Product Quality Control In Thin Wall Injection Molding Based On Knowledge

Thin Wall Injection Molding (TWIM) technology is a new injection molding process, which owns the apparent advantages such as saving materials greatly, reducing cost, application weight and shape size, improving productivity and better integrated design and assembly. Consumer electronic market demands 3C products (mobile telephones, portable computers and PDA, and etc) is on increase. The development trend of these products is light, thin, short, and small. TWIM technology has being developing quickly.The technology needs special thin wall injection molding machine and higher injection pressure, higher injection speed, higher melt temperature and it can use less material to product thin-wall applications which have the same mechanical performance with traditional products. The technology is a very important further technique in the field of plastic forming. Because of the special molding condition, its processing is very complicated, its mold ability is very poor, its process is very sensitive, and defects emerge more easily. So it is very hard to master the technique. In this case, it is very necessary to study process condition by simulation and control the defects.This paper researched the principle of expert system, the strategy and arithmetic of reasoning, and the material flow of reasoning mechanism. A thin-wall molding technology parameter optimization expert system has been constructed. The main research work of the thesis includes:First, this paper introduces the research background; technology summary and research present situation both at home and abroad of thin wall injection molding technology. And it presents the technology using in the thin wall injection molding system, including expert system, knowledge engineering, fuzzy logic and system inference machine. The diagram of fuzzy expert system in product quality control, the inference flow diagram of the system and the structure of the system is given in the paper. Second, by acquiring expert experience and other knowledge, translating them into a production rule format that can be accepted by CLIPS, we establish knowledge base, including process parameter base, resin material base, process and defect base, defects base and fuzzy rule base. The membership function of process parameters and defects and the fuzzy rules are constructed. The membership function reflects the fault fuzzy and fuzzy connection. Fuzzy reasoning which is based on the Rule is realized.Based on the above theory study, Knowledge Base and Knowledge Base Management System are mainly researched by applying the theory of artificial intelligent, database and fuzzy theory. The basic structure of the system fuzzy inference is given. The procedure of embedding CLIPS on VC platform is given, the information exchange between VC and CLIPS, database access, inference process control and check, inference mechanism are also thoroughly studied. On the basis of the above theories and technologies, combining common principles and methods of expert systems development, a thin wall injection molding parameter optimization expert system is developed using GUI tool-Visual C++6.0 and expert system tool-CLIPS on windows XP platform. At Last this paper gives the applying example to test the Knowledge Base Management System and defects optimization Module.