Research On The Quotation Method For Injection Molds Based On CBR

The high-speed development of the Mould industry brought many opportunities for moulds manufactural enterprises, but at the same time it brought a dog-eat-dog market. In order to survival and development, moulds manufactural enterprises must make arduous efforts for more order forms. So how to quote a mold price accurately and quickly was become an important problem when moulds manufactural enterprises were in the face of clients' ask. Thus a research was made by this paper. After referring many literatures, this paper selected man-hour method which was described in the book named《the reference manual of computing dies & moulds' methods》compiled by China Dies & Moulds Industry Association as the theoretic base to compute moulds price, and then introduced cased-based reasoning method into man-hour method to improved it's adaptability, finally combining other technology, such as database technology, A computer aided prototype system of injection molds quotation has been developed. The contents of every chapters of this paper as follow:Chapter one discussed the important meanings of mould quotation quickly and exactly at first, and then looked back mould quotation methods and current situation of computer aided mould quotation system in domestic and international. Lastly it proposed the research contents and goals of the paper.Chapter two presented basic principle and pivotal technology of case-based reasoning. At first it explained definition of CBR, and simply described the idea how to solve problem by CBR. Then narrated CBR's phylogeny, from its origin and status quo which was a study hotspot in AI domain to development trend. Afterward expatiated on pivotal technology in every process of CBR, such as case presentation, case retrieval, case adaptation and case-library update etc.Chapter three built a model to estimate machining man-hour of injection molds using case-based reasoning. Depend on the domain knowledge of injection molds, a few features was extracted to present a machining man-hour case of injection molds in collusion with machining man-hour. It adopted nearest neighbour method as the case retrieval method, gived relevant formulas or tables to calculate similarity between cases or between features. In order to overcome the handicap in obtaining case feature weights in nearest neighbour method, a mathematic model was built to obtain case feature weights automatically by the case data stored in case-base, and then a genetic algorithm (GA) approach is designed to solve the model. Then it gived case adaptation method which had two steps: first the system would give the advised values which were come from every feature difference between cases, and then it would be modified further by human experts. At last how to update and maintain the machining man-hour case-library of injection molds was also expatiated.Chapter four built the quotation model of injection molds which would be adopted by this paper. It described that how to compute machining cost, material cost, design cost, mould test cost, special tools cost, packaging- transport cost, profit and revenue.Chapter five, on the basis of theory research in former chapters, built a computer aided prototype system of injection molds quotation with Visual C++ 6.0 as development tool and SQL Server 2000 as DBMS, and then introduced every function block of this system in detail. In the end of this chapter, an example of injection molds has been used to test the system.Chapter six summarized the main research contents and achievements of this paper, and put forward the research direction for the future.