Linear MPC based on data-driven Artificial Neural Networks for large-scale nonlinear distributed parameter systems

Weiguo Xie; Ioannis Bonis; Constantinos Theodoropoulos

doi:10.1016/B978-0-444-59520-1.50101-9

Linear MPC based on data-driven Artificial Neural Networks for large-scale nonlinear distributed parameter systems

Weiguo Xie, Ioannis Bonis, Constantinos Theodoropoulos

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

Process controller synthesis with detailed models is a challenging task, which may lead to many advantageous closed-loop features. Model reduction such as Proper Orthogonal Decomposition (POD) and (adaptive) linearization can be applied to tackle with the arising problems, whereas process data can be directly used to build accurate models via training of artificial neural networks (ANN). In this contribution, we present two methodologies we have recently developed, which combine ANN with POD, for use in the context of MPC: the process at hand is represented as a sum of products of time- varying coefficients (computed with ANN) with the POD basis functions computed from plant " snapshots" The resulting accurate model can be used in NMPC, or trajectory piecewise linearization along a reference path can be applied on the ANN, yielding a series of linear models, suitable for linear MPC. © 2012 Elsevier B.V.

Original language	English
Title of host publication	Computer Aided Chemical Engineering\|Comput. Aided Chem. Eng.
Publisher	Elsevier BV
Pages	1212-1216
Number of pages	4
Volume	30
DOIs	https://doi.org/10.1016/B978-0-444-59520-1.50101-9
Publication status	Published - 2012
Event	22nd European Symposium on Computer Aided Process Engineering - University College London, Gower Street, London, UK, WC1E 6BT Duration: 17 Jun 2012 → 20 Jun 2012

Conference

Conference	22nd European Symposium on Computer Aided Process Engineering
City	University College London, Gower Street, London, UK, WC1E 6BT
Period	17/06/12 → 20/06/12

Keywords

Model predictive control
Model reduction for (non-)linear predictive control
Neural network training
Proper orthogonal decomposition
Reduced order NMPC

Access to Document

10.1016/B978-0-444-59520-1.50101-9

Cite this

@inproceedings{92e127aebed74fbabf9c6ec3e23731cd,

title = "Linear MPC based on data-driven Artificial Neural Networks for large-scale nonlinear distributed parameter systems",

abstract = "Process controller synthesis with detailed models is a challenging task, which may lead to many advantageous closed-loop features. Model reduction such as Proper Orthogonal Decomposition (POD) and (adaptive) linearization can be applied to tackle with the arising problems, whereas process data can be directly used to build accurate models via training of artificial neural networks (ANN). In this contribution, we present two methodologies we have recently developed, which combine ANN with POD, for use in the context of MPC: the process at hand is represented as a sum of products of time- varying coefficients (computed with ANN) with the POD basis functions computed from plant {"} snapshots{"} The resulting accurate model can be used in NMPC, or trajectory piecewise linearization along a reference path can be applied on the ANN, yielding a series of linear models, suitable for linear MPC. {\textcopyright} 2012 Elsevier B.V.",

keywords = "Model predictive control, Model reduction for (non-)linear predictive control, Neural network training, Proper orthogonal decomposition, Reduced order NMPC",

author = "Weiguo Xie and Ioannis Bonis and Constantinos Theodoropoulos",

year = "2012",

doi = "10.1016/B978-0-444-59520-1.50101-9",

language = "English",

volume = "30",

pages = "1212--1216",

booktitle = "Computer Aided Chemical Engineering|Comput. Aided Chem. Eng.",

publisher = "Elsevier BV",

address = "Netherlands",

note = "22nd European Symposium on Computer Aided Process Engineering ; Conference date: 17-06-2012 Through 20-06-2012",

}

Xie, W, Bonis, I & Theodoropoulos, C 2012, Linear MPC based on data-driven Artificial Neural Networks for large-scale nonlinear distributed parameter systems. in Computer Aided Chemical Engineering|Comput. Aided Chem. Eng.. vol. 30, Elsevier BV, pp. 1212-1216, 22nd European Symposium on Computer Aided Process Engineering, University College London, Gower Street, London, UK, WC1E 6BT, 17/06/12. https://doi.org/10.1016/B978-0-444-59520-1.50101-9

Linear MPC based on data-driven Artificial Neural Networks for large-scale nonlinear distributed parameter systems. / Xie, Weiguo; Bonis, Ioannis; Theodoropoulos, Constantinos.
Computer Aided Chemical Engineering|Comput. Aided Chem. Eng.. Vol. 30 Elsevier BV, 2012. p. 1212-1216.

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Linear MPC based on data-driven Artificial Neural Networks for large-scale nonlinear distributed parameter systems

AU - Xie, Weiguo

AU - Bonis, Ioannis

AU - Theodoropoulos, Constantinos

PY - 2012

Y1 - 2012

N2 - Process controller synthesis with detailed models is a challenging task, which may lead to many advantageous closed-loop features. Model reduction such as Proper Orthogonal Decomposition (POD) and (adaptive) linearization can be applied to tackle with the arising problems, whereas process data can be directly used to build accurate models via training of artificial neural networks (ANN). In this contribution, we present two methodologies we have recently developed, which combine ANN with POD, for use in the context of MPC: the process at hand is represented as a sum of products of time- varying coefficients (computed with ANN) with the POD basis functions computed from plant " snapshots" The resulting accurate model can be used in NMPC, or trajectory piecewise linearization along a reference path can be applied on the ANN, yielding a series of linear models, suitable for linear MPC. © 2012 Elsevier B.V.

AB - Process controller synthesis with detailed models is a challenging task, which may lead to many advantageous closed-loop features. Model reduction such as Proper Orthogonal Decomposition (POD) and (adaptive) linearization can be applied to tackle with the arising problems, whereas process data can be directly used to build accurate models via training of artificial neural networks (ANN). In this contribution, we present two methodologies we have recently developed, which combine ANN with POD, for use in the context of MPC: the process at hand is represented as a sum of products of time- varying coefficients (computed with ANN) with the POD basis functions computed from plant " snapshots" The resulting accurate model can be used in NMPC, or trajectory piecewise linearization along a reference path can be applied on the ANN, yielding a series of linear models, suitable for linear MPC. © 2012 Elsevier B.V.

KW - Model predictive control

KW - Model reduction for (non-)linear predictive control

KW - Neural network training

KW - Proper orthogonal decomposition

KW - Reduced order NMPC

U2 - 10.1016/B978-0-444-59520-1.50101-9

DO - 10.1016/B978-0-444-59520-1.50101-9

M3 - Conference contribution

VL - 30

SP - 1212

EP - 1216

BT - Computer Aided Chemical Engineering|Comput. Aided Chem. Eng.

PB - Elsevier BV

T2 - 22nd European Symposium on Computer Aided Process Engineering

Y2 - 17 June 2012 through 20 June 2012

ER -

Linear MPC based on data-driven Artificial Neural Networks for large-scale nonlinear distributed parameter systems

Abstract

Conference

Keywords

Access to Document

Fingerprint

Cite this