Moving window KPCA with reduced complexity for nonlinear dynamic process monitoring

Ines Jaffel, Okba Taouali, Mohamed-Faouzi Harkat, Hassani Messaoud

Research output: Contribution to journalArticle

32 Citations (Scopus)

Abstract

This paper proposes an improved Reduced Kernel Principal Component Analysis (RKPCA) for handling nonlinear dynamic systems. The proposed method is entitled Moving Window Reduced Kernel Principal Component Analysis (MW-RKPCA). It consists firstly in approximating the principal components (PCs) of the KPCA model by a reduced data set that approaches “properly” the system behavior in the order to elaborate an RKPCA model. Secondly, the proposed MW-RKPCA consists on updating the RKPCA model using a moving window. The relevance of the proposed MW-RKPCA technique is illustrated on a Tennessee Eastman process.

Original languageEnglish
Pages (from-to)184-192
Number of pages9
JournalISA Transactions
Volume64
DOIs
Publication statusPublished - 1 Sep 2016
Externally publishedYes

Fingerprint

Kernel Principal Component Analysis
Process Monitoring
Nonlinear Process
Process monitoring
Dynamic Process
principal components analysis
Principal component analysis
Nonlinear Dynamics
Nonlinear Dynamic System
Principal Components
Updating
Dynamical systems
Model

Keywords

  • Fault detection
  • KPCA
  • MW-RKPCA
  • Nonlinear dynamic process
  • RKPCA

ASJC Scopus subject areas

  • Instrumentation
  • Computer Science Applications
  • Electrical and Electronic Engineering
  • Applied Mathematics

Cite this

Moving window KPCA with reduced complexity for nonlinear dynamic process monitoring. / Jaffel, Ines; Taouali, Okba; Harkat, Mohamed-Faouzi; Messaoud, Hassani.

In: ISA Transactions, Vol. 64, 01.09.2016, p. 184-192.

Research output: Contribution to journalArticle

Jaffel, Ines ; Taouali, Okba ; Harkat, Mohamed-Faouzi ; Messaoud, Hassani. / Moving window KPCA with reduced complexity for nonlinear dynamic process monitoring. In: ISA Transactions. 2016 ; Vol. 64. pp. 184-192.
@article{47891d5e0fb244bc89d550dc06ac8a4f,
title = "Moving window KPCA with reduced complexity for nonlinear dynamic process monitoring",
abstract = "This paper proposes an improved Reduced Kernel Principal Component Analysis (RKPCA) for handling nonlinear dynamic systems. The proposed method is entitled Moving Window Reduced Kernel Principal Component Analysis (MW-RKPCA). It consists firstly in approximating the principal components (PCs) of the KPCA model by a reduced data set that approaches “properly” the system behavior in the order to elaborate an RKPCA model. Secondly, the proposed MW-RKPCA consists on updating the RKPCA model using a moving window. The relevance of the proposed MW-RKPCA technique is illustrated on a Tennessee Eastman process.",
keywords = "Fault detection, KPCA, MW-RKPCA, Nonlinear dynamic process, RKPCA",
author = "Ines Jaffel and Okba Taouali and Mohamed-Faouzi Harkat and Hassani Messaoud",
year = "2016",
month = "9",
day = "1",
doi = "10.1016/j.isatra.2016.06.002",
language = "English",
volume = "64",
pages = "184--192",
journal = "ISA Transactions",
issn = "0019-0578",
publisher = "ISA - Instrumentation, Systems, and Automation Society",

}

TY - JOUR

T1 - Moving window KPCA with reduced complexity for nonlinear dynamic process monitoring

AU - Jaffel, Ines

AU - Taouali, Okba

AU - Harkat, Mohamed-Faouzi

AU - Messaoud, Hassani

PY - 2016/9/1

Y1 - 2016/9/1

N2 - This paper proposes an improved Reduced Kernel Principal Component Analysis (RKPCA) for handling nonlinear dynamic systems. The proposed method is entitled Moving Window Reduced Kernel Principal Component Analysis (MW-RKPCA). It consists firstly in approximating the principal components (PCs) of the KPCA model by a reduced data set that approaches “properly” the system behavior in the order to elaborate an RKPCA model. Secondly, the proposed MW-RKPCA consists on updating the RKPCA model using a moving window. The relevance of the proposed MW-RKPCA technique is illustrated on a Tennessee Eastman process.

AB - This paper proposes an improved Reduced Kernel Principal Component Analysis (RKPCA) for handling nonlinear dynamic systems. The proposed method is entitled Moving Window Reduced Kernel Principal Component Analysis (MW-RKPCA). It consists firstly in approximating the principal components (PCs) of the KPCA model by a reduced data set that approaches “properly” the system behavior in the order to elaborate an RKPCA model. Secondly, the proposed MW-RKPCA consists on updating the RKPCA model using a moving window. The relevance of the proposed MW-RKPCA technique is illustrated on a Tennessee Eastman process.

KW - Fault detection

KW - KPCA

KW - MW-RKPCA

KW - Nonlinear dynamic process

KW - RKPCA

UR - http://www.scopus.com/inward/record.url?scp=84992184518&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84992184518&partnerID=8YFLogxK

U2 - 10.1016/j.isatra.2016.06.002

DO - 10.1016/j.isatra.2016.06.002

M3 - Article

AN - SCOPUS:84992184518

VL - 64

SP - 184

EP - 192

JO - ISA Transactions

JF - ISA Transactions

SN - 0019-0578

ER -