分类[J].阀门,2007,(6):44-47.
• [5]Uppal,Faisel J Patton,Ron J Witczak.Marcin Aneuro-fuzzy multiple-model observer approach to ro-bust fault diagnosis based on the DAMADICS bench-mark problem[J].Control Engineering Practice,2006,14(6):699-717.
• [6]Bocaniala,Cosmin Danut,Sa da Costa,et al.Applica-tion of a novel fuzzy classifier to fault detection and iso-lation of the DAMADICS benchmark problem[J].Control Engineering Practice,2006,14(6):653-669.
• [7]谢宏,何怡刚,彭敏放,等.离散Hopfield神经网络在混烧控制系统故障诊断中的应用[J].湖南大学学报(自然科学版),2007,(3):38-40.
• [8]时文刚,刘树林,张嘉钟,等.基于支持向量机的往复泵泵阀故障诊断方法[J].机械强度,2002,(3):50-52.
• [9]王成栋,朱永生,张优云,等.时频分析与支持向量机在柴油机气阀故障诊断中的应用[J].内燃机学报,2004,(3):50-52.
• [10]蒋浩天.工业系统的故障检测与诊断[M].北京:机械工业出版社,2003.
• [11]Vapnik VN.The Nature of Statistical Learning Theo-ry[M].New York:Springer,1995.
• [12]Vapnik V N.Statistical Learning Theory[M].NewYork:John Wiley and Sons,1998.
• [13]Joachims T,Transductive Inference for Text Classifi-cation Using Support Vector Machines[A].Proceed-ings of the 16th International Conference on MachineLearning[C].San Francisco:Morgan KaufmannPublishers,1999.200-209.