Research on satisfaction-degree based evaluation for behavior knowledge in intelligent CAD

To improve the reuse ability and management level of knowledge in intelligent CAD, a new method of knowledge evaluation is presented. It represents knowledge of product model with reduced semantic dependency graph, extracts user´s behavior with semantic dependency graph, determines the type of user´s behavior with the constraint relations among features modified by current operation and the feature corresponding to knowledge evaluated and evaluates the knowledge by the process of behavior transitions. This method can not only increase the performance of knowledge reuse and management, but also improve the efficiency of design in Intelligent CAD system. Experiments on computer show that this new method is more adaptable and practicable.