Abstract: For the difficult issues of process selection about NC camshaft grinding, on the basis of synthesizing the advantages of rough set (RS) theory and case-based reasoning (CBR), a new RS-CBR process expert system is put forward. In the system, the discrete algorithm and its reduction of RS is used to carry out automatic calculation of process feature attribute weights; the CBR is used to calculate the similarity of process cases. A series of experiments are accomplished by using uniform design principle on CNC8312A NC camshaft grinder, and several groups of data are obtained to establish process case database. On this basis, the RS-CBR model of NC camshaft grinding is established to match grinding process intelligently. A satisfied result is obtained by applying the matched process case to actual production.

Key words: Case-based reasoning, Expert system, NC camshaft grinding, Process matching, Rough set theory