Taguchi-aided search method for design optimization of engineering systems

K'uang J. Ku, Singiresu S Rao, Li Chen

Research output: Contribution to journalArticle

60 Citations (Scopus)

Abstract

The Taguchi method emphasizes the design of a robust product that is insensitive to disturbances (noises). This feature of the Taguchi method is useful in mechanical design since an optimal design obtained through the application of the Taguchi method will also be inherently robust. The Taguchi concept of design of experiments (DOE) is used to solve unconstrained and constrained optimization problems. The signal-to-noise ratio used in the conventional Taguchi method uses only positive function values and hence is not applicable to design optimization directly. In this work, a modified form of signal-to-noise ratio is proposed so that both positive and negative objective function values can be handled. In the case of unconstrained optimization, the search directions are determined by using the Taguchi philosophy. The partial factorial design of experiments provides an efficient way of determining a good search direction. The procedure is extended to constrained optimization problems in the presence of both single and multiple objective functions. Detailed solution procedures are presented along with three numerical examples for illustration.

Original languageEnglish
Pages (from-to)1-23
Number of pages23
JournalEngineering Optimization
Volume30
Issue number1
StatePublished - Dec 1 1998
Externally publishedYes

Fingerprint

Taguchi Method
Taguchi methods
Systems Engineering
Systems engineering
Search Methods
Design of Experiments
Constrained optimization
Unconstrained Optimization
Constrained Optimization Problem
Design of experiments
Signal to noise ratio
Objective function
Mechanical Design
Factorial Design
Multiple Objectives
Value Function
Disturbance
Partial
Numerical Examples
Design optimization

Keywords

  • Nonlinear optimization
  • Robust design
  • Taguchi method

ASJC Scopus subject areas

  • Management Science and Operations Research
  • Engineering (miscellaneous)

Cite this

Taguchi-aided search method for design optimization of engineering systems. / Ku, K'uang J.; Rao, Singiresu S; Chen, Li.

In: Engineering Optimization, Vol. 30, No. 1, 01.12.1998, p. 1-23.

Research output: Contribution to journalArticle

@article{b211a258417a4cdd8c9f39adc1dc8ed9,
title = "Taguchi-aided search method for design optimization of engineering systems",
abstract = "The Taguchi method emphasizes the design of a robust product that is insensitive to disturbances (noises). This feature of the Taguchi method is useful in mechanical design since an optimal design obtained through the application of the Taguchi method will also be inherently robust. The Taguchi concept of design of experiments (DOE) is used to solve unconstrained and constrained optimization problems. The signal-to-noise ratio used in the conventional Taguchi method uses only positive function values and hence is not applicable to design optimization directly. In this work, a modified form of signal-to-noise ratio is proposed so that both positive and negative objective function values can be handled. In the case of unconstrained optimization, the search directions are determined by using the Taguchi philosophy. The partial factorial design of experiments provides an efficient way of determining a good search direction. The procedure is extended to constrained optimization problems in the presence of both single and multiple objective functions. Detailed solution procedures are presented along with three numerical examples for illustration.",
keywords = "Nonlinear optimization, Robust design, Taguchi method",
author = "Ku, {K'uang J.} and Rao, {Singiresu S} and Li Chen",
year = "1998",
month = "12",
day = "1",
language = "English",
volume = "30",
pages = "1--23",
journal = "Engineering Optimization",
issn = "0305-215X",
publisher = "Taylor and Francis Ltd.",
number = "1",

}

TY - JOUR

T1 - Taguchi-aided search method for design optimization of engineering systems

AU - Ku, K'uang J.

AU - Rao, Singiresu S

AU - Chen, Li

PY - 1998/12/1

Y1 - 1998/12/1

N2 - The Taguchi method emphasizes the design of a robust product that is insensitive to disturbances (noises). This feature of the Taguchi method is useful in mechanical design since an optimal design obtained through the application of the Taguchi method will also be inherently robust. The Taguchi concept of design of experiments (DOE) is used to solve unconstrained and constrained optimization problems. The signal-to-noise ratio used in the conventional Taguchi method uses only positive function values and hence is not applicable to design optimization directly. In this work, a modified form of signal-to-noise ratio is proposed so that both positive and negative objective function values can be handled. In the case of unconstrained optimization, the search directions are determined by using the Taguchi philosophy. The partial factorial design of experiments provides an efficient way of determining a good search direction. The procedure is extended to constrained optimization problems in the presence of both single and multiple objective functions. Detailed solution procedures are presented along with three numerical examples for illustration.

AB - The Taguchi method emphasizes the design of a robust product that is insensitive to disturbances (noises). This feature of the Taguchi method is useful in mechanical design since an optimal design obtained through the application of the Taguchi method will also be inherently robust. The Taguchi concept of design of experiments (DOE) is used to solve unconstrained and constrained optimization problems. The signal-to-noise ratio used in the conventional Taguchi method uses only positive function values and hence is not applicable to design optimization directly. In this work, a modified form of signal-to-noise ratio is proposed so that both positive and negative objective function values can be handled. In the case of unconstrained optimization, the search directions are determined by using the Taguchi philosophy. The partial factorial design of experiments provides an efficient way of determining a good search direction. The procedure is extended to constrained optimization problems in the presence of both single and multiple objective functions. Detailed solution procedures are presented along with three numerical examples for illustration.

KW - Nonlinear optimization

KW - Robust design

KW - Taguchi method

UR - http://www.scopus.com/inward/record.url?scp=0031677271&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0031677271&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:0031677271

VL - 30

SP - 1

EP - 23

JO - Engineering Optimization

JF - Engineering Optimization

SN - 0305-215X

IS - 1

ER -