Research On Knowledge Reasoning Technology Of Drawing Die Design For Auto Body

Die design for Auto Body is a complicated course, which depends on designers’experience strictly. How to improve the intelligent degree of die design for body is the key point of die design for this field all along.As computer technologies are applied popularly and artificial intelligence develops rapidly, how to use these technologies to make die design for auto body more normalized, to be able to choose process parameters scientifically and fix on addendum surface, drawbead and drawing threshold, binder surface, drawing direction reasonably has become an important subject that many factories need to solve urgently.The thesis introduces KBE(Knowledge Based Engineering) technology into die design field for auto body, which systematically fixes die design experience and knowledge, utilizes knowledge in the knowledge base as a guidance, figures out problems by use of inference engine, and builds up KBE system for drawing die design for auto body. It can provide theory basis and technology support for auto body die design.Aiming at die design for auto body, a drawing process database for the auto body is set up, and according to concrete characteristics of die design divides the system into subdata bases of different functions. It utilizes combination of Visual C++ 6.0 and SQL Server 2000, reaches inquiry functions and management functions, carries out real-time inquiry and unified management of all sorts of parameters of parts, part materials, fabrication facilities, drawing processes etc..It’s far from enough to build a process database for drawing die design for auto body. In light of the problem that die design technological parameters are hard to be fixed, the thesis, according to basic theories of die design, adopts the KBE reasoning technology linked up with experts’experience to bring up the special KBE system for drawing die design for auto body based on VC++ 6.0 and SQL Server 2000.On the base of selection principles of technological parameters and design of optimization models for targets, the optimization design for dies is achieved by way that the knowledge representation method based on rules is utilized, knowledge base of an expert system is built up combined with SQL Server 2000, and VC++ 6.0 programming is utilized to accomplish the reasoning mechanism of every model in the system and die optimal design based on mixed reasoning strategies of case and rule reasoning. The system utilizes the object–oriented programming technology and the relational database system to render a friendly man-machine interface, which is easy to be supplemented and improved.