A model-free approach for auto-tuning of model predictive control

Nhat Tran; J. Scholten; Leyla Ozkan; A.C.P.M. Backx

doi:10.3182/20140824-6-ZA-1003.01494

A model-free approach for auto-tuning of model predictive control

Nhat Tran, J. Scholten, Leyla Ozkan, A.C.P.M. Backx

Research output: Contribution to journal › Conference article › Academic › peer-review

10 Citations (Scopus)

Abstract

A two-layer approach for the auto-tuning of model predictive control (MPC) is proposed. The bottom layer computes the weighting matrices of the cost function from a desired closed-loop bandwidth while the top layer aims at finding the optimal bandwidth. This optimum corresponds to the optimal balance between the robustness and nominal performance of the closed-loop system. To find the optimal bandwidth, the extremum seeking (ES) algorithm, a form of non-model-based adaptive optimisation, is proposed. The auto-tuning approach is tested on a binary distillation column model. It is shown that the auto-tuning approach enables the MPC system to track its optimal closed-loop bandwidth and therefore obtain the minimum output variance.

Original language	English
Pages (from-to)	2189-2194
Number of pages	6
Journal	IFAC Proceedings Volumes
Volume	47
Issue number	3
DOIs	https://doi.org/10.3182/20140824-6-ZA-1003.01494
Publication status	Published - 2014
Event	19th World Congress of the International Federation of Automatic Control (IFAC 2014 World Congress) - Cape Town International Convention Centre, Cape Town, South Africa Duration: 24 Aug 2014 → 29 Aug 2014 Conference number: 19 http://www.ifac2014.org

Keywords

Auto-tuning
Closed-loop bandwidth
Extremum seeking
Model predictive control

Access to Document

10.3182/20140824-6-ZA-1003.01494

Cite this

@article{86e3c3d113444861b05b2571a281dc4d,

title = "A model-free approach for auto-tuning of model predictive control",

abstract = "A two-layer approach for the auto-tuning of model predictive control (MPC) is proposed. The bottom layer computes the weighting matrices of the cost function from a desired closed-loop bandwidth while the top layer aims at finding the optimal bandwidth. This optimum corresponds to the optimal balance between the robustness and nominal performance of the closed-loop system. To find the optimal bandwidth, the extremum seeking (ES) algorithm, a form of non-model-based adaptive optimisation, is proposed. The auto-tuning approach is tested on a binary distillation column model. It is shown that the auto-tuning approach enables the MPC system to track its optimal closed-loop bandwidth and therefore obtain the minimum output variance.",

keywords = "Auto-tuning, Closed-loop bandwidth, Extremum seeking, Model predictive control",

author = "Nhat Tran and J. Scholten and Leyla Ozkan and A.C.P.M. Backx",

year = "2014",

doi = "10.3182/20140824-6-ZA-1003.01494",

language = "English",

volume = "47",

pages = "2189--2194",

journal = "IFAC Proceedings Volumes",

issn = "1474-6670",

publisher = "Elsevier",

number = "3",

note = "19th World Congress of the International Federation of Automatic Control (IFAC 2014 World Congress), IFAC 2014 ; Conference date: 24-08-2014 Through 29-08-2014",

url = "http://www.ifac2014.org",

}

TY - JOUR

T1 - A model-free approach for auto-tuning of model predictive control

AU - Tran, Nhat

AU - Scholten, J.

AU - Ozkan, Leyla

AU - Backx, A.C.P.M.

N1 - Conference code: 19

PY - 2014

Y1 - 2014

N2 - A two-layer approach for the auto-tuning of model predictive control (MPC) is proposed. The bottom layer computes the weighting matrices of the cost function from a desired closed-loop bandwidth while the top layer aims at finding the optimal bandwidth. This optimum corresponds to the optimal balance between the robustness and nominal performance of the closed-loop system. To find the optimal bandwidth, the extremum seeking (ES) algorithm, a form of non-model-based adaptive optimisation, is proposed. The auto-tuning approach is tested on a binary distillation column model. It is shown that the auto-tuning approach enables the MPC system to track its optimal closed-loop bandwidth and therefore obtain the minimum output variance.

AB - A two-layer approach for the auto-tuning of model predictive control (MPC) is proposed. The bottom layer computes the weighting matrices of the cost function from a desired closed-loop bandwidth while the top layer aims at finding the optimal bandwidth. This optimum corresponds to the optimal balance between the robustness and nominal performance of the closed-loop system. To find the optimal bandwidth, the extremum seeking (ES) algorithm, a form of non-model-based adaptive optimisation, is proposed. The auto-tuning approach is tested on a binary distillation column model. It is shown that the auto-tuning approach enables the MPC system to track its optimal closed-loop bandwidth and therefore obtain the minimum output variance.

KW - Auto-tuning

KW - Closed-loop bandwidth

KW - Extremum seeking

KW - Model predictive control

UR - http://www.scopus.com/inward/record.url?scp=84929723623&partnerID=8YFLogxK

U2 - 10.3182/20140824-6-ZA-1003.01494

DO - 10.3182/20140824-6-ZA-1003.01494

M3 - Conference article

AN - SCOPUS:84929723623

VL - 47

SP - 2189

EP - 2194

JO - IFAC Proceedings Volumes

JF - IFAC Proceedings Volumes

SN - 1474-6670

IS - 3

T2 - 19th World Congress of the International Federation of Automatic Control (IFAC 2014 World Congress)

Y2 - 24 August 2014 through 29 August 2014

ER -

A model-free approach for auto-tuning of model predictive control

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this