期刊名称：International Journal of Industrial Engineering Computations
印刷版ISSN：1923-2926
电子版ISSN：1923-2934
出版年度：2019
卷号：10
期号：1
页码：133-148
DOI：10.5267/j.ijiec.2018.2.001
语种：English
出版社：Growing Science Publishing Company
摘要：Designing a production process normally is involved with some important constraints such as uncertainty, trade-off between production costs and quality, customer’s expectations and production tolerances. In this paper, a novel multi-objective robust optimization model is introduced to investigate the best levels of design variables. The primary objective is to minimize the production cost while increasing robustness and performance. The response surface methodology is utilized as a common approximation model to fit the relationship between responses and design variables in the worst-case scenario of uncertainties. The target mean ratio α is applied to ensure the quality of the process by providing the robustness for all types of quality characteristics and with a trade-off between variability and deviance from the ideal point. The Lp metric method is used to integrate all objectives in one overall function. In order to estimate target value of the quality loss by considering production tolerances, the process capability ratio (Cpm) is applied. At the end, a numerical chemical mixture problem is served to show the applicability of the proposed method.
关键词：Robust design; Loss function; Uncertainty; Response surface methodology; Process optimization
