基于本体的机电产品设计知识管理系统研究

发布时间：2018-06-29 15:57

本文选题：机电产品 + 知识管理　；参考：《南京航空航天大学》2016年硕士论文

【摘要】：目前机电产品设计生产企业逐步表现出快速化、知识化的发展趋势,并且根据研究显示大约80%以上的产品设计需要查询过往设计资源,而目前雷达类的机电产品研发过程中形成的大量设计、制造等方面的文档性知识尚未进行有效的管理,且行业中只有零星的设计知识应用,对于这些宝贵的设计资源,缺乏对资源的组织和重用方法严重制约了雷达类机电产品研发水平的提高和效率的提升。本文在国防基础科研“十二五”重大项目的支持下,针对产品设计知识管理系统进行了若干研究,主要涉及的研究内容如下:(1)对基于本体的机电产品设计知识元建模进行了研究。提出以专业领域维度划分的知识组织方式,将领域本体作为知识元建模的语义模型,通过对象属性将本体中各维度等进行融合,最后,研究了本体导出的OWL文件在关系数据库中结构化存储的方式。(2)对知识的语义信息提取和文档特征提取方法进行了研究。针对本体中蕴含的语义信息,本文提出多影响因子综合加权的语义相似度算法,将本体中概念间的相似度量化表示。在此基础上,提出基于语义增量的关键词权重计算方法,利用关键词特征对文档建模。(3)对基于推理技术的知识检索方法进行了研究。利用本体的推理功能,提出将用户输入的查询进行线上推理与线下推理,分别达到对查询逻辑关系一致性检测和查询关键词扩展的目标。(4)产品设计知识管理原型系统的开发。提出了知识管理系统的开发思路,设计了系统的功能框架,介绍了系统的软件环境,并完成了知识管理系统相应模块的开发。
[Abstract]:At present, mechanical and electrical product design and production enterprises are gradually showing a trend of rapid and knowledge-based development. And according to research results, about 80% of product design needs to inquire about past design resources. However, at present, a large amount of design and manufacturing documents have not been effectively managed in the R & D process of radar mechanical and electrical products, and there is only sporadic application of design knowledge in the industry. For these valuable design resources, The lack of resource organization and reuse seriously restricts the improvement of R & D level and efficiency of radar electromechanical products. Supported by the "12th Five-Year Plan" of the National Defense basic Scientific Research Project, this paper has carried out some research on the product design knowledge management system. The main contents are as follows: (1) Ontology-based knowledge element modeling of electromechanical product design is studied. This paper proposes a knowledge organization method based on domain dimension division, takes domain ontology as semantic model of knowledge metamodel, merges each dimension of ontology through object attributes. Finally, The way of structured storage of owl files exported by ontology in relational database is studied. (2) semantic information extraction and document feature extraction of knowledge are studied. In view of the semantic information contained in ontology, this paper proposes a semantic similarity algorithm which is weighted by multiple influence factors, and quantifies the similarity between concepts in ontology. On this basis, a method of keyword weight calculation based on semantic increment is proposed, and the document is modeled by keyword feature. (3) knowledge retrieval method based on reasoning technology is studied. By using the reasoning function of ontology, it is proposed that users inputted queries can be inferred on line and offline respectively, and the objectives of checking the consistency of query logical relationships and extending query keywords are achieved respectively. (4) the development of product design knowledge management prototype system. This paper puts forward the development idea of knowledge management system, designs the function frame of the system, introduces the software environment of the system, and completes the development of the corresponding module of the knowledge management system.
【学位授予单位】：南京航空航天大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TP311.52;TH122

【参考文献】