| Summary: | 碩士 === 國立高雄第一科技大學 === 運輸倉儲營運所 === 93 === In past vendor selection problem (VSP) research, most of them are under single-decision-maker scenario. Also, such decision was usually carried out by single department, i.e., purchasing department, which lacks of expertise and consensus among different departments. Even for group decision-making scenarios, simple geometric average method was used, which could not reasonably address differences in decision weights among decision makers.
In this research, a framework combining analytical hierarchy process (AHP) and non-preemptive goal programming (NPGP) is proposed to determine optimum vendor mix to achieve maximum total value of purchasing (TVP) and minimum total cost of purchasing (TCP). Both qualitative and quantitative criteria along with decision makers’ professional capabilities are considered in the framework.
A TFT-LCD manufacturer (case company A) is investigated as an example to demonstrate the applicability of the framework. Final decision derived by the framework is compared with current status of the case company. Implications of differences are verified by the top management.
To recognize differences brought by different decision-making models, results derived by AHP, LP (Linear Programming), AHP+LP, and AHP+NPGP models are compared with respect to cost and defective-rate metrics. It can be shown that the method proposed (AHP+NPGP) can deal with qualitative and quantitative criteria and can trade-off between cost and defective-rate measures, which is a more comprehensive decision-making model for a VSP.
|
|---|
