Sequential l1 quadratic programming for nonlinear model predictive control

Research output: Contribution to journalConference articleResearchpeer-review

45 Downloads (Pure)

Abstract

In this paper, we present and describe a computationally efficient sequential l1 quadratic programming (Sl1QP) algorithm for Nonlinear Model Predictive Control (NMPC). We use a tailored trust region sequential quadratic programming for the solution of the optimal control problem (OCP) involved in the NMPC algorithm. We use a multiple shooting approach for numerical integration and sensitivity computation. A second order correction ensures a faster convergence of the SQP algorithm. We exploit the structure of the OCP by using an efficient primal-dual interior point algorithm based on Riccati factorizations and a block diagonal BFGS update of the Hessian matrix. The complexity scales linearly with the prediction horizon length. We numerically evaluate and compare the performance of our algorithm on a numerical example.

Original languageEnglish
Book seriesIFAC-PapersOnLine
Volume52
Issue number1
Pages (from-to)474-479
ISSN2405-8963
DOIs
Publication statusPublished - 1 Jan 2019
Event12th IFAC Symposium on Dynamics and Control of Process Systems - Jurerê Beach Village Hotel, Florianópolis , Brazil
Duration: 23 Apr 201926 Apr 2019
Conference number: 12
https://dycopscab2019.sites.ufsc.br/

Conference

Conference12th IFAC Symposium on Dynamics and Control of Process Systems
Number12
LocationJurerê Beach Village Hotel
CountryBrazil
CityFlorianópolis
Period23/04/201926/04/2019
Internet address

Keywords

  • Nonlinear model predictive control
  • Sequential quadratic programming
  • Trust region algorithm

Cite this

@inproceedings{f532466eb743443ba22f906386f462ab,
title = "Sequential l1 quadratic programming for nonlinear model predictive control",
abstract = "In this paper, we present and describe a computationally efficient sequential l1 quadratic programming (Sl1QP) algorithm for Nonlinear Model Predictive Control (NMPC). We use a tailored trust region sequential quadratic programming for the solution of the optimal control problem (OCP) involved in the NMPC algorithm. We use a multiple shooting approach for numerical integration and sensitivity computation. A second order correction ensures a faster convergence of the SQP algorithm. We exploit the structure of the OCP by using an efficient primal-dual interior point algorithm based on Riccati factorizations and a block diagonal BFGS update of the Hessian matrix. The complexity scales linearly with the prediction horizon length. We numerically evaluate and compare the performance of our algorithm on a numerical example.",
keywords = "Nonlinear model predictive control, Sequential quadratic programming, Trust region algorithm",
author = "Dimitri Boiroux and J{\o}rgensen, {John Bagterp}",
year = "2019",
month = "1",
day = "1",
doi = "10.1016/j.ifacol.2019.06.107",
language = "English",
volume = "52",
pages = "474--479",
journal = "IFAC-PapersOnLine",
issn = "2405-8963",
publisher = "Elsevier",
number = "1",

}

Sequential l1 quadratic programming for nonlinear model predictive control. / Boiroux, Dimitri; Jørgensen, John Bagterp.

In: IFAC-PapersOnLine, Vol. 52, No. 1, 01.01.2019, p. 474-479.

Research output: Contribution to journalConference articleResearchpeer-review

TY - GEN

T1 - Sequential l1 quadratic programming for nonlinear model predictive control

AU - Boiroux, Dimitri

AU - Jørgensen, John Bagterp

PY - 2019/1/1

Y1 - 2019/1/1

N2 - In this paper, we present and describe a computationally efficient sequential l1 quadratic programming (Sl1QP) algorithm for Nonlinear Model Predictive Control (NMPC). We use a tailored trust region sequential quadratic programming for the solution of the optimal control problem (OCP) involved in the NMPC algorithm. We use a multiple shooting approach for numerical integration and sensitivity computation. A second order correction ensures a faster convergence of the SQP algorithm. We exploit the structure of the OCP by using an efficient primal-dual interior point algorithm based on Riccati factorizations and a block diagonal BFGS update of the Hessian matrix. The complexity scales linearly with the prediction horizon length. We numerically evaluate and compare the performance of our algorithm on a numerical example.

AB - In this paper, we present and describe a computationally efficient sequential l1 quadratic programming (Sl1QP) algorithm for Nonlinear Model Predictive Control (NMPC). We use a tailored trust region sequential quadratic programming for the solution of the optimal control problem (OCP) involved in the NMPC algorithm. We use a multiple shooting approach for numerical integration and sensitivity computation. A second order correction ensures a faster convergence of the SQP algorithm. We exploit the structure of the OCP by using an efficient primal-dual interior point algorithm based on Riccati factorizations and a block diagonal BFGS update of the Hessian matrix. The complexity scales linearly with the prediction horizon length. We numerically evaluate and compare the performance of our algorithm on a numerical example.

KW - Nonlinear model predictive control

KW - Sequential quadratic programming

KW - Trust region algorithm

U2 - 10.1016/j.ifacol.2019.06.107

DO - 10.1016/j.ifacol.2019.06.107

M3 - Conference article

VL - 52

SP - 474

EP - 479

JO - IFAC-PapersOnLine

JF - IFAC-PapersOnLine

SN - 2405-8963

IS - 1

ER -