Modelling and predictive control of a neutralisation reactor using sparse support vector machine Wiener models

Maciej Ławryńczuk

Abstract

This paperhastwoobjectives:(a)itdescribestheproblemof finding apreciseanduncomplicatedmodel of aneutralisationprocess,(b)itdetailsdevelopmentofanonlinearModelPredictiveControl(MPC) algorithm fortheplant.ThemodelhasacascadeWienerstructure,i.e.alineardynamicpartisfollowed by anonlinearsteady-stateone.ALeast-SquaresSupportVectorMachine(LS-SVM)approximatorisused as thesteady-statepart.AlthoughtheLS-SVMhasexcellentapproximationabilitiesanditmaybefound easily,itsuffersfromahugenumberofparameters.TwopruningmethodsoftheLS-SVMWienermodel are describedandcomparedwithaclassicalpruningalgorithm.Thedescribedpruningmethodsmakeit possible toremoveasmuchas70%ofsupportvectorswithoutanysignificant deteriorationofmodel accuracy.Next,theprunedmodelisusedinacomputationallyefficient MPCalgorithminwhichalinear approximationofthepredictedoutputtrajectoryissuccessivelyfoundon-lineandusedforprediction. The controlprofile iscalculatedon-linefromaquadraticoptimisationproblem.Itisdemonstratedthat the describedMPCalgorithmwithon-linelinearisationbasedontheprunedLS-SVMWienermodelgives practicallythesametrajectoriesasthoseobtainedinthecomputationallycomplexMPCapproachbased on thefullmodelwithon-linenonlinearoptimisationrepeatedateachsamplinginstant.
Author Maciej Ławryńczuk IAiIS
Maciej Ławryńczuk,,
- The Institute of Control and Computation Engineering
Journal seriesNeurocomputing, ISSN 0925-2312
Issue year2016
Vol2016
No205
Pages311-328
Publication size in sheets0.85
Keywords in EnglishProcess control Neutralisationreactorcontrol Model PredictiveControl Wiener models Least-SquaresSupportVectorMachines
DOIDOI:10.1016/j.neucom.2016.03.066
URL http://www.sciencedirect.com/science/article/pii/S0925231216302995
projectDevelopment of methodology of control, decision support and production management. Project leader: Zieliński Cezary, , Phone: 5102, start date 19-05-2015, end date 31-12-2016, 504/02233/1031, Completed
WEiTI Działalność statutowa
Languageen angielski
File
Lawrynczuk Neurocomp2016.pdf 1.88 MB
Score (nominal)30
ScoreMinisterial score = 30.0, 27-03-2017, ArticleFromJournal
Ministerial score (2013-2016) = 30.0, 27-03-2017, ArticleFromJournal
Publication indicators WoS Impact Factor: 2016 = 3.317 (2) - 2016=3.211 (5)
Citation count*10 (2018-07-15)
Cite
Share Share

Get link to the record
msginfo.png


* presented citation count is obtained through Internet information analysis and it is close to the number calculated by the Publish or Perish system.
Back