چکیده:
In most manufacturing processes, each product may contain a variety of quality characteristics which are of the interest to be optimized simultaneously through determination of the optimum setting of controllable factors. Although, classic experimental design presents some solutions for this regard, in a fuzzy environment, and in cases where the response data follow non-normal distributions, the available methods do not apply any more. In this paper, a general framework is introduced in which NORTA inverse transformation technique and fuzzy goal programming are used to deal with non-normality distribution of the response data and the fuzziness of response targets respectively. Moreover, the presented framework uses a simulation approach to investigate the effectiveness of the determined setting of controllable factors obtained from statistical analysis, for optimization of sink mark index, deflection rate and volumetric shrinkage in plastic molding manufacturing processes. The accuracy of the proposed method is verified through a real case study.
خلاصه ماشینی:
"com A Fuzzy Goal Programming Approach for Optimizing Non-normal Fuzzy Multiple Response Problems Hamed Mogouiea, Amir Farshbaf-Geranmayeh*b, Amirhossein Amiric and Mahdi Bashiric a Department of Industrial and Systems Engineering, Isfahan University of Technology, Isfahan, Iran b School of Industrial Engineering, College of Engineering, University of Tehran, Tehran, Iran c Industrial Engineering Department, Faculty of Engineering, Shahed University, Tehran, Iran Abstract In most manufacturing processes, each product may contain a variety of quality characteristics which are of the interest to be optimized simultaneously through determination of the optimum setting of controllable factors.
Although, classic experimental design presents some solutions for this regard, in a fuzzy environment, and in cases where the response data follow non-normal distributions, the available methods do not apply any more.
Moreover, the presented framework uses a simulation approach to investigate the effectiveness of the determined setting of controllable factors obtained from statistical analysis, for optimization of sink mark index, deflection rate and volumetric shrinkage in plastic molding manufacturing processes.
Fuzzy multi-objective modeling of the problem After transforming the initial responses into normally distributed ones in the previous step, it is possible to perform a regression analysis for each of these responses separately.
5. Conclusions In this paper a new approach was proposed for a multi response optimization problem where the target values of responses were fuzzy numbers, and some responses followed non-normal distribution.
For future studies, it is recommended to apply other regression models for mixed continuous – discrete responses and to use the obtained regression functions as the objectives directly without the need to transform the non-normally distributed data."