Abstract: It is widely agreed that an important problem in AI is the representation of commonsense knowledge. combined withthe study on the mechanical design expert system, a new knowledge representation method used to select the material ofspare parts based fuzzy decisions is presented in this paper. With the quantification of the knowledge and experienceof designing engineers, the expert system with the aid of this method can make decisions to select the most suitable materi-al after making simple numerical calculation and judgement of threshold value. It is indicated that the new method has leftout the complex inference and raised the design efficiency.

Key words: fuzzy decision, threshold value, knowledge representation, mechanical design

摘要： 常识性知识的表示，是人工智能的重要课题，结合机械设计专家系统研制实际，本文提出了应用模糊综合决策进行零件材料选择的新方法，该法将设计工程师知识和经验量化，通过简单的数值运算和阈值判断实现最佳材料的选定，从而略去了繁杂的模拟设计工程师推理判断过程，提高了机械设计专家系统的效率。经实现了的机械设计专家系统的举例设计，证明该方法是可行的。

关键词: 模糊综合决策, 阈值, 知识表示, 机械设计