Research On Ontology-driven Knowledge Modeling Method For Multi-disciplinary Collaborative Design

With the continuously improvement on the functions and quality requirements of mechanical products, product design becomes more and more complex, which increasingly becomes a collaborative work among different designers in multi-disciplinary. Different knowledge modeling methods are used because of different knowledge background of the designers in multi-disciplinary, therefore, there would be different descriptions on the same product design, which causes knowledge could not be shared normally and it would affect the efficiency of multidisciplinary collaborative design (MCD). The new type of ontology modeling method is applied in this paper for MCD knowledge, which resolves the problem of knowledge sharing and provides a unified expression.First of all, semantic heterogeneous are summarized in MCD knowledge modeling by analyzing the characteristics of MCD knowledge. Heterogeneous are turned to semantic layer using ontology-driven modeling, at the same time, semantic heterogeneous could be mediated between different ontology by building bridge in semantic to achieve knowledge share and reuse.Secondly, principles of MCD knowledge collecting are defined, including principle of class extraction, Domain and Range design, properties extraction and values range. In order to ensure the efficiency of the synergy design, product design is divided into several sub-tasks with the guide of the task-decomposition rules, including rules of appropriate decomposition granularity, equivalent task amount, consistent completion time, relative independence and design experience preferred.To make designers understand the same semantic of one product design from multi-disciplinary, sub-tasks are described in Ontology languages OWL and stored using Ontology editor of protégé4.0.2.Then task-ontology for each sub-task is obtained.The relationship of knowledge containing and the consistent for class is verified according to Ontology Reasoning. In the process of reasoning, it can be deduced whether the class hierarchy is consistent with the reality.