Abstract
Process planning is a key step for obtaining high machine quality and efficiency in the polishing. By studying the polishing process planning idea of the skilled technician, a novel process planning method combining with artificial intelligence principle is proposed in this paper. Polishing planning model based on fuzzy theory and case-based reasoning (CBR) technology is investigated in detail, which consists of fuzzy comprehensive evaluation of material machinability, case retrieval, case inference, and case modification. Fuzzy comprehensive evaluation standard based on material physical mechanics performance index is used for determining material cutting performance level. Specific steps are as follows: establishing the factor set, establishing the weight set, establishing evaluation set, fuzzy evaluation of single factor, and fuzzy comprehensive evaluation. The primary cases are chosen according to the grade of material cutting performance. In case retrieval, all primary cases are retrieved in terms of the nearest neighbor principle and the similarity between two cases is calculated according to Euler distance. The retrieval features include the surface roughness before polishing, material characteristics, and the surface roughness requirements after polishing. In the case inference, the method of the cross-correlation coefficient is used for reasoning all cases retrieved in order to evaluate the impact of each process parameter on the surface quality and identify the relevance of each process parameters on the surface quality. In the case modification, the methods of linear extrapolation and parameter adjustment are used for adjusting and revising the process parameters of case retrieved according to the correlation coefficients. At last, example verification is finished and the experiment results are generally acceptable. It is concluded that it is feasible to solve the problem of polishing process parameters selection using this method.