An Improved Multivariate Chart Using Partial Least Squares with Continuous Ranked Probability Score

Fouzi Harrou, Ying Sun, Muddu Madakyaru, Benamar Bouyedou

Research output: Contribution to journalArticle

9 Citations (Scopus)

Abstract

Reliable fault detection systems in industrial processes provide pertinent information for improving the safety and process reliability and reducing manpower costs. Here, we present a flexible and efficient fault detection approach based on the continuous ranked probability score (CRPS) metric to detect faults in multivariate data. This approach merges partial least squares (PLS) models and the CRPS metric to separate normal from abnormal features by simultaneously taking advantage of the feature representation ability of a PLS and the fault detection capacity of a CRPS-based scheme. The proposed approach uses PLS to generate residuals, and then apply the CRPS-based chart to reveal any abnormality. Specifically, two monitoring schemes based on the CRPS measure have been introduced in this paper. The first approach uses the Shewhart scheme to evaluate the CRPS of the response variables residuals from the PLS model. The second approach merges the CRPS into the exponentially weighted moving average monitoring chart. We assess the effectiveness of these approaches by using real and simulated distillation column data. We also compare the detection quality of PLS-based CRPS charts with that of PLS-based $T^{2}$ , $Q$ , multivariate cumulative sum, and multivariate exponentially weighted moving average methods. Results show that the capacity of the proposed scheme can reliably detect faults in multivariate processes.

Original languageEnglish
Article number8396256
Pages (from-to)6715-6726
Number of pages12
JournalIEEE Sensors Journal
Volume18
Issue number16
DOIs
Publication statusPublished - 15-08-2018

Fingerprint

charts
fault detection
Fault detection
manpower
Monitoring
distillation
Distillation columns
abnormalities
safety
costs
Costs

All Science Journal Classification (ASJC) codes

  • Instrumentation
  • Electrical and Electronic Engineering

Cite this

Harrou, Fouzi ; Sun, Ying ; Madakyaru, Muddu ; Bouyedou, Benamar. / An Improved Multivariate Chart Using Partial Least Squares with Continuous Ranked Probability Score. In: IEEE Sensors Journal. 2018 ; Vol. 18, No. 16. pp. 6715-6726.
@article{1340f6518e734de58581de90779b186f,
title = "An Improved Multivariate Chart Using Partial Least Squares with Continuous Ranked Probability Score",
abstract = "Reliable fault detection systems in industrial processes provide pertinent information for improving the safety and process reliability and reducing manpower costs. Here, we present a flexible and efficient fault detection approach based on the continuous ranked probability score (CRPS) metric to detect faults in multivariate data. This approach merges partial least squares (PLS) models and the CRPS metric to separate normal from abnormal features by simultaneously taking advantage of the feature representation ability of a PLS and the fault detection capacity of a CRPS-based scheme. The proposed approach uses PLS to generate residuals, and then apply the CRPS-based chart to reveal any abnormality. Specifically, two monitoring schemes based on the CRPS measure have been introduced in this paper. The first approach uses the Shewhart scheme to evaluate the CRPS of the response variables residuals from the PLS model. The second approach merges the CRPS into the exponentially weighted moving average monitoring chart. We assess the effectiveness of these approaches by using real and simulated distillation column data. We also compare the detection quality of PLS-based CRPS charts with that of PLS-based $T^{2}$ , $Q$ , multivariate cumulative sum, and multivariate exponentially weighted moving average methods. Results show that the capacity of the proposed scheme can reliably detect faults in multivariate processes.",
author = "Fouzi Harrou and Ying Sun and Muddu Madakyaru and Benamar Bouyedou",
year = "2018",
month = "8",
day = "15",
doi = "10.1109/JSEN.2018.2850804",
language = "English",
volume = "18",
pages = "6715--6726",
journal = "IEEE Sensors Journal",
issn = "1530-437X",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "16",

}

An Improved Multivariate Chart Using Partial Least Squares with Continuous Ranked Probability Score. / Harrou, Fouzi; Sun, Ying; Madakyaru, Muddu; Bouyedou, Benamar.

In: IEEE Sensors Journal, Vol. 18, No. 16, 8396256, 15.08.2018, p. 6715-6726.

Research output: Contribution to journalArticle

TY - JOUR

T1 - An Improved Multivariate Chart Using Partial Least Squares with Continuous Ranked Probability Score

AU - Harrou, Fouzi

AU - Sun, Ying

AU - Madakyaru, Muddu

AU - Bouyedou, Benamar

PY - 2018/8/15

Y1 - 2018/8/15

N2 - Reliable fault detection systems in industrial processes provide pertinent information for improving the safety and process reliability and reducing manpower costs. Here, we present a flexible and efficient fault detection approach based on the continuous ranked probability score (CRPS) metric to detect faults in multivariate data. This approach merges partial least squares (PLS) models and the CRPS metric to separate normal from abnormal features by simultaneously taking advantage of the feature representation ability of a PLS and the fault detection capacity of a CRPS-based scheme. The proposed approach uses PLS to generate residuals, and then apply the CRPS-based chart to reveal any abnormality. Specifically, two monitoring schemes based on the CRPS measure have been introduced in this paper. The first approach uses the Shewhart scheme to evaluate the CRPS of the response variables residuals from the PLS model. The second approach merges the CRPS into the exponentially weighted moving average monitoring chart. We assess the effectiveness of these approaches by using real and simulated distillation column data. We also compare the detection quality of PLS-based CRPS charts with that of PLS-based $T^{2}$ , $Q$ , multivariate cumulative sum, and multivariate exponentially weighted moving average methods. Results show that the capacity of the proposed scheme can reliably detect faults in multivariate processes.

AB - Reliable fault detection systems in industrial processes provide pertinent information for improving the safety and process reliability and reducing manpower costs. Here, we present a flexible and efficient fault detection approach based on the continuous ranked probability score (CRPS) metric to detect faults in multivariate data. This approach merges partial least squares (PLS) models and the CRPS metric to separate normal from abnormal features by simultaneously taking advantage of the feature representation ability of a PLS and the fault detection capacity of a CRPS-based scheme. The proposed approach uses PLS to generate residuals, and then apply the CRPS-based chart to reveal any abnormality. Specifically, two monitoring schemes based on the CRPS measure have been introduced in this paper. The first approach uses the Shewhart scheme to evaluate the CRPS of the response variables residuals from the PLS model. The second approach merges the CRPS into the exponentially weighted moving average monitoring chart. We assess the effectiveness of these approaches by using real and simulated distillation column data. We also compare the detection quality of PLS-based CRPS charts with that of PLS-based $T^{2}$ , $Q$ , multivariate cumulative sum, and multivariate exponentially weighted moving average methods. Results show that the capacity of the proposed scheme can reliably detect faults in multivariate processes.

UR - http://www.scopus.com/inward/record.url?scp=85049150705&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85049150705&partnerID=8YFLogxK

U2 - 10.1109/JSEN.2018.2850804

DO - 10.1109/JSEN.2018.2850804

M3 - Article

VL - 18

SP - 6715

EP - 6726

JO - IEEE Sensors Journal

JF - IEEE Sensors Journal

SN - 1530-437X

IS - 16

M1 - 8396256

ER -