摘要

Based on the characteristics of the injection molding process, an integrated intelligent model employing case-based reasoning (CBR) and fuzzy inference has been constructed, considering the molding personnel's thought during the trial runs and the advantages of CBR and fuzzy inference. The ideology of CBR is adopted for the initial process parameters setting, which simulates the molding personnel's behavior that they often recall a previous case and set the initial process parameters of the current one by referring to that old case. The case design and matching have been described in detail, and four case adaptation strategies have been discussed. The ideology of fuzzy inference based on expert knowledge and practical experience reflects the test moldings for defects correction and process parameters optimization. Therefore, a fuzzy inference model was built for the correction and optimization process. Its key implementation techniques, such as fuzzification and defuzzification strategies, fuzzy rules definition, membership functions, etc, have been discussed. Finally, a corresponding intelligent system has been developed that is integrated with the injection machine by communicating with the controller. The system can be used to determine the initial process parameters and optimize them online. An experimental study has been carried out for verification.