The Research And Development Of Dispersed Knowledge Bases Based Press Process Decision System

In the fierce market competition, small and middle press mold enterprises are on the inferior position, because of lacking of experienced press process decision experts. At the same time, it is difficult for these enterprises to establish press process decision system just depended on their limited knowledge source. Based on the technology of Internet/Web and artificial intelligence, this purpose of this paper is to research on dispersed knowledge bases based multi-reasoning press process decision system, in which DXF files is employed as interface between CAD system and the decision system. The decision system can complete feasibility analyses of press process. By utilizing the technology of Internet/Web, the dispersed press process knowledge bases have been organized into case bases and rule bases, and a neural network of adaptive resonance theory (ART1) is designed to compute index code of case, which is used to search the most matching press process case. When the most similar case can't be found, the rule-based reasoning net is employed to plan press process. On the base of the research work above, the author of this paper has designed and developed a prototype system. This paper is significant for the research on the development method and technology of the press process intelligent decision system based on the technology of Internet/Web.