本研究的主要目的乃是提出一個模糊多準則決策模式以處理工廠區位選擇的問題。在本研究所構建的決策模式中，利用語意變數的表示評估準則的權重及各項的評估值，藉以降低評估過程的模糊性。這些語意變數均可以模糊數加以表示。在既定的信賴程度值下，本研究利用區間分析定義一個偏好關係函數(preference relation function)以計算任兩個區位方案間的偏好程度。同時，證明此偏好關係函數具有互倒性(reciprocal)與遞移性(transitive)的特質。基於這些特性，本研究提出一個新的模糊數排序方法，以進行區位方案的評選。最後，利用一範例來說明此決策模式的演算過程，以及在不同信賴程度值下，各個區位方案排序的比較結果。

　　　　　The aim of this paper is to present a new multiple criteria decision-making method to solve plant location selection problem under fuzzy environment. In the proposed method, the criteria ratings and the corresponding importance weightings are assessed in linguistic variables which are described by fuzzy numbers. For a given confidence level, a preference relation between each pair of plant locations is defined based on the interval analysis. This preference relation function is verified that it is reciprocal and transitive. And then, a new ranking method is proposed to determine the ranking order of all candidate locations. Finally, a numerical example is solved with different confidence levels to illustrate the procedure of the proposed method in the end of this paper.