Liquid-Liquid Dispersion Performance Prediction and Uncertainty Quantification Using Recurrent Neural Networks

We demonstrate the application of a recurrent neural network (RNN) to perform multistep and multivariate time-series performance predictions for stirred and static mixers as exemplars of complex multiphase systems. We employ two network architectures in this study, fitted with either long short-term...

Full description

Saved in:
Bibliographic Details
Published in:Industrial & engineering chemistry research Vol. 63; no. 17; p. 7853
Main Authors: Liang, Fuyue, Valdes, Juan P, Cheng, Sibo, Kahouadji, Lyes, Shin, Seungwon, Chergui, Jalel, Juric, Damir, Arcucci, Rossella, Matar, Omar K
Format: Journal Article
Language:English
Published: United States 01.05.2024
ISSN:0888-5885
Online Access:Get more information
Tags: Add Tag
No Tags, Be the first to tag this record!
Abstract We demonstrate the application of a recurrent neural network (RNN) to perform multistep and multivariate time-series performance predictions for stirred and static mixers as exemplars of complex multiphase systems. We employ two network architectures in this study, fitted with either long short-term memory and gated recurrent unit cells, which are trained on high-fidelity, three-dimensional, computational fluid dynamics simulations of the mixer performance, in the presence and absence of surfactants, in terms of drop size distributions and interfacial areas as a function of system parameters; these include physicochemical properties, mixer geometry, and operating conditions. Our results demonstrate that while it is possible to train RNNs with a single fully connected layer more efficiently than with an encoder-decoder structure, the latter is shown to be more capable of learning long-term dynamics underlying dispersion metrics. Details of the methodology are presented, which include data preprocessing, RNN model exploration, and methods for model performance visualization; an ensemble-based procedure is also introduced to provide a measure of the model uncertainty. The workflow is designed to be generic and can be deployed to make predictions in other industrial applications with similar time-series data.
AbstractList We demonstrate the application of a recurrent neural network (RNN) to perform multistep and multivariate time-series performance predictions for stirred and static mixers as exemplars of complex multiphase systems. We employ two network architectures in this study, fitted with either long short-term memory and gated recurrent unit cells, which are trained on high-fidelity, three-dimensional, computational fluid dynamics simulations of the mixer performance, in the presence and absence of surfactants, in terms of drop size distributions and interfacial areas as a function of system parameters; these include physicochemical properties, mixer geometry, and operating conditions. Our results demonstrate that while it is possible to train RNNs with a single fully connected layer more efficiently than with an encoder-decoder structure, the latter is shown to be more capable of learning long-term dynamics underlying dispersion metrics. Details of the methodology are presented, which include data preprocessing, RNN model exploration, and methods for model performance visualization; an ensemble-based procedure is also introduced to provide a measure of the model uncertainty. The workflow is designed to be generic and can be deployed to make predictions in other industrial applications with similar time-series data.
We demonstrate the application of a recurrent neural network (RNN) to perform multistep and multivariate time-series performance predictions for stirred and static mixers as exemplars of complex multiphase systems. We employ two network architectures in this study, fitted with either long short-term memory and gated recurrent unit cells, which are trained on high-fidelity, three-dimensional, computational fluid dynamics simulations of the mixer performance, in the presence and absence of surfactants, in terms of drop size distributions and interfacial areas as a function of system parameters; these include physicochemical properties, mixer geometry, and operating conditions. Our results demonstrate that while it is possible to train RNNs with a single fully connected layer more efficiently than with an encoder-decoder structure, the latter is shown to be more capable of learning long-term dynamics underlying dispersion metrics. Details of the methodology are presented, which include data preprocessing, RNN model exploration, and methods for model performance visualization; an ensemble-based procedure is also introduced to provide a measure of the model uncertainty. The workflow is designed to be generic and can be deployed to make predictions in other industrial applications with similar time-series data.We demonstrate the application of a recurrent neural network (RNN) to perform multistep and multivariate time-series performance predictions for stirred and static mixers as exemplars of complex multiphase systems. We employ two network architectures in this study, fitted with either long short-term memory and gated recurrent unit cells, which are trained on high-fidelity, three-dimensional, computational fluid dynamics simulations of the mixer performance, in the presence and absence of surfactants, in terms of drop size distributions and interfacial areas as a function of system parameters; these include physicochemical properties, mixer geometry, and operating conditions. Our results demonstrate that while it is possible to train RNNs with a single fully connected layer more efficiently than with an encoder-decoder structure, the latter is shown to be more capable of learning long-term dynamics underlying dispersion metrics. Details of the methodology are presented, which include data preprocessing, RNN model exploration, and methods for model performance visualization; an ensemble-based procedure is also introduced to provide a measure of the model uncertainty. The workflow is designed to be generic and can be deployed to make predictions in other industrial applications with similar time-series data.
