Multiobjective optimization using variable complexity modelling for control system design

Valceres V R Silva; Peter J. Fleming; Jungiro Sugimoto; Ryuichi Yokoyama

doi:10.1016/j.asoc.2007.02.004

Multiobjective optimization using variable complexity modelling for control system design

Valceres V R Silva^*, Peter J. Fleming, Jungiro Sugimoto, Ryuichi Yokoyama

^*Corresponding author for this work

Research output: Contribution to journal › Article › peer-review

22 Citations (Scopus)

Abstract

A multi-stage design approach that uses a multiobjective genetic algorithm as the framework for optimization and multiobjective preference articulation, and an H_infty loop-shaping technique are used to design controllers for a gas turbine engine. A non-linear model is used to assess performance of the controller. Because the computational load of applying multiobjective genetic algorithm to this control strategy is very high, a neural network and response surface models are used in order to speed up the design process within the framework of a multiobjective genetic algorithm. The final designs are checked using the original non-linear model.

Original language	English
Pages (from-to)	392-401
Number of pages	10
Journal	Applied Soft Computing Journal
Volume	8
Issue number	1
DOIs	https://doi.org/10.1016/j.asoc.2007.02.004
Publication status	Published - 2008 Jan

Keywords

H_infty control
Multiobjective genetic algorithms
Neural networks
Optimization
Variable complexity modelling

ASJC Scopus subject areas

Software
Computer Science Applications
Artificial Intelligence

Access to Document

10.1016/j.asoc.2007.02.004

Cite this

@article{4cbc12eb02494887849b5ba5f09735e1,

title = "Multiobjective optimization using variable complexity modelling for control system design",

abstract = "A multi-stage design approach that uses a multiobjective genetic algorithm as the framework for optimization and multiobjective preference articulation, and an H_infty loop-shaping technique are used to design controllers for a gas turbine engine. A non-linear model is used to assess performance of the controller. Because the computational load of applying multiobjective genetic algorithm to this control strategy is very high, a neural network and response surface models are used in order to speed up the design process within the framework of a multiobjective genetic algorithm. The final designs are checked using the original non-linear model.",

keywords = "H_infty control, Multiobjective genetic algorithms, Neural networks, Optimization, Variable complexity modelling",

author = "Silva, {Valceres V R} and Fleming, {Peter J.} and Jungiro Sugimoto and Ryuichi Yokoyama",

year = "2008",

month = jan,

doi = "10.1016/j.asoc.2007.02.004",

language = "English",

volume = "8",

pages = "392--401",

journal = "Applied Soft Computing",

issn = "1568-4946",

publisher = "Elsevier BV",

number = "1",

}

TY - JOUR

T1 - Multiobjective optimization using variable complexity modelling for control system design

AU - Silva, Valceres V R

AU - Fleming, Peter J.

AU - Sugimoto, Jungiro

AU - Yokoyama, Ryuichi

PY - 2008/1

Y1 - 2008/1

N2 - A multi-stage design approach that uses a multiobjective genetic algorithm as the framework for optimization and multiobjective preference articulation, and an H_infty loop-shaping technique are used to design controllers for a gas turbine engine. A non-linear model is used to assess performance of the controller. Because the computational load of applying multiobjective genetic algorithm to this control strategy is very high, a neural network and response surface models are used in order to speed up the design process within the framework of a multiobjective genetic algorithm. The final designs are checked using the original non-linear model.

AB - A multi-stage design approach that uses a multiobjective genetic algorithm as the framework for optimization and multiobjective preference articulation, and an H_infty loop-shaping technique are used to design controllers for a gas turbine engine. A non-linear model is used to assess performance of the controller. Because the computational load of applying multiobjective genetic algorithm to this control strategy is very high, a neural network and response surface models are used in order to speed up the design process within the framework of a multiobjective genetic algorithm. The final designs are checked using the original non-linear model.

KW - H_infty control

KW - Multiobjective genetic algorithms

KW - Neural networks

KW - Optimization

KW - Variable complexity modelling

UR - http://www.scopus.com/inward/record.url?scp=34548515057&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=34548515057&partnerID=8YFLogxK

U2 - 10.1016/j.asoc.2007.02.004

DO - 10.1016/j.asoc.2007.02.004

M3 - Article

AN - SCOPUS:34548515057

SN - 1568-4946

VL - 8

SP - 392

EP - 401

JO - Applied Soft Computing

JF - Applied Soft Computing

IS - 1

ER -

Multiobjective optimization using variable complexity modelling for control system design

Abstract

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this