Key-performance-indicator-related state monitoring based on kernel canonical correlation analysis

被引：49

作者：

Chen, Qing ^{[1
]}

Wang, Youqing ^{[1
,2
]}

机构：

[1] Beijing Univ Chem Technol, Coll Informat Sci & Technol, Beijing 100029, Peoples R China

[2] Shandong Univ Sci & Technol, Coll Elect Engn & Automat, Qingdao 266590, Peoples R China

来源：

CONTROL ENGINEERING PRACTICE | 2021年 / 107卷

基金：

中国国家自然科学基金;

关键词：

Kernel canonical correlation analysis; Key performance indicator; State monitoring; Fault detection; FAULT-DETECTION METHODS; QUALITY; DECOMPOSITION; DIAGNOSIS;

D O I：

10.1016/j.conengprac.2020.104692

中图分类号：

TP [自动化技术、计算机技术];

学科分类号：

0812 ;

摘要：

As a multivariate statistical analysis method, canonical correlation analysis (CCA) performs well for state monitoring of linear processes, but most industrial processes are nonlinear. To solve this problem, kernel canonical correlation analysis (KCCA) has been adopted; however, KCCA still has key performance indicators (KPI)-related issue. In this paper, two improved KCCA methods are proposed to deal with KPI-related issue. One is performing singular value decomposition (SVD) on the correlation coefficient matrix, then the kernel matrix can be divided into KPI-related and KPI-unrelated parts. Another one is performing general singular value decomposition (GSVD) on two coefficient matrices. In addition, this paper also performs fault detectability analysis and computational complexity analysis on these two methods. Finally, the Tennessee Eastman (TE) process is used in this study to verify the efficacy of these two proposed methods.

引用

页数：12

共 32 条

[1] Kernel independent component analysis [J].

Bach, FR ;

Jordan, MI .

JOURNAL OF MACHINE LEARNING RESEARCH, 2003, 3 (01) :1-48

[2]

Chen Z., 2017, Data-Driven Fault Detection for Industrial Processes, DOI [10.1007/978-3-658-16756-1, DOI 10.1007/978-3-658-16756-1]

[3] Improved canonical correlation analysis-based fault detection methods for industrial processes [J].

Chen, Zhiwen ;

Zhang, Kai ;

Ding, Steven X. ;

Shardt, Yuri A. W. ;

Hu, Zhikun .

JOURNAL OF PROCESS CONTROL, 2016, 41 :26-34

[4] Canonical correlation analysis-based fault detection methods with application to alumina evaporation process [J].

Chen, Zhiwen ;

Ding, Steven X. ;

Zhang, Kai ;

Li, Zhebin ;

Hu, Zhikun .

CONTROL ENGINEERING PRACTICE, 2016, 46 :51-58

[5] Identification of faulty sensors using principal component analysis [J].

Dunia, R ;

Qin, SJ ;

Edgar, TF ;

McAvoy, TJ .

AICHE JOURNAL, 1996, 42 (10) :2797-2812

[6] Online monitoring of nonlinear multivariate industrial processes using filtering KICA-PCA [J].

Fan, Jicong ;

Qin, S. Joe ;

Wang, Youqing .

CONTROL ENGINEERING PRACTICE, 2014, 22 :205-216

[7] Review of Recent Research on Data-Based Process Monitoring [J].

Ge, Zhiqiang ;

Song, Zhihuan ;

Gao, Furong .

INDUSTRIAL & ENGINEERING CHEMISTRY RESEARCH, 2013, 52 (10) :3543-3562

[8]

Golub G. H., 2013, Matrix Computations, V3, DOI [10.2307/3621013, DOI 10.2307/3621013]

[9] MODIFIED PARTIAL LEAST SQUARE FOR DIAGNOSING KEY-PERFORMANCE-INDICATOR-RELATED FAULTS [J].

He, Siyuan ;

Wang, Youqing ;

Liu, Changqing .

CANADIAN JOURNAL OF CHEMICAL ENGINEERING, 2018, 96 (02) :444-454

[10] Relations between two sets of variates [J].

Hotelling, H .

BIOMETRIKA, 1936, 28 :321-377

← 1 2 3 4 →