An optimization method for handling incomplete and conflicting opinions in quality function deployment based on consistency and consensus reaching process

Abstract

Quality Function Deployment (QFD) is a highly regarded customer-oriented quality management tool, and its key step is to obtain the importance ratings of Customer Requirements (CRs) from customer opinions. However, customers usually provide incomplete and conflicting opinions due to cognitive limitations and differences in backgrounds. To this end, this study integrates an optimization-based consensus reaching process into QFD for dealing with conflicting customer opinions, where customer opinions are modeled with incomplete linguistic distribution assessments. First, a consistency-driven optimization model is designed to convert incomplete linguistic distribution assessments into complete linguistic distribution assessments by maximizing the consistency index of complete linguistic distribution assessments. Then, consensus reaching process based on a two-stage-based consensus model is developed to yield the consensual collective linguistic distribution assessments, which are further utilized to derive the consensual importance ratings of CRs. The first stage of the consensus model aims to minimize the adjustment between the original linguistic distribution assessments and the adjusted linguistic distribution assessments, and the second stage aims to minimize the symbolic distance between the original linguistic distribution assessments and the adjusted linguistic distribution assessments. Finally, a case study of product development for EVCARD in China, a sensitivity analysis, and a comparative analysis are presented to show the effectiveness and feasibility of our proposal.