Author Matar, Omar K
Cheng, Sibo
Liang, Fuyue
Kahouadji, Lyes
Shin, Seungwon
Juric, Damir
Arcucci, Rossella
Valdes, Juan P
Chergui, Jalel
Author_xml – sequence: 1
  givenname: Fuyue
  surname: Liang
  fullname: Liang, Fuyue
  organization: Department of Chemical Engineering, Imperial College London, London SW7 2AZ, U.K
– sequence: 2
  givenname: Juan P
  orcidid: 0000-0003-3249-5194
  surname: Valdes
  fullname: Valdes, Juan P
  organization: Department of Chemical Engineering, Imperial College London, London SW7 2AZ, U.K
– sequence: 3
  givenname: Sibo
  surname: Cheng
  fullname: Cheng, Sibo
  organization: CEREA, École des Ponts ParisTech-EdF R&D, Champs-sur-Marne 77455, France
– sequence: 4
  givenname: Lyes
  surname: Kahouadji
  fullname: Kahouadji, Lyes
  organization: Department of Chemical Engineering, Imperial College London, London SW7 2AZ, U.K
– sequence: 5
  givenname: Seungwon
  surname: Shin
  fullname: Shin, Seungwon
  organization: Department of Mechanical and System Design Engineering, Hongik University, Seoul 04066, Republic of Korea
– sequence: 6
  givenname: Jalel
  surname: Chergui
  fullname: Chergui, Jalel
  organization: Centre National de la Recherche Scientifique (CNRS), Laboratoire Interdisciplinaire des Sciences du Numérique (LISN), Université Paris Saclay, Orsay 91400, France
– sequence: 7
  givenname: Damir
  surname: Juric
  fullname: Juric, Damir
  organization: Department of Applied Mathematics and Theoretical Physics, University of Cambridge, Cambridge CB3 0WA, U.K
– sequence: 8
  givenname: Rossella
  surname: Arcucci
  fullname: Arcucci, Rossella
  organization: Department of Earth Science & Engineering, Imperial College London, London SW7 2AZ, U.K
– sequence: 9
  givenname: Omar K
  orcidid: 0000-0002-0530-8317
  surname: Matar
  fullname: Matar, Omar K
  organization: Department of Chemical Engineering, Imperial College London, London SW7 2AZ, U.K
BackLink https://www.ncbi.nlm.nih.gov/pubmed/38706982$$D View this record in MEDLINE/PubMed
BookMark eNo1kDtPwzAUhT0U0QfsTCgjS4rj2Ik9ovKUKiiIztGNc40MjdPajlD_PSmF6ZO-e3Skc6dk5DqHhFxkdJ5Rll2DDnOL2s-5ppRmfEQmVEqZCinFmExD-By0EJyfknEuS1ooySbELe2ut016RHJrwxZ9sJ1LVuhN51twGpOVx8bqeNDgmmQ9OB_BurhPXntw0Rqr4fe8DtZ9JG-oe-_RxeQZew-bAfG781_hjJwY2AQ8_-OMrO_v3heP6fLl4Wlxs0yBcxHTUoMqclNggybPQdGS1yjzUktFa2N40RQgSikV0wKGTcikUVTVjao541qxGbk69m59t-sxxKq1QeNmAw67PlQ5FRlnRZkfopd_0b5usam23rbg99X_j9gPSkxsWg
CitedBy_id crossref_primary_10_1016_j_ijmultiphaseflow_2024_104936
crossref_primary_10_1007_s00530_025_01917_z
crossref_primary_10_1038_s41598_025_87370_4
crossref_primary_10_1002_cjce_25379
crossref_primary_10_1016_j_cjche_2024_09_028
ContentType Journal Article
Copyright 2024 The Authors. Published by American Chemical Society.
Copyright_xml – notice: 2024 The Authors. Published by American Chemical Society.
