Publication detail

A Levenberg-Marquardt Method for Nonlinear Model Predictive Control Design

DOKOUPIL, J. PIVOŇKA, P. BURLAK, V.

Original Title

A Levenberg-Marquardt Method for Nonlinear Model Predictive Control Design

Czech Title

Návrh Prediktivního Regulátoru Nelineárního Modelu Metodou Levenberg-Marquardt

English Title

A Levenberg-Marquardt Method for Nonlinear Model Predictive Control Design

Type

conference paper

Language

en

Original Abstract

This article deals with a nonlinear model predictive control design (NMPC) which applies numerical optimization using the Levenberg-Marquardt method in iterative batch mode adaptation. The proposed approach enables asymptotic tracking of a reference trajectory by a prediction of a general nonlinear model. Investigation of the properties of the NMPC is performed using the Wiener nonlinear model which is suitable to describe an unknown process dynamics. The work therefore also seeks to formulate the optimal prediction of Wiener model output in state space representation.

Czech abstract

Článek pojednává o návrhu nelineárního prediktivního regulátoru (NMPC) numerickou optimalizací metodou Levenberg-Marquardt (LM) v iterativním dávkovém režimu učení. Navržený přístup umožňuje asymtotické sledováni referenční trajektorie predikcí obecně nelineárního modelu. Vyšetřování vlastností NMPC je uskutečněno na Wienerově nelineárním modelu vhodného pro popis procesu neznámých dynamických vlastností. V rámci práce je proto rovněž řešena problematika formulace optimální predikce výstupu Wienerova modelu ve stavovém prostoru.

English abstract

This article deals with a nonlinear model predictive control design (NMPC) which applies numerical optimization using the Levenberg-Marquardt method in iterative batch mode adaptation. The proposed approach enables asymptotic tracking of a reference trajectory by a prediction of a general nonlinear model. Investigation of the properties of the NMPC is performed using the Wiener nonlinear model which is suitable to describe an unknown process dynamics. The work therefore also seeks to formulate the optimal prediction of Wiener model output in state space representation.

Keywords

nonlinear model predictive control, Wiener model, Levenberg-marquardt method, innovative state space model

RIV year

2011

Released

22.11.2011

Publisher

DAAAM International Vienna

Location

TU Wien Karlsplatz 13/311 A-1040 Vienna Austria

ISBN

978-3-901509-83-4

Book

Annals & Proceedings of 22nd DAAAM World Symposiums

Edition

2011

Edition number

1

Pages from

861

Pages to

862

Pages count

2

BibTex


@inproceedings{BUT74597,
  author="Jakub {Dokoupil} and Petr {Pivoňka} and Vladimír {Burlak}",
  title="A Levenberg-Marquardt Method for Nonlinear Model Predictive Control Design",
  annote="This article deals with a nonlinear model predictive control design (NMPC) which applies numerical optimization using the Levenberg-Marquardt method in iterative batch mode adaptation. The proposed approach enables asymptotic tracking of a reference trajectory by a prediction of a general nonlinear model. Investigation of the properties of the NMPC is performed using the Wiener nonlinear model which is suitable to describe an unknown process dynamics. The work therefore also seeks to formulate the optimal prediction of Wiener model output in state space representation.",
  address="DAAAM International Vienna",
  booktitle="Annals & Proceedings of 22nd DAAAM World Symposiums",
  chapter="74597",
  edition="2011",
  howpublished="print",
  institution="DAAAM International Vienna",
  year="2011",
  month="november",
  pages="861--862",
  publisher="DAAAM International Vienna",
  type="conference paper"
}