Research On Some Problems Of Semantic-Based Manufacturing Knowledge Management Under Networked Manufacturing Environment

Knowledge management is one of key technologies of networked manufacturing. Manufacturing knowledge (MK) is the core knowledge asset under networked manufacturing environment. Thus it is definitely required to manage manufacturing knowledge effectively for networked manufacturing in the age of knowledge economics. The aim of manufacturing knowledge management (MKM) is to facilitate dissemination and share of advanced manufacturing knowledge in networked manufacturing environment to encourage collaboration among enterprises.Supported by the National 863/CIMS research project —"ASP based integrated service and system for Internet-based Manufacturing", MKMS-SW is a prototype system based on the Semantic Web, which is especially designed for MK representation, representation and recognition of web-based manufacturing case knowledge, extraction of relational database knowledge and manufacturing collaboration based on knowledge service. This dissertation mainly explores key technologies and innovative researches of MKMS-SW.In the 1^st chapter, the background of KM is discusses. We firstly summarize the meaning of KM. Its research content, development history and current advances are also reviewed. Then, after analyzing architecture and characteristics of MK of networked manufacturing, the deficiencies of approaches to applying current KM to MKM are pointed out.In the 2^nd chapter, the architecture of MKMS-SW is presented. The design objective of MKMS-SW is addressed firstly. After analyzing and comparing current systems, the logical hierarchy model based on the Semantic Web is brought forward. This model includes an ontology layer which make semantics of MK clear to provide deep supports for MK representation, acquisition, retrieval and networked manufacturing collaboration based on MK. On the basis of this model, both function model and architecture of MKMS-SW are proposed. MKMS-SW chooses s J2EE as its development platform. The technical evaluation of its architecture is given finally.In the 3^rd chapter, a MK representation model is discussed in detail. The requirements of representation of MK under networked manufacturing are analyzed firstly. Then, after exploring current approaches to knowledge representation, a MK representation model, which is based on weighed semantic hyper-graph, is presented. This model is implemented using the ontology and XTM. It has a three-layer structure: ontology layer, XTM layer and resource layer. Besides, in order to facilitate applications to check affectivity and consistency of this model, a formal concept model based on hyper-graph data structure is proposed. Finally, an ontology of MK is presented as well as a scheme of heterogeneous ontology integration.In the 4^th Chapter, the technology of representation and recognition of web-based manufacturing case model is discussed. In order to solve the deficiency of lack of practicability ofcurrent feature-based approaches, a manufacturing case model based on geometry is presented. This model is divided into two layers: case ontology layer and case XTM layer. In addition, an indexing scheme is proposed. A method of calculating case similarity is also presented. Finally, the course of acquisition of case knowledge is discussed as well as its key algorithms.In the 5th chapter, we discuss the acquisition of semantic knowledge stored in a relational database by legacy systems. An acquisition method based on reverse engineering of the relational database is presented. This method includes two courses: the mapping between ER scheme and knowledge ontology, the mapping between table records and instances of knowledge ontology. The mapping rules of the above two mappings are defined. The first mapping is on the basis of reverse engineering of the relational database, thus a new approach to reverse engineering based on ontology is given which includes such two phases as reverse engineering of table definition and reverse engineering of table records.In the 6* chapter, we address a scheme of networked manufacturing collaboration based on KSP (Knowledge Service Provider). And a model of knowledge service-based networked manufacturing collaboration is presented. The key technologies of this model are discussed detailedly, which include application scheme, manufacturing knowledge service modelling, service ontology based on OWL-S, decomposition and synthesis of service based on knowledge service flow, and calculation of service capability.In the 7th chapter, a case study is presented. We create a test environment for CAPP on the basis of process-level of collaborations among enterprises. Through this test environment, the main functions of MKMS-SW are introduced.