Optimization using higher order approximations and parallel processing

Charles E. Seeley, Aditi Chattopadhyay

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

the time associated with solving realistic optimization problems is greatly reduced by employing sequential approximate optimization techniques. First order techniques are most commonly employed since conventional wisdom suggests that too many exact function evaluations are necessary to justify higher order techniques. Parallel processing allows several computations to be performed simultaneously and makes the use of higher order approaches more attractive. In this paper, the efficiency of both first and second order approximations is investigated. Parallel processing concepts, which could be implemented on a cluster of workstations, are discussed. A test problem, based on frequency spacing of a composite laminate, is presented. Results indicate that significant time savings can be obtained using a combined sequential approximate optimization approach which utilizes benefits of both first and second order techniques and is implemented using parallel processing.

Original languageEnglish (US)
Title of host publicationCollection of Technical Papers - AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics and Materials Conference
Editors Anon
PublisherAIAA
Pages449-462
Number of pages14
Volume1
StatePublished - 1998
EventProceedings of the 1998 39th AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics, and Materials Conference and Exhibit and AIAA/ASME/AHS Adaptive Structures Forum. Part 1 (of 4) - Long Beach, CA, USA
Duration: Apr 20 1998Apr 23 1998

Other

OtherProceedings of the 1998 39th AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics, and Materials Conference and Exhibit and AIAA/ASME/AHS Adaptive Structures Forum. Part 1 (of 4)
CityLong Beach, CA, USA
Period4/20/984/23/98

Fingerprint

Processing
Function evaluation
Laminates
Composite materials

ASJC Scopus subject areas

  • Architecture

Cite this

Seeley, C. E., & Chattopadhyay, A. (1998). Optimization using higher order approximations and parallel processing. In Anon (Ed.), Collection of Technical Papers - AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics and Materials Conference (Vol. 1, pp. 449-462). AIAA.

Optimization using higher order approximations and parallel processing. / Seeley, Charles E.; Chattopadhyay, Aditi.

Collection of Technical Papers - AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics and Materials Conference. ed. / Anon. Vol. 1 AIAA, 1998. p. 449-462.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Seeley, CE & Chattopadhyay, A 1998, Optimization using higher order approximations and parallel processing. in Anon (ed.), Collection of Technical Papers - AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics and Materials Conference. vol. 1, AIAA, pp. 449-462, Proceedings of the 1998 39th AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics, and Materials Conference and Exhibit and AIAA/ASME/AHS Adaptive Structures Forum. Part 1 (of 4), Long Beach, CA, USA, 4/20/98.
Seeley CE, Chattopadhyay A. Optimization using higher order approximations and parallel processing. In Anon, editor, Collection of Technical Papers - AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics and Materials Conference. Vol. 1. AIAA. 1998. p. 449-462
Seeley, Charles E. ; Chattopadhyay, Aditi. / Optimization using higher order approximations and parallel processing. Collection of Technical Papers - AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics and Materials Conference. editor / Anon. Vol. 1 AIAA, 1998. pp. 449-462
@inproceedings{5f1cc2e19e6541a5bbc83a72a1a57250,
title = "Optimization using higher order approximations and parallel processing",
abstract = "the time associated with solving realistic optimization problems is greatly reduced by employing sequential approximate optimization techniques. First order techniques are most commonly employed since conventional wisdom suggests that too many exact function evaluations are necessary to justify higher order techniques. Parallel processing allows several computations to be performed simultaneously and makes the use of higher order approaches more attractive. In this paper, the efficiency of both first and second order approximations is investigated. Parallel processing concepts, which could be implemented on a cluster of workstations, are discussed. A test problem, based on frequency spacing of a composite laminate, is presented. Results indicate that significant time savings can be obtained using a combined sequential approximate optimization approach which utilizes benefits of both first and second order techniques and is implemented using parallel processing.",
author = "Seeley, {Charles E.} and Aditi Chattopadhyay",
year = "1998",
language = "English (US)",
volume = "1",
pages = "449--462",
editor = "Anon",
booktitle = "Collection of Technical Papers - AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics and Materials Conference",
publisher = "AIAA",

}

TY - GEN

T1 - Optimization using higher order approximations and parallel processing

AU - Seeley, Charles E.

AU - Chattopadhyay, Aditi

PY - 1998

Y1 - 1998

N2 - the time associated with solving realistic optimization problems is greatly reduced by employing sequential approximate optimization techniques. First order techniques are most commonly employed since conventional wisdom suggests that too many exact function evaluations are necessary to justify higher order techniques. Parallel processing allows several computations to be performed simultaneously and makes the use of higher order approaches more attractive. In this paper, the efficiency of both first and second order approximations is investigated. Parallel processing concepts, which could be implemented on a cluster of workstations, are discussed. A test problem, based on frequency spacing of a composite laminate, is presented. Results indicate that significant time savings can be obtained using a combined sequential approximate optimization approach which utilizes benefits of both first and second order techniques and is implemented using parallel processing.

AB - the time associated with solving realistic optimization problems is greatly reduced by employing sequential approximate optimization techniques. First order techniques are most commonly employed since conventional wisdom suggests that too many exact function evaluations are necessary to justify higher order techniques. Parallel processing allows several computations to be performed simultaneously and makes the use of higher order approaches more attractive. In this paper, the efficiency of both first and second order approximations is investigated. Parallel processing concepts, which could be implemented on a cluster of workstations, are discussed. A test problem, based on frequency spacing of a composite laminate, is presented. Results indicate that significant time savings can be obtained using a combined sequential approximate optimization approach which utilizes benefits of both first and second order techniques and is implemented using parallel processing.

UR - http://www.scopus.com/inward/record.url?scp=0031707952&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0031707952&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:0031707952

VL - 1

SP - 449

EP - 462

BT - Collection of Technical Papers - AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics and Materials Conference

A2 - Anon, null

PB - AIAA

ER -