Computational efficiency of Laguerre MPC using active set method

Bilal Khan*, Anthony Rossiter

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

2 Scopus citations

Abstract

This paper explores the computational efficiency of the recently developed Laguerre optimal MPC. Laguerre OMPC facilitates substantial improvements in feasibility with lit- tle or no detriment to performance while not increasing the number of optimisation variables. However, it is shown that some of the structure of the optimisation problem is lost when using a Laguerre parameterisation. In contrast, Optimal MPC can has a strong structure which can be exploited in the active set method thus allowing relatively in- expensive optimisation with large numbers of degrees of freedom. This paper contrasts the two approaches to predictive control and in particular the computational load us- ing both a generic optimiser and active set methods.

Original languageEnglish
Title of host publicationProceedings of the 13th IASTED International Conference on Intelligent Systems and Control, ISC 2011
Pages148-153
Number of pages6
DOIs
StatePublished - 2011
Externally publishedYes
Event13th IASTED International Conference on Intelligent Systems and Control, ISC 2011 - Cambridge, United Kingdom
Duration: 11 Jul 201113 Jul 2011

Publication series

NameProceedings of the IASTED International Conference on Intelligent Systems and Control
ISSN (Print)1025-8973

Conference

Conference13th IASTED International Conference on Intelligent Systems and Control, ISC 2011
Country/TerritoryUnited Kingdom
CityCambridge
Period11/07/1113/07/11

Keywords

  • Active set methods
  • Computational efficiency
  • Feasibility
  • MPC

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Software
  • Control and Systems Engineering
  • Modeling and Simulation
  • Artificial Intelligence

Fingerprint

Dive into the research topics of 'Computational efficiency of Laguerre MPC using active set method'. Together they form a unique fingerprint.

Cite this