给出了一个产品设计实例推理的Agent结构,该结构由9个模块构成,其中核心模块是学习模块、推理模块和决策模块。学习模块主要采用基于实例的学习方法对实例进行学习训练,推理模块主要采用产生式规则获取推理规则,决策模块采用相似度决策方法对获取知识进行决策判断。
An agent structure for product design case based reasoning is presented. The structure is made of nine modules. Key modules in the structure include a leaning nodule, a reasoning module and a decision-making module. Case based leaning methods are used to train cases in the leaning module. Reasoning rules in the reasoning module are obtained through production rule reasoning. A similarity decision-making method is adopted for decision-making in decision-making.