Robust Fuzzy Quality Function Deployment Based on the Mean-End-chain Concept: Service Station Evaluation Problem for Rail Catering Services.

Xin Wu,Lei Nie,Meng Xu

DOI: https://doi.org/10.1016/j.ejor.2017.05.036

IF: 6.4

2017-01-01

European Journal of Operational Research

Abstract:One task that catering services for high-speed railways (CSHRs) must accomplish is to identify and evaluate potential service stations before the design phase of the distribution system. Fuzzy quality function deployment (F-QFD) is one approach for processing the evaluation scheme by translating the basic requirements described in vague terms into actionable alternatives. However, fuzzy importance ratings obtained using F-QFD might be misleading because the approach does not consider the random fluctuations of the fuzzy importance ratings. This paper first proposes a two-phase robust F-QFD process that is integrated with a robust analysis to consider how the QFD process can interact with both the fuzziness and the randomness found in real-world management. Two indicators that measure absolute and relative robustness are proposed. Second, following the mean-end-chain concept, this paper considers the close relationship between the two phases by developing a set of robustness-oriented fuzzy goal programming (RFGP) models to determine the locations of potential service stations. Two robustness indicators are introduced into the two-phase RFGP models to mitigate the adverse effect of random fluctuations. To address the fuzzy and binary variables in the model of phase 2, a hybrid cross-entropy method (HCEA) is developed. The overall framework is termed two-phase robust F-QFD based on the mean-end-chain (MEC) concept (R2-F-QFD-MEC). A series of computational experiments demonstrate both the effectiveness of the framework and the benefits of the robustness-oriented F-QFD. A case study regarding 33 potential service stations along the Beijing-Shanghai high-speed corridor is used to demonstrate the applicability of the method. (C) 2017 Elsevier B.V. All rights reserved.