基于特征子空间的KPCA及其在故障检测与诊断中的应用

被引：21

作者：

付克昌

吴铁军

机构：

[1] 浙江大学工业控制国家重点实验室,智能系统与决策研究所

来源：

化工学报 | 2006年 / 11期

关键词：

主成分分析; PCA; 核PCA; 故障检测; 故障诊断;

D O I：

暂无

中图分类号：

TP277 [监视、报警、故障诊断系统];

学科分类号：

0804 ; 080401 ; 080402 ;

摘要：

针对标准KPCA(kernelprincipalcomponentanalysis)不适合大样本分析的缺点,提出了一种基于特征子空间的KPCA(FSKPCA)及其故障检测与诊断方法,该方法通过构建具有较小维数的特征子空间上的正交基来简化核矩阵,从而降低KPCA的计算复杂性.与标准KPCA方法相比,FSKPCA方法具有更高的计算效率且只需较小的计算机存储空间.通过非等温连续反应釜过程的故障检测与诊断的应用实例,说明了本算法的有效性.

引用

页码：2664 / 2669

页数：6

共 12 条

[1]

Fault detection and identification of nonlinear processes based on kernel PCA. Choi Sang Wook,Lee Changkyu,Lee Jong-Min,Park Jin Hyun,Lee In-Beum. Chemometrics and Intelligent Laboratory Systems . 2005

[2]

Nonlinear component analysis as a kernel eigenvalue problem. Scholkopf B,Smola A,Muller K. Neural Computation . 1998

[3]

Nonlinear process monitoring using kernel principal component analysis. Lee Jong-Min,Yoo Changkyoo,Sang Wook,Wanrolleghem Peter A,Lee In-Beum. Chemical Engineering Science . 2004

[4]

Identification of faulty sensors using principal component analysis. Dunia R,Qin S J,Edgar T F,Mcavoy TJ. American Institute of Chemical Engineers Journal . 1996

[5]

Nonlinear principal component analysis based on principal curves and neural networks. Dong D,McAvoy T J. Computers and Chemistry . 1996

[6]

Fault detection and diagnosis in industrial systems. Leo H Chiang,Richard D Braatz. . 2001

[7]

Nonlinear principal component analysis using autoassociative neural networks. Kramer M A. American Institute of Chemical Engineers Journal . 1991

[8]

Process Modeling,Si mulation,and Control for Chemical Engineers. Luyhen W. . 1988

[9]

Nonlinear principal component analysis based on principal curves and neural networks. Dong D,Mcavoy T J. Computers and Chemistry . 1996

[10]

Nonlinear principal components analysis with application to process fault detection. Jia F,Martin E B,Morris A J. International Journal of Systems Science . 2001

← 1 2 →