Abstract
Quality function deployment (QFD) has been widely recognised as an effective means to develop quality products that can maximise customer satisfactions. Traditionally, fuzzy numbers and fuzzy arithmetic operations are employed to quantify and analyse the vague design information in QFD. However, besides the difficulties in determining an appropriate fuzzy membership function, the rapid enlargement of fuzzy boundary intervals due to fuzzy arithmetic operations is another important issue, which leads to low discerniblility of design objectives. In this work, two novel concepts termed rough number and rough boundary interval are proposed based on rough set theory and are integrated with interval arithmetic to formulate a novel approach to QFD analysis. A comparison between using rough numbers and fuzzy numbers in QFD analysis is conducted through a case study. The results show that the proposed rough number enabled QFD analysis is able to provide more insights into the design information and can significantly suppress the enlargement of boundary intervals, compared to the traditional method using fuzzy numbers.