A new mathematical program based on Principal Component Analysis for multiple response optimization

Author

Shahrabi, J. ; Hejazi, T.H.

Author_Institution

Dept. of Ind. Eng. & Manage. Syst., Amirkabir Univ. of Tehran, Tehran, Iran

fYear

2011

fDate

14-17 Sept. 2011

Firstpage

445

Lastpage

449

Abstract

In many industrial systems, improvement of quality characteristics is a main step in product design. Design of experiments is a well-grounded approach to optimize quality characteristics (Response variables) in terms of design parameters (Factors). The aim of this study is to develop the mathematical model for continuous optimization and analysis of the multiple response surfaces whereas the responses are correlated. This study proposes the novel mathematical model based on PCA and Lp-metric function for continuous pareto optimization of MRS problem when the responses are correlated and might have different importance and dimension.

Keywords

Pareto optimisation; mathematical programming; principal component analysis; product design; quality control; Lp-metric function; MRS problem; PCA; continuous Pareto optimization; design of experiments; design parameters; industrial systems; mathematical program; multiple-response optimization; multiple-response surface analysis; principal component analysis; product design; quality characteristic improvement; response variables; Correlation; Equations; Mathematical model; Optimization; Principal component analysis; Reliability; Response surface methodology; Correlated responses; Mathematical programming; Multiple Response Surface (MRS); Principal Component Analysis (PCA);

fLanguage

English

Publisher

ieee

Conference_Titel

Quality and Reliability (ICQR), 2011 IEEE International Conference on

Conference_Location

Bangkok

Print_ISBN

978-1-4577-0626-4

Type

conf

DOI

10.1109/ICQR.2011.6031758

Filename

6031758