A fuzzy goal programming-regression approach to optimize process performance of multiple responses under uncertainty. Issue 1 (2nd January 2019)

• Record Type: Journal Article
• Authors: Al-Refaie, Abbas; Bani Domi, Ghaith; Abdullah, Rasha
• Is Part Of: International journal of management science and engineering management. Volume 14:Issue 1(2019)
• Page Start: 20
• Page End: 32
• Publication Date: 2019-01-02
• Subjects: Fuzzy regression -- Fuzzy goal programming -- desirability function -- optimization; Management science -- Periodicals; Engineering -- Management -- Periodicals; Engineering -- Management; Management science; Periodicals; 658.005
• DOI: 10.1080/17509653.2018.1467802
• Languages: English
• ISSNs: 1750-9661

Abstract

ABSTRACT: This paper proposes an approach to optimize performance of manufacturing processes for multiple quality responses using a fuzzy goal programming-regression approach. Firstly, a multiple nonlinear regression model is formulated for each response replicate and then utilized in fuzzy goal programing to obtain the combinations of optimal factor levels. Fuzzy regression model is then developed for each quality response. In order to determine the combination of optimal factor settings for multiple responses, the desirability functions and pay-off matrices are constructed and then employed to formulate the lower, middle, and upper optimization models. Three case studies, which were investigated in previous literature, are employed for illustration. Compared to the Taguchi method, artificial neural networks, fuzzy regression, and grey-Taguchi method, the proposed approach provides larger anticipated improvement in quality responses, efficiently deals with fuzziness and irregular process performance, and effectively considers preferences on process settings and response values in all the three case studies. Finally, a confirmation experiments in plastic pipes industry to validate the obtained results. In conclusion, the proposed approach may provide valuable support to process engineers in optimizing process performance of multiple quality responses under uncertainty in a wide range of industrial applications.