DBID NPM
7X8
DOI 10.1021/acs.iecr.4c00014
DatabaseName PubMed
MEDLINE - Academic
DatabaseTitle PubMed
MEDLINE - Academic
DatabaseTitleList PubMed
MEDLINE - Academic
Database_xml – sequence: 1
  dbid: NPM
  name: PubMed
  url: http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed
  sourceTypes: Index Database
– sequence: 2
  dbid: 7X8
  name: MEDLINE - Academic
  url: https://search.proquest.com/medline
  sourceTypes: Aggregation Database
DeliveryMethod no_fulltext_linktorsrc
Discipline Engineering
ExternalDocumentID 38706982
Genre Journal Article
GroupedDBID -~X
.DC
.K2
4.4
53G
55A
5GY
5VS
6TJ
7~N
AABXI
ABMVS
ABQRX
ABUCX
ACGFO
ACJ
ACS
ADHLV
AEESW
AENEX
AFEFF
AGXLV
AHGAQ
ALMA_UNASSIGNED_HOLDINGS
AQSVZ
BAANH
CS3
CUPRZ
DU5
EBS
ED~
F5P
GGK
GNL
IH9
JG~
LG6
NPM
P2P
ROL
TAE
TN5
UI2
VF5
VG9
W1F
WH7
~02
7X8
ABBLG
ABLBI
ID FETCH-LOGICAL-a445t-7ca963f6edef33a9074be837c890bff46d6a578892c5a554e28f909bd9b424c92
IEDL.DBID 7X8
ISICitedReferencesCount 8
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=001242278900001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN 0888-5885
IngestDate Sun Nov 09 14:04:56 EST 2025
Thu Jan 02 22:33:22 EST 2025
IsDoiOpenAccess false
IsOpenAccess true
IsPeerReviewed true
IsScholarly true
Issue 17
Language English
License 2024 The Authors. Published by American Chemical Society.
LinkModel DirectLink
MergedId FETCHMERGED-LOGICAL-a445t-7ca963f6edef33a9074be837c890bff46d6a578892c5a554e28f909bd9b424c92
Notes ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 23
ORCID 0000-0003-3249-5194
0000-0002-0530-8317
OpenAccessLink https://hal.science/hal-04647255
PMID 38706982
PQID 3051426739
PQPubID 23479
ParticipantIDs proquest_miscellaneous_3051426739
pubmed_primary_38706982
PublicationCentury 2000
PublicationDate 2024-05-01
PublicationDateYYYYMMDD 2024-05-01
PublicationDate_xml – month: 05
  year: 2024
  text: 2024-05-01
  day: 01
PublicationDecade 2020
PublicationPlace United States
PublicationPlace_xml – name: United States
PublicationTitle Industrial & engineering chemistry research
PublicationTitleAlternate Ind Eng Chem Res
PublicationYear 2024
SSID ssj0005544
Score 2.4917078
Snippet We demonstrate the application of a recurrent neural network (RNN) to perform multistep and multivariate time-series performance predictions for stirred and...
SourceID proquest
pubmed
SourceType Aggregation Database
Index Database
StartPage 7853
Title Liquid-Liquid Dispersion Performance Prediction and Uncertainty Quantification Using Recurrent Neural Networks
URI https://www.ncbi.nlm.nih.gov/pubmed/38706982
https://www.proquest.com/docview/3051426739
Volume 63
WOSCitedRecordID wos001242278900001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
hasFullText
inHoldings 1
isFullTextHit
isPrint
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwpV1JS8QwFA7qeNCD-zJuRPDasZOmS04i6uBBSxUH5lbSLNBLZ-mM4L_3vabDeBEELy2UtoSX5X1578v3CLmJOLh9GQJys77FNKPwCtO3XiT7jMXaKq78pthEnKbJaCSyNuBWt7TK5ZrYLNR6rDBGfhugUDeL4kDcTaYeVo3C7GpbQmOddAKAMkjpikcrtXBwldyhSDxdlIRtmhLc2q1Uda8EYNaDtuA24XeA2Tiawe5_m7hHdlqISe_dmNgna6Y6INs_hAcPSfVSThel9tyNPpaoF45xM5qtDhLQbIZZHOw5KitNh_Cs4Q_Mv-jbQjqWUdOxtCEe0HeM3aPaE0XJD2hC6jjm9REZDp4-Hp69tvKCJzkP516sJExMGxltbBBI3EAXBrayKhF-YS2PdCRhqieCqVCClQ1LrPBFoUXBGVeCHZONalyZU0IRghQKXuPa5_0E_sWZDqU1OmY84rZLrpfGzGFkY7pCVma8qPOVObvkxPVIPnESHHmA6VmRsLM_fH1OthggEcdSvCAdC_PaXJJN9Tkv69lVM2Tgmmav3x3zzk8
linkProvider ProQuest
openUrl ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=Liquid-Liquid+Dispersion+Performance+Prediction+and+Uncertainty+Quantification+Using+Recurrent+Neural+Networks&rft.jtitle=Industrial+%26+engineering+chemistry+research&rft.au=Liang%2C+Fuyue&rft.au=Valdes%2C+Juan+P&rft.au=Cheng%2C+Sibo&rft.au=Kahouadji%2C+Lyes&rft.date=2024-05-01&rft.issn=0888-5885&rft.volume=63&rft.issue=17&rft.spage=7853&rft_id=info:doi/10.1021%2Facs.iecr.4c00014&rft_id=info%3Apmid%2F38706982&rft_id=info%3Apmid%2F38706982&rft.externalDocID=38706982
thumbnail_l http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=0888-5885&client=summon
thumbnail_m http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=0888-5885&client=summon
thumbnail_s http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=0888-5885&client=summon