Design of Fuzzy Classifier for Complicated Process Analysis and Its Performance

Application of pattern recognition and AI is a promisive approach to this problem. In the paper, pattern recognition method is used to test the quality of sinter. On the basis of the algorithm of fuzzy clustering analysis, the classifiers based on fuzzy clustering and dynamic state clustering in microcomputer are set up. The performance of the classifiers is tested. The results show that, in the case of lacking priori knowledge, fuzzy clustering analysis is superior to dynamic state clustering analysis when the number of samples is small.