Selecting distribution centre location using an improved fuzzy MCDM approach

AbstractA distribution centre (DC) is usually supplied by manufacturing factories and in turn it supplies the consumers. To reduce transportation costs and enhance operation efficiency, selecting a suitable DC location has become a very important issue for both manufacturing and distribution industries. Recently, Chen [1] proposed a fuzzy MCDM (multiple criteria decision-making) approach to selecting the location of a distribution centre. Despite its merits, two limitations were found in his works, i.e. follows: (1) the final fuzzy evaluation value Pi was treated as a triangular fuzzy number. This is not correct because it may not produce a triangular shape and (2) the difference of two final fuzzy evaluation values, i.e. Pi−Pj, was regarded as a triangular fuzzy number. This is also incorrect. Obviously, Pi−Pj may not produce triangular shape because Pi and Pj may not be triangular fuzzy numbers. To resolve the above limitations and enhance the applicability of the fuzzy MCDM to the DC location selection problem, an improved fuzzy MCDM approach is proposed. In this work, the membership functions of Pi and Pi−Pj are developed to construct an improved fuzzy preference relation, which is further employed with the Chen ranking procedure to complete the proposed method. Finally, a numerical example demonstrates the feasibility of the proposed method.