Multi-objective supply chain network planning decisions in uncertain environments

• Main Authors: HO, CHUN-YU, 何俊諭
• Other Authors: LIANG, TIAN-FU
• Format: Others
• Language: zh-TW
• Published: 2017
• Online Access: http://ndltd.ncl.edu.tw/handle/w2r72q

碩士 === 修平科技大學 === 精實生產管理碩士班 === 105 === In real-world supply chain network planning decisions (SCNPD), related environmental coefficients and parameters are frequently fuzzy/imprecise in nature owing to incomplete and unavailable information, and the decision maker (DM) must simultaneously handle conflicting objectives over the intermediate planning horizon. This work develops an interactive mixed possibilistic/fuzzy goal programming (MPFGP) technique to solve multi-product, multi-period and multi-echelon SCNPD problems with multiple imprecise goals in uncertain environments. The original SCNPD model designed here attempts to simultaneously minimize the total manufacture/distribution costs and minimize total delivery time with reference to available machine capacity and labor levels at each manufacturer, as well as demand and available warehouse space at each destination, and considers the constraints on total budget and the time value of money of related operating cost categories. Moreover, an interactive two-phases MPFGP method is developed for solving the original multi-objective SCNPD problems based on the fuzzy sets, possibility theory and mathematical programming techniques. Additionally, a systematic solution procedure is formulated to provide a suitable fuzzy decision-making framework of the decision maker (DM) for solving the SCNPD problems, enabling a DM to interactively modify the imprecise data and parameters until a preferred efficient compromise solution and a suitable SCNPD plan are derived. Finally, an industrial case is utilized to implement the feasibility of applying the proposed interactive two-phases MPFGP method to realistic SCNPD problems and several significant implications regarding the practical application of the proposed method are presented. Overall, the significant characteristics that differentiate the proposed MPFGP method from the existing main SCNPD techniques include multiple objectives, wide-ranging decision information, flexible decision-making processes, and considering the time value of money of related operating cost categories. Computational methodology developed in this work can effectively handle the practical SCNPD problems and can easily be extended to relevant management decisions in uncertain environments.