Föderatives Produktdatenmanagement anhand semantischer Informationsmodellierung
Föderatives Produktdatenmanagement anhand semantischer Informationsmodellierung
复制标题
联邦产品数据管理和语义信息模型
DOI：
10.3929/ethz-a-001638863
复制
发表时间：
1996
影响因子：
2.8
通讯作者：
R. Montau
中科院分区：
文献类型：
--
作者：
R. Montau
本论文的动机来自于缺乏现有的信息系统，以满足当今企业的信息需求。数据量的增加显示了当前系统在数据结构方面的局限性，并强调了使用语义建模技术的必要性。所介绍的“使用语义信息建模的联合产品数据管理”的方法改进了设计过程中的计算机辅助和与其他活动领域的信息交换。这项工作的主要方面是·需求建模，·零件检索，·零件编号映射和·基于系统联邦的协调。为了支持设计任务的澄清，这是必不可少的和开创性的每一个设计过程，任务澄清和评估的集成系统的开发。在这个系统的中心概念是搜索矩阵作为一个有效的和广泛的建模需求的有条不紊的工具。此外，该矩阵的逻辑结构被用来自动生成用于在评估过程中加权的标准结构。在综合现有成组技术方法的基础上，开发了集成零件检索系统CAOS（Component and Assembly Ordering System），以支持现有解决方案的系统重用。因此，第一次保证了用于整个部件谱的明确搜索机制。对于产品数据的识别，在大多数情况下使用复合编号系统，与纯计数相比，复合编号系统提供更大的语义。然而，它们在独特性和持久性方面是不够的。为了实现这两种编号技术的优点，零件编号映射的方法。这种方法通过利用现有的数据库功能确保分类的完整性。基于系统联邦的协调以最小的自主性和灵活性损失为CIM应用程序的特定任务连接提供服务。协调的基础是两级存储库概念，其用负责本地和全局表示之间的映射的代理来扩展每个组件系统。在协调器中，只有系统间的依赖关系被集中存储，而生产数据则保留在创建它们的系统中。
The motivation for this thesis arises from the lack of existing information systems to satisfy the information requirements of today's enterprises. The increase in the amount of data shows the limitations of current systems in terms of data structu¬ ring and highlights the need for the use of semantic modelling techniques. The in¬ troduced approach of "federated product data management using semantic infor¬ mation modelling" improves both computer assistance within the design process and the exchange of information to other fields of activity. The major aspects of this work are • modelling of requirements, • part retrieval, • part numbering mapping and • coordination based on system federation. To support the clarification of design tasks, which is both indispensable and pathbreaking for every design process, an integrated system for task clarification and evaluation is developed. The central concept in this system is the search matrix as a methodical tool for an efficient and extensive modelling ofrequirements. More¬ over, the logical structure of this matrix is exploited to automatically generate a criteria structure for weighting in evaluation processes. Based on a synthesis of current group technology methods, the integrated part retrieval system CAOS (Component and Assembly Ordering System) is de¬ veloped to support a systematic reuse of existing solutions. As a result, for the first time, definite search mechanisms for the whole spectrum of parts are gua¬ ranteed. For the identification of product data, in the majority of cases compound numbe¬ ring systems are used which, in contrast to pure counting numbers, provide grea¬ ter semantics. However, they are inadequate in terms of uniqueness and perma¬ nence. To achieve the advantages of both numbering techniques, the method of part numbering mapping is developed. This method ensures the integrity of the classification through the exploitation of existing database functionality. Coordination on the basis of system federation serves a task-specific connection of CIM applications with a minimum loss of autonomy and flexibility. The basis for coordination is a two-level repository concept which extends each component system with an agent responsible for mapping between local and global represen¬ tations. In the coordinator, only inter-system dependencies are stored centrally and the productive data remains in the systems which created them.