A Fuzzy Modelling Approach to Robust Design via Loss Functions

Abstract

Especially in a world where industrial development is reinforced by globalization tendencies, competitive companies know that satisfying customers' needs and running a successful operation requires a process that is reliable, predictable and robust. Therefore, many of quality improvement techniques focus on reducing process variation in line with the “loss to society” concept. The upside-down normal loss function is a weighted loss function that has the ability to evaluate losses with a more reasonable risk assessment. In this study, we introduce a fuzzy modelling approach based on expected upside-down normal loss function where the mean and standard deviation responses are fitted by response surface models. The proposed method aims to identify a set of operating conditions to maximize the degree of satisfaction with respect to the expected loss. Additionally, the proposed approach provides a more informative and realistic approach for comparing competing sets of conditions depending upon how much better or worse a process is. We demonstrate the proposed approach in a well-known design of experiment by comparing it with existing methods.

Keywords

• fuzzy modeling,response surface methodology,robust parameter design,upside-down normal loss function

References

1. G.E.P. Box and K.B. Wilson, On the Experimental Attainment of Optimum Conditions, J. Roy. Statist. Soc. Ser. B Metho. 13 (1951) 1-45.
2. G.G. Vining and R.H. Myers, Combining Taguchi and Response Surface Philosophies: A Dual Response Approach, J. Qual. Technol. 22 (1990) 38-45.
3. E. Del Castillo and D.C. Montgomery, A Nonlinear Programming Solution to the Dual Response Problem, J. Qual. Technol. 25 (1993) 199-204.
4. D.K.J. Lin and W. Tu, Dual Response Surface Optimization, J. Qual. Technol. 27 (1995) 34-39.
5. K. Kim and D.K.J. Lin, Dual Response Surface Optimization: A Fuzzy Modeling Approach, J. Qual. Technol. 30 (1998) 1-10.
6. A.C. Shoemaker, K.L, Tsui, and C.F.J. Wu, Economical Experimentation Methods for Robust Parameter Design, Technometrics. 33 (1991) 415-427.
7. K.A. Copeland and P.R. Nelson, Dual Response Optimization via Direct Function Minimization, J. Qual. Technol. 28 (1996) 331-336.
8. J.M. Lucas, How to Achieve a Robust Process Using Response Surface Methodology, J. Qual. Technol. 26 (1994). 248-260.

9. O. Köksoy and N. Doganaksoy, Joint Optimization of Mean and Standard Deviation in Response Surface Experimentation, J. Qual. Technol. 35 (2003) 239-252.
10. O. Köksoy, Multiresponse Robust Design: Mean Square Error (MSE criterion), Appl. Math. Comput. 175 (2006) 1716-1729.
11. D.C. Drain and A.M. Gough, Applications of the Upside-down Normal Loss Function, IEEE T. Semıconduct M. 9 (1996) 143–145.
12. O. Köksoy and S.S. Fan, An Upside-down Normal Loss Function-based Method for Quality Improvement, Eng. Optimiz. 44 (2012) 935-945.
13. M. Zeybek and O. Köksoy, Optimization of Correlated Multi-response Quality Engineering by the Upside-down Normal Loss Function, Eng. Optimiz. 48 (2016) 1419-1431.
14. H.J. Zimmermann, Fuzzy Sets, Decision Making, and Expert Systems, Kluwer Academic Publishers, Boston, M.A. (1987).
15. C.W. Kirkwood, Strategic Decision Making, Duxbury Press, Belmont, C.A. (1996).
16. H. Moskowitz and K. Kim, On Assessing the H Value in Fuzzy Linear Regression, Fuzzy Sets Syst. 58 (1993) 303-327.
17. G.E.P. Box and N.R. Draper, Empirical Model Building and Response Surfaces, John&Sons, New York, N.Y. (1987).