Conceptual system design evaluation: Rating and ranking versus compliance analysis

The conceptual design phase is a critical part of the systems engineering approach, in terms of commitment of both time and money. Structure and discipline at this time are fundamental to the efficient and effective development of competitive products and systems. However, the extent and resolution of information available during this phase, in the form of parameter values and priorities, is typically characterized by imprecision and subjectivity. Such uncertainty can be represented by fuzzy variables and linguistic values. Discrimination between competing product concepts thus necessitates identification of a defuzzification technique to allow comparison of imprecise requirements with imprecise expected values for conceptual alternatives. Since fuzzy numbers do not exhibit complete order, many existing defuzzification techniques are based on the philosophy of rating and ranking fuzzy sets. This paper argues that a defuzzification technique based on the degree of compliance between required and anticipated parameter values is more suited to concept evaluation and selection. The advantage of such an approach is demonstrated and a fuzzy logic based framework for this purpose is proposed. Finally, techniques for accelerating such a compliance analysis are suggested. © 2004 Wiley Periodicals, Inc. Syst Eng 7: 338–351, 2004