Research On Reusability Of Artifact Design Information And Key Technology Of Design Resouce Management

The reusability of artifact design information and key technology of design resource management have been presented. Around reusability for artifact design information, Muilt-view and Muilt-granularity reusable artifact design information model has been brought forward for organizing the reusable artifact design information, the key technology of artifact design information reuse during the artifact design process has been studied., including reusable acquiring technology of artifact design information, reusable retrieving technology of artifact design information, reusable repository management technology etc. Based on these technologies, and supported by China 863 Hi-tech Program and Zhejiang province manufacture enterprise information project program, a GEMC-PDM system for reuse artifact design information management has been developed. This research not only has important value in theory , but also gains fine effects in several projects and mechanical enterprises .In chapter 1, research actualities and develop process of artifact design information reusable theory and technology have been firstly reviewed, and then the current research situation and problems in different technologies, such as modeling technology, acquiring technology, retrieving technology, and management technology of artifact design repository and resource, have been analyzed. Finally , the research background , entire content and the significance of this dissertation are discussed .The basic theory of the reusable artifact design information is put forward in chapter 2. At first, the contents, the characteristics, and the classified system of artifact design information have been addressed. In the following, the basic concept of the reusability for artifact design information has been given, and the evaluable target for reusability and reusable measurement of artifact design information have been established. Based on above researched, the basic methodologies for improving the reusability of artifact design information has been presented, which is the theory base of the following chapters.Analyzing the reusable characteristics of artifact design information, a new organization model for the reusable artifact design information has been put forward. Using by the model, the design reuse process has also been discussed. This model is a information synthesis for supporting reuse artifact design information . It is composed by a self-contained information set. In the model, structural information and top level feature information have been associated. The structure is also reasonable, all reusable levels has been involved. The reusability of artifact design information has been greatly improved by using this model. Finally, the UML and STEP form of model has beenpresented.In chapter 4, Discussion emphasizes on the methods of acquiring of the reusable artifact design information. Two methods have been discussed, including online method and offline method. The offline method, for example, the technology of acquiring for feature information based on RBF has been discussed. A novel method for feature coding has been given. Using by RBF neural network, the efficiency and flexibility of feature information acquiring have been improved. As a example of online method, the acquiring method of artifact function and structure information has been discussed, based on the assemble feature model, function information and structure information has been associated reasonably. By this result, the evolution of artifact design information from function feature to structure feature has been completed.Design retrieve method for the reusable artifact design information has been mainly discussed in chapter 5. Customer-centered reusable design is one of important feature of DFMC (Design for mass customization). Design retrieval is one of the key technology of design reuse. In allusion to insufficiency of design retrieval based on similarity , the Neural Network is introduced into design of case retrieval and a model based Neural Network is proposed. By modeling for the fu...