A hybrid genetic algorithm for mixed-discrete design optimization

Singiresu S Rao, Ying Xiong

Research output: Contribution to journalArticle

58 Citations (Scopus)

Abstract

A new hybrid genetic algorithm is presented for the solution of mixed-discrete nonlinear design optimization. In this approach, the genetic algorithm (GA) is used mainly to determine the optimal feasible region that contains the global optimum point, and the hybrid negative subgradient method integrated with discrete one-dimensional search is subsequently used to replace the GA to find the final optimum solution. The hybrid genetic algorithm, combining the advantages of random search and deterministic search methods, can improve the convergence speed and computational efficiency compared with some other GAs or random search methods. Several practical examples of mechanical design are tested using the computer program developed. The numerical results demonstrate the effectiveness and robustness of the proposed approach.

Original languageEnglish
Pages (from-to)1100-1112
Number of pages13
JournalJournal of Mechanical Design, Transactions Of the ASME
Volume127
Issue number6
DOIs
StatePublished - Nov 1 2005

Fingerprint

Genetic algorithms
Computational efficiency
Computer program listings
Design optimization

ASJC Scopus subject areas

  • Mechanical Engineering

Cite this

A hybrid genetic algorithm for mixed-discrete design optimization. / Rao, Singiresu S; Xiong, Ying.

In: Journal of Mechanical Design, Transactions Of the ASME, Vol. 127, No. 6, 01.11.2005, p. 1100-1112.

Research output: Contribution to journalArticle

@article{0ef9811919074e28b28745b3f0009fed,
title = "A hybrid genetic algorithm for mixed-discrete design optimization",
abstract = "A new hybrid genetic algorithm is presented for the solution of mixed-discrete nonlinear design optimization. In this approach, the genetic algorithm (GA) is used mainly to determine the optimal feasible region that contains the global optimum point, and the hybrid negative subgradient method integrated with discrete one-dimensional search is subsequently used to replace the GA to find the final optimum solution. The hybrid genetic algorithm, combining the advantages of random search and deterministic search methods, can improve the convergence speed and computational efficiency compared with some other GAs or random search methods. Several practical examples of mechanical design are tested using the computer program developed. The numerical results demonstrate the effectiveness and robustness of the proposed approach.",
author = "Rao, {Singiresu S} and Ying Xiong",
year = "2005",
month = "11",
day = "1",
doi = "10.1115/1.1876436",
language = "English",
volume = "127",
pages = "1100--1112",
journal = "Journal of Mechanical Design - Transactions of the ASME",
issn = "1050-0472",
publisher = "American Society of Mechanical Engineers(ASME)",
number = "6",

}

TY - JOUR

T1 - A hybrid genetic algorithm for mixed-discrete design optimization

AU - Rao, Singiresu S

AU - Xiong, Ying

PY - 2005/11/1

Y1 - 2005/11/1

N2 - A new hybrid genetic algorithm is presented for the solution of mixed-discrete nonlinear design optimization. In this approach, the genetic algorithm (GA) is used mainly to determine the optimal feasible region that contains the global optimum point, and the hybrid negative subgradient method integrated with discrete one-dimensional search is subsequently used to replace the GA to find the final optimum solution. The hybrid genetic algorithm, combining the advantages of random search and deterministic search methods, can improve the convergence speed and computational efficiency compared with some other GAs or random search methods. Several practical examples of mechanical design are tested using the computer program developed. The numerical results demonstrate the effectiveness and robustness of the proposed approach.

AB - A new hybrid genetic algorithm is presented for the solution of mixed-discrete nonlinear design optimization. In this approach, the genetic algorithm (GA) is used mainly to determine the optimal feasible region that contains the global optimum point, and the hybrid negative subgradient method integrated with discrete one-dimensional search is subsequently used to replace the GA to find the final optimum solution. The hybrid genetic algorithm, combining the advantages of random search and deterministic search methods, can improve the convergence speed and computational efficiency compared with some other GAs or random search methods. Several practical examples of mechanical design are tested using the computer program developed. The numerical results demonstrate the effectiveness and robustness of the proposed approach.

UR - http://www.scopus.com/inward/record.url?scp=28244439808&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=28244439808&partnerID=8YFLogxK

U2 - 10.1115/1.1876436

DO - 10.1115/1.1876436

M3 - Article

AN - SCOPUS:28244439808

VL - 127

SP - 1100

EP - 1112

JO - Journal of Mechanical Design - Transactions of the ASME

JF - Journal of Mechanical Design - Transactions of the ASME

SN - 1050-0472

IS - 6

ER -