摘要

Rule based reasoning (RBR) embodied in cam grinding will extend the application of knowledge-based system by utilizing previous knowledge and experience. However, it is very difficult to make rules without exact exceptions. Knowledge acquisition also is a bottleneck problem. Additionally, traditional methods of induction learning from previous experiences limit the development of RBR methodology. This article presents a novel methodology to apply Rough Set (RS) algorithm in rule induction and rule base establishing for RBR system. The RS-RBR system is used in cubic boron nitride (CBN) grinding wheel dressing for cam numerical control (NC) grinding. In the system, the RS is used to process the rule extraction; the RBR is used to assess the dressing parameters. A series of experiments have been accomplished by CNC8312A NC camshaft grinder. Finally, a case study is also implemented to prove the proposed method.