Multi-objective optimization-based TOPSIS method for sustainable product design under epistemic uncertainty
Jing Zhou,Tangfan Xiahou,Yu Liu
DOI: https://doi.org/10.1016/j.asoc.2020.106850
IF: 8.7
2021-01-01
Applied Soft Computing
Abstract:<p>Sustainable product design has captured considerable attention over recent years due to the growing customer demands of sustainability. To improve the environmental performance of products at the early stage of product design, a variety of economic, social, and environmental factors, such as manufacturing cost and time, product yield, capacity, customer preferences, and pollutant emissions, have to be taken into account jointly. However, due to the lack of knowledge and ambiguity of customers and experts, some of these factors may contain epistemic uncertainties, overlooking them may lead to an infeasible design. To fill the gap, we propose a new multi-objective optimization-based technique for order preference by similarity to ideal solution (TOPSIS) method to facilitate sustainable product design under epistemic uncertainty. In the proposed method, we develop a fuzzy Mahalanobis-Taguchi system method to address the epistemic uncertainty of customer preferences on optimization objectives. Meanwhile, we introduce the <em>Me</em> measure to manipulate the epistemic uncertainty of experts' judgments on process parameters and variables during the manufacturing process. Subsequently, we implement the new TOPSIS method to obtain the optimal design scheme. We provide an example of sustainable substrate design, along with sensitivity analysis scenarios and comparative studies, to elaborate on the performance of the proposed method.</p>
computer science, artificial intelligence, interdisciplinary applications