Research On Representation, Acquisition And Application Of Multi-domain Knowledge For Product Design

Knowledge is the sum of cognition and experience in the practice of changing the world. As a kind resource of the manufacturing industry, knowledge can improve technology and value for products, which is a key factor to sustain competitiveness. In order to bring vitality to market, meet market demand, build competitive advantage, improve efficiency and quality, the knowledge-driven design pattern is exploreed. This pattern is a collaborative process based on knowledge, where design teams improve or create new product on the demand and resource constraints.The pattern not only needs effective collaborative mechanism and platform, but also needs to express, acquire, integrate and synthesize knowledge.So it has practical significance to express and acquire knowledge for product design.Knowledge has characteristic of multi-domain, multi-disciplinary and diversity, which is throughout the whole life cycle of product design. A conceptual, formal and definite model should be created to meet the demand for product design. The knowledge not only includes the explicit knowledge, but also includes the tacit knowledge hidden in the minds of experts. In the views of most designers, the tacit knowledge has more value than the explicit knowledge, which is an important source of innovation. The explicit knowledge is the key factor to the success of product design and innovation, so how to effectively mine and use them has become a hot topic. In addition, during the design process, especially in the conceptual design stage, the incomplete knowledge is often encountered. Such knowledge accompanied with the redundancy directly influences the quality of design.It aims to meet the requirement of the knowledge in product design, the researches of knowledge representation, knowledge visualization, similarity calculation and tacit knowledge acquisition are studied in this paper. Specifically, with classification of multi-domain knowledge, a method of modeling ontology based on the work breakdown structure is proposed. The ontology model and the framework are established to provide foundation work to link knowledge with design process. It aims to find the similar knowledge quickly and effectively from the knowledge base, the semantics is studied. Concepts, attributes and relationships of knowledge are mapped to nodes and edges of the graph, combined with the graph theory. The comprehensibility is improved and the content of knowledge is visually demonstrated. At the same time, a method of calculating similarity based on distance and attribute is discussed. In consideration of experts are the carrier of the tacit knowledge, the idea that expert model is to express tacit knowledge is proposed. The Problem that information of expert overloaded and expert selection for group discussion have been solved by the mechanism, process and retrieval algorithm studied in this paper. The engine of expert recommendation is designed to express and acquire the tacit knowledge to improve the quality of product design. By attempting to use the rough set theory, a model of incomplete knowledge and a method of analysising redundancy are studied. This has a certain reference value to acquire and mine incomplete knowledge from product design.In addition, with the 863 project "research on method and technology for complex product innovation design based on multi-domain knowledge driven" supported, a prototype system which provides methods and techniques to support the product design is designed and implemented in the paper. At the same time, the feasibility and effectiveness of research result are illustrated by this prototype system.