Product design based on clustering analyzing and fuzzy recognition

Based on clustering analyzing, a structure selecting method was proposed for design-oriented expert system during product conceptual design. The clustering analyzing method was introduced to find out the attribute sets in different levels matched with structure tree so as to set up a product characteristic space. In this space, each sub-space is matched with a type of product structure. Similarly, the standard model of product can be set up according to the clustering result and all the models are formed as standard structure tree. Then, the fuzzy recognition method is adopted to determine the type of product structure by mapping one certain node of the tree with the design requirements. The method can be wide used in applications of mechanical product design.