Research On Discoverying And Reusing Process Knowledge

In economical society, the enterprise must adapt itself to the competition by rapid product development. The process knowledge, as enterprise's the important intellective resource, has been carring weight to the product development. The process knowledge is urgently needed to enhance the quality and level of the process planning.The Application of CAPP (Computer Aided Process Planning) improves the efficiency of process planning. However, exiting CAPP System can't satisfy the enterprise with the intelligence because it is quite difficult of the process knowledge acquisition. This dissertation presents the technology of the process knowledge discovery and reuse. The Knowledge Discovery in Database (KDD) theory is applied to acquire the plentiful and effective process knowledge. The key research works covered in the dissertation are as follows:1. The significance of the process knowledge acquisition in database is analyzed. The research about the process knowledge acquisition and KDD is synthetically elucidated. The requirement of applying KDD in the process knowledge acquisition is put forward.2. The work flow of discovering the process knowledge in database is studied out. Three key technology of the process knowledge discovery are dissertated, the technology of theprocess data pretreatment, the technology of process data mining, and the technology of process knowledge evaluation. Moreover, the oriented-object flexible process data coding technology is discussed to make the computer distinguish the text about process planning.3. The method of defining and modeling the mining target is constructed. The process data extract language (PDEML) is put forward to describe the mining target model.4. The technology of mining the typical process instance is expounded. The typical process route as an exmple is mined based on the Hierarchical Clustering algorithm. A mathematics model based on the data matrix was built to describe the process route. The process route similarity is calculated by the distance equations. The process route clustering method is validated by the calculating example.5. To judge if the candidate of the typical process instance can become the typical process instance, the multi-criterion group decision method is put forward. These criterions are the production cost, the production efficiency, the machining process performanceand the procuct qualitY.6. This dissertation analyzes the process of the process planning utilizing process knowledge. The process knowledge reuse technology based on the process knowledge search and revise is presented. The process knowledge search employs the hierarchical intelligent search algtithms. The process knowledge recvise is realized by the parameteric revise and the modularized revise of typical process instance.7. A prototype system is designed and developed. An enterprise's database as the process data source is applied to validate the theory, technologies, methods and system discussed in this dissertation.