Research On Key Technologies Of Knowledge Service For Mechanical Product Design

In recent years, with the further downturn and volatile of international economic environment as well as economic globalization, China’s manufacturing enterprises are facing with the dilemma of loss of manufacturing industry business and continue to decline in profits. For Chinese manufacturing enterprises, to improve enterprise research and development capabilities and enhance the core competitiveness, is the only way to get rid of this dilemma. Product design is a spiral process of knowledge accumulation and reuse. However, the difficulty in knowledge acquisition, poor knowledge application efficiency and bad knowledge base maintenance situation, affect the knowledge reuse in enterprises and hinder the enhancement of research and development ability, threatening the company’s survival and development. Therefore, the product design oriented knowledge service architecture and some key technologies of knowledge fusion, knowledge integration and knowledge recommendation is researched in this paper, to improve the efficiency of knowledge reuse and enhance the research and development capability and core competitiveness of the companies, extending their value chain to the high-end segment.According to the characteristics of designer knowledge requirement, the status quo of the enterprise data and knowledge management, and the different demand of designers for knowledge service mode, the knowledge application service mode is presented besides the traditional knowledge acquisition service mode, the knowledge service framework consist of knowledge resource layer, knowledge organization layer, knowledge supply layer and design task execution layer, is established, forming the knowledge service process from discrete knowledge organization to knowledge initiative supply. The difficulty of knowledge application is reduced by knowledge fusion, the efficiency of knowledge retrieval is increased by knowledge integration and the way of knowledge acquisition is improved by knowledge recommendation, so the theoretical foundation is laied for improving the efficiency of knowledge reuse.For traditional knowledge-based engineering tool build methods have a bad performance in generality and ease of use, it is hard to meets the requirement of knowledge integration work, then, taking product design guider as the knowledge fusion carrier, a knowledge fusion mode is put forward based on the method of product design guider development method. According to the characteristic of the product variant design, the variant design product hierarchy model is established, product design knowledge is classified and knowledge constraint between design elements is defined, then a knowledge-constraint-based dynamic process planning method is presented and the product design guider model is established, at last the product design guider rapid development method process is given, provides sysmatic method and technical support for knowledge fusion work. Based on this method, product design guider rapid development platform is designed and implemented, which provides an effective toolkit for knowledge fusion work.In order to overcome the difficulty of assembly model retrieval and access in the knowledge service system, a knowledge integration framework based on semantic annotation method is put forword. An assembly semantic annotation method is established based on standard semantic structure, realizes the automatic annotation of assembly model semantics, improve the efficiency and ensure uniformity of assembly semantic annotation work, realize the integration of assembly model. In consideration of the important influence of spatial structure for engineering semantics of assembly model, a spatial structure based similarity matching method of assembly model was proposed, expand the selection scope of standard semantic structure and improve the adaptability of assembly automatic semantics annotation method.In order to improve the accuracy of knowledge recommendation, the designer knowledge requirement model is built which consist of User profile and Task Profile based on the diversity of design task and personalization and dynamic of designers knowledge requirement. Designers knowledge requirement acquirement and updating method is established based on the their behavior of knowledge reuse and the memory-forgetting model, according to this method designer knowledge requirement can be automatically obtained with dynamic update, and it is avoid the subjectivity, low efficiency and lack of ability to track the designer knowledge requirement dynamic change of traditional manual method, increases the efficiency and accuracy of knowledge requirement acquirement. At last, the knowledge recommendation engine based on the User Profile and Task Profile is established, achieves the personalized knowledge recommendation.Finally, the operate flow of the knowledge service system is established based on the abovementioned theory, method and technology. Then, the knowledge service system is developed according to the customizable PLM (product lifecycle management) system of the Shandong shanda hoteamsoft co., LTD.And the system has been applied and verified at last.