Determining Importance Weights for Engineering Characteristics in QFD Considering the Correlations Under Uncertainties

Abstract

Quality function deployment (QFD) is a systematic tool for translating customer requirements (CRs) into feasible engineering characteristics (ECs) of a product. In QFD, it is crucial to derive the importance of ECs. Owing to the typical vagueness or ambiguity in the product development process, this paper applies the uncertainty theory to QFD, and an uncertain expected value method for getting the importance of ECs and ranking them is presented, in which linguistic inputs are expressed as uncertain variables. For convenience sake, we also give a general formula about the expected value of the product of two linear uncertain variables, as well as three linear uncertain variables. Furthermore, in getting the importance of ECs, we consider not only the relative weights of CRs and the relationships between CRs and ECs, but also the correlations among ECs. Finally, an example of a digital camera improvement illustrates the performance of the proposed approach, and the experimental results are compared with the situation that ignores the correlations among ECs.