Research On Key Technology Of Manufacturing Oriented Knowledge Based Engineering System

During the development of information technology in manufacturing industry, how to make use of new technology to build the most perspective enterprise management platform has become the principal problem. The application of new high technology needs more effective and efficient management. This paper is about how to get rapidly and accurately, the enterprise needed information-knowledge, from the abundant, structural and non-structural information source, to manage efficiently, and to serve the decision-making in the end. The managed object is information and knowledge extracted from partial application system such as Enterprise Resource Plan (ERP), Customer Relationship Management (CRM), Supply Chain Management (SCM) and Computer Integrated Manufacturing System (CIMS). The consistent object can be treated as a whole. Aiming at the manufacturing enterprise's feature and problem, we give the outline of strategy. The knowledge based engineering system is the specific concretion of knowledge engineering in many respects. The management and application of knowledge is to get the decision support for the leader to conduct the enterprise strategically. From the point of practical requirement, according to the research approach of knowledge engineering, the study discusses about the knowledge essential, knowledge expression, knowledge procurement, and inference technology, building up the share of knowledge. Comparing different knowledge expressing methods with one another, analyzing the different inference technologies and modeling patterns, choosing from the most appropriate platform, we make a conclusion that system architecture design is feasible. And then, a solution to found the enterprise standard platform is brought forward; using advanced mid-ware technology, databases with high performance (real application clusters). The enterprise solution with open source and distributed architecture is put forward-the Knowledge Based Engineering system conceptual model based on J2EE platform. We have implemented partially the kernel functions. With object-oriented method, we make the knowledgebase integrate with XML technology, and build it in the relational database management systems. According the theory of rule based system architecture design, we choose the inference engine-Mandarax, which is an open source Java rule platform based on the logic programming concepts of derivation rules with negation-as-failure and backward reasoning, as the inference component of KBE system. A web application of decision support system based on KBE is established. This application is the implementation of intelligent query knowledgebase system. At last, in order to make the system compatible and expansible, we discuss the system self-study function model and the core of knowledge discovery process, according the theory of data mining and knowledge discovery development.