中文摘要：

作者提出了一种基于标准论域确定隶属函数的新方法．该方法对语言值集进行通盘考虑，而不是孤立地确定某一个语言值的隶属函数．作者首先明确提出了标准论域的概念，在标准论域上，语言值的隶属函数形式简单而且易于得到，然后利用语言值在基础论域和标准论域上的峰值点，采用插值或拟合方法确定基础论域到标准论域上的变换，进而由标准论域上的隶属函数得到基础论域上的隶属函数．

英文摘要：

If we decide to use the technologies based determine membership function of a linguistic value on fuzzy logic, say fuzzy control, the first step is to （a word）, and then we can use a fuzzy set represent the linguistic value （the word）. However, there isn＇t a universal method to do it until now. In this paper, a new method to determine membership function is introduced, which is based on normalized uni- verse of discourse （NUOD）. Instead of consider only one linguistic value, this method consider the membership functions of all the linguistic values we are interested in of a linguistic variable at the same time. First, the concept of ＂normalized universe of discourse＂ （NUOD） is introduced. The membership functions defined on NUOD is simple and easy to obtain. Second, using the information of the peak points of linguistic values, determine the mapping from the universe of discourse （UOD） to the NUOD by interpolation or fitting. So we can determine the membership functions on NUOD. membership functions defined on UOD through the membership functions on NUOD.