A comparison of various optimization algorithms on a multilevel problem
A comparison of various optimization algorithms on a multilevel problem
In many problems in science and engineering, there are often a number of computational models that can be used to simulate the problem at hand. Models of physical systems can differ according to computational cost, accuracy and precision. This paper presents the concept of multilevel optimization, where different models of the problem are used in combination. This initial study compares several strategies for combining fast evaluations of limited accuracy with a few accurate calculations. It also attempts to show how different optimizers work under these different combination strategies. A specially designed test function is used to carry out these comparisons. Of the proposed strategies and optimisers, a sequential mixing strategy applied to a genetic algorithm with clustering gives the best results. This paper highlights the need to develop specialized optimization algorithms for this kind of problem.
optimization, multilevel problems, genetic algorithms
639-654
El-Beltagy, M.A
820b78f8-f595-4508-bb65-9ac44db711db
Keane, A.J.
26d7fa33-5415-4910-89d8-fb3620413def
1999
El-Beltagy, M.A
820b78f8-f595-4508-bb65-9ac44db711db
Keane, A.J.
26d7fa33-5415-4910-89d8-fb3620413def
El-Beltagy, M.A and Keane, A.J.
(1999)
A comparison of various optimization algorithms on a multilevel problem.
Engineering Applications of Artificial Intelligence, 12 (5), .
(doi:10.1016/S0952-1976(99)00033-0).
Abstract
In many problems in science and engineering, there are often a number of computational models that can be used to simulate the problem at hand. Models of physical systems can differ according to computational cost, accuracy and precision. This paper presents the concept of multilevel optimization, where different models of the problem are used in combination. This initial study compares several strategies for combining fast evaluations of limited accuracy with a few accurate calculations. It also attempts to show how different optimizers work under these different combination strategies. A specially designed test function is used to carry out these comparisons. Of the proposed strategies and optimisers, a sequential mixing strategy applied to a genetic algorithm with clustering gives the best results. This paper highlights the need to develop specialized optimization algorithms for this kind of problem.
Text
elbe_99b.pdf
- Accepted Manuscript
More information
Published date: 1999
Keywords:
optimization, multilevel problems, genetic algorithms
Identifiers
Local EPrints ID: 23618
URI: http://eprints.soton.ac.uk/id/eprint/23618
ISSN: 0952-1976
PURE UUID: 55c4ae3f-b766-411f-a26c-0d58a09ffbbb
Catalogue record
Date deposited: 01 Feb 2007
Last modified: 16 Mar 2024 02:53
Export record
Altmetrics
Contributors
Author:
M.A El-Beltagy
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics
