Evaluation of multi-attribute decision making systems applied during the concept design of new microplasma devices
详细信息 查看全文
- 作者：Elizabeth Lennon ; John Farr ; Ronald Besser
- 关键词：MADM ; multi-attribute decision making ; Pugh ; Pugh&rsquo ; s controlled convergence method ; QFD ; Quality Function Deployment ; UBAF ; Utility Based Axiomatic Framework ; AHP ; Analytic Hierarchy Process ; GRA ; Grey Relational Analysis
- 刊名：Expert Systems with Applications
- 出版年：2013
- 出版时间：15 November, 2013
- 年：2013
- 卷：40
- 期：16
- 页码：6321-6329
- 全文大小：1093 K
文摘
Various multi-attribute decision making (MADM) systems can be implemented to narrow a field of new concept designs down to those with high likelihoods of surpassing state-of-the-art technologies. This research investigated the conceptual design phase of new microplasma devices in order to create metrics that evaluate the efficiency, effectiveness, and overall utility of representative MADM systems studied in previous engineering design applications. Device attributes and concept alternatives for the microplasma devices were identified from open-ended expert surveys. Efficiency metrics were defined based on the number of manual user inputs. Published device literature and testing were used to gauge how closely device concepts satisfied multi-attribute criteria, forming the basis of an effectiveness metric. A weighted average of the efficiency and effectiveness defined a MADM system¡¯s overall utility. Varying the effectiveness weight provided further insight into the conditions under which particular MADM approaches exhibited higher utility values. The MADM systems found to possess the highest overall quantified utilities were based on Pugh¡¯s controlled convergence, Utility Based Axiomatic Framework, and Grey Relational Analysis. The MADM method with the lowest overall utility was the analytical hierarchy process. These findings indicate that consensus building and utility-based MADM systems are especially helpful to engineering design teams during the early design phases of novel technologies when resources are constrained or historical data is limited.