Fuzzing Ratings for Multiattribute Design Decision-Making

Abstract

Early in the design process, problems can arise when information is incomplete and goals are not known precisely. When preliminary design evaluation is approached as a multiattribute decision-making problem, both the levels of attributes and their relative importance can be treated as fuzzy numbers elicited from the designer. However, information regarding estimated attribute levels might be lost in limiting the designer to the standard universe of discourse. Another problem is that the attribute weights might be difficult for the designer to determine. A methodology is demonstrated for ranking alternatives based on the fuzzy distance from a fuzzy goal. The concept of a fuzzy line segment is introduced in order to make the universe of discourse continuous, thus not restricting the designer to a small set of fuzzy inputs. The fuzzy line segment makes it possible to more closely reflect the designer’s estimates of performance of design alternatives and the relative weight assigned to each attribute. It facilitates more accurate and precise linguistic input, and also provides a way to “fuzzify” numeric input. As a result, Saaty’s Analytic Hierarchy Process (AHP) can be employed to assist the designer in more accurately determining attribute weights.

Application of new fuzzy-weighted average (NFWA) method to engineering design evaluation

“…In Hsiao (1998), both weights of objectives and performance levels are quanti-® ed with the membership functions of fuzzy sets. However, contrary to Thurston and Carnahan (1992) and Carnahan et al (1994) in which the overall desirability of an 1149 NFWA method applied to engineering design evaluation alternative is fuzzy (a fuzzy number), representing its imprecision, the scores assigned during the evaluation process are crisp numbers (degrees of membership of fuzzy numbers). Therefore, this approach is less powerful and credible, as the results do not represent the imprecision of the judgements.…”

“…In the former, the alternative with the highest fuzzy rating is selected as the`best' ; in the latter, the alternative selected is which is closest to a fuzzy goal. Others, such as Carnahan et al (1994), Chen (1996), Khoo and Ho (1996), Zhou (1997Zhou ( , 1998 and Wang (1999), have followed a similar procedure in engineering design evaluation and quality function deployment (QFD). In these applications, fuzzy numbers characterizes the linguistic terms, and the overall desirability is calculated by the FWA without the extended division.…”

Design concept evaluation in product development using rough sets and grey relation analysis

“…In this respect, Thurston and Carnahan (1992) stated that fuzzy set theory could be applied to the early stage of preliminary design evaluation under multiple attributes or in situations based on semantic assessment of relative performance levels. Many researchers (Carnahan, Thurston, & Liu, 1994;Khoo & Ho, 1996;Sun, Kalenchuk, Xue, & Gu, 2000;Tsai & Hsiao, 2004;Wang, 2001) have successfully used fuzzy sets in engineering design evaluation. More recently, Huang, Bo, and Chen (2006) had integrated fuzzy sets with genetic algorithms and neural networks to formulate an integrated approach for design concept generation and evaluation.…”

A new integrated design concept evaluation approach based on vague sets

“…Thurston and Carnahan (1992) indicated that fuzzy set theory could be applied in design concept evaluation in situations based on linguistic judgments. Many researchers (Carnahan, Thurston, & Liu, 1994;Khoo & Ho, 1996;Sun, Kalenchuk, Xue, & Gu, 2000;Tsai & Hsiao, 2004) have successfully applied fuzzy sets in design concept evaluation. Huang, Bo, and Chen (2006) proposed an integrated approach which combined fuzzy sets with genetic algorithms and neural networks for supporting design concept evaluation, but this approach lacked efficiency in the real world due to its complex algorithm structures and lengthy training process.…”