A simple model-free butterfly shape-based detection (BSD) method integrated with deep learning CNN for valve stiction detection and quantification

Control valve stiction is a long-standing problem within process industries. In most methods for shape-based stiction detection, they rely heavily on the traditional controller output (OP) and process variable (PV) plot (i.e. PV-OP plot) that tends to produce an �elliptical� shape which is the w...

Full description

Main Authors: Kamaruddin, B., Zabiri, H., Mohd Amiruddin, A.A.A., Teh, W.K., Ramasamy, M., Jeremiah, S.S.
Format: Article
Institution: Universiti Teknologi Petronas
Record Id / ISBN-0: utp-eprints.23109 /
Published: Elsevier Ltd 2020
Online Access: https://www.scopus.com/inward/record.uri?eid=2-s2.0-85078084604&doi=10.1016%2fj.jprocont.2020.01.001&partnerID=40&md5=b8a2b9783078100a1c34ac030c30c86a
http://eprints.utp.edu.my/23109/
Tags: Add Tag
No Tags, Be the first to tag this record!
id utp-eprints.23109
recordtype eprints
spelling utp-eprints.231092021-08-19T05:26:54Z A simple model-free butterfly shape-based detection (BSD) method integrated with deep learning CNN for valve stiction detection and quantification Kamaruddin, B. Zabiri, H. Mohd Amiruddin, A.A.A. Teh, W.K. Ramasamy, M. Jeremiah, S.S. Control valve stiction is a long-standing problem within process industries. In most methods for shape-based stiction detection, they rely heavily on the traditional controller output (OP) and process variable (PV) plot (i.e. PV-OP plot) that tends to produce an �elliptical� shape which is the widely acknowledged pattern indication for the presence of stiction. However, many of the methods suffered from unsatisfactory generalization capability when subjected to different loop dynamics. In this paper, a �butterfly� shape derived from the manipulation of the standard PV and OP data, which is more robust towards different loop dynamics, is developed for stiction detection. This simple model-free butterfly shape-based detection (BSD) method uses Stenman's one parameter stiction model, which results in a distinctive �butterfly� pattern in the presence of stiction. The proposed BSD is tested on simulated data, as well as 26 benchmark industrial case studies and has shown a relatively higher generalization capability with relatively higher successful detection rate on stiction loops and on non-stiction loops. A simple quantification algorithm based on BSD-convolutional neural network (BSD-CNN) framework is then developed to quantify the stiction severity. Based on the 15 benchmark industrial loops with stiction, the proposed BSD-CNN quantification algorithm has shown reasonable accuracy when compared to other published quantification methods in literature. © 2020 Elsevier Ltd Elsevier Ltd 2020 Article NonPeerReviewed https://www.scopus.com/inward/record.uri?eid=2-s2.0-85078084604&doi=10.1016%2fj.jprocont.2020.01.001&partnerID=40&md5=b8a2b9783078100a1c34ac030c30c86a Kamaruddin, B. and Zabiri, H. and Mohd Amiruddin, A.A.A. and Teh, W.K. and Ramasamy, M. and Jeremiah, S.S. (2020) A simple model-free butterfly shape-based detection (BSD) method integrated with deep learning CNN for valve stiction detection and quantification. Journal of Process Control, 87 . pp. 1-16. http://eprints.utp.edu.my/23109/
institution Universiti Teknologi Petronas
collection UTP Institutional Repository
description Control valve stiction is a long-standing problem within process industries. In most methods for shape-based stiction detection, they rely heavily on the traditional controller output (OP) and process variable (PV) plot (i.e. PV-OP plot) that tends to produce an �elliptical� shape which is the widely acknowledged pattern indication for the presence of stiction. However, many of the methods suffered from unsatisfactory generalization capability when subjected to different loop dynamics. In this paper, a �butterfly� shape derived from the manipulation of the standard PV and OP data, which is more robust towards different loop dynamics, is developed for stiction detection. This simple model-free butterfly shape-based detection (BSD) method uses Stenman's one parameter stiction model, which results in a distinctive �butterfly� pattern in the presence of stiction. The proposed BSD is tested on simulated data, as well as 26 benchmark industrial case studies and has shown a relatively higher generalization capability with relatively higher successful detection rate on stiction loops and on non-stiction loops. A simple quantification algorithm based on BSD-convolutional neural network (BSD-CNN) framework is then developed to quantify the stiction severity. Based on the 15 benchmark industrial loops with stiction, the proposed BSD-CNN quantification algorithm has shown reasonable accuracy when compared to other published quantification methods in literature. © 2020 Elsevier Ltd
format Article
author Kamaruddin, B.
Zabiri, H.
Mohd Amiruddin, A.A.A.
Teh, W.K.
Ramasamy, M.
Jeremiah, S.S.
spellingShingle Kamaruddin, B.
Zabiri, H.
Mohd Amiruddin, A.A.A.
Teh, W.K.
Ramasamy, M.
Jeremiah, S.S.
A simple model-free butterfly shape-based detection (BSD) method integrated with deep learning CNN for valve stiction detection and quantification
author_sort Kamaruddin, B.
title A simple model-free butterfly shape-based detection (BSD) method integrated with deep learning CNN for valve stiction detection and quantification
title_short A simple model-free butterfly shape-based detection (BSD) method integrated with deep learning CNN for valve stiction detection and quantification
title_full A simple model-free butterfly shape-based detection (BSD) method integrated with deep learning CNN for valve stiction detection and quantification
title_fullStr A simple model-free butterfly shape-based detection (BSD) method integrated with deep learning CNN for valve stiction detection and quantification
title_full_unstemmed A simple model-free butterfly shape-based detection (BSD) method integrated with deep learning CNN for valve stiction detection and quantification
title_sort simple model-free butterfly shape-based detection (bsd) method integrated with deep learning cnn for valve stiction detection and quantification
publisher Elsevier Ltd
publishDate 2020
url https://www.scopus.com/inward/record.uri?eid=2-s2.0-85078084604&doi=10.1016%2fj.jprocont.2020.01.001&partnerID=40&md5=b8a2b9783078100a1c34ac030c30c86a
http://eprints.utp.edu.my/23109/
_version_ 1741196622328496128
score 11.62408