基于知识融合的汽车覆盖件模具设计方法研究

针对产品设计过程中,分散异构知识融合不足导致知识检索准确性不高、知识重用效率低等问题,提出一种知识融合模型.利用知识需求模型传递知识需求信息,采用知识元对各类异构知识进行统一表示,通过概念本体对语义概念进行有序组织,与知识元之间形成映射关系.通过语义距离计算获取候选知识元集,并采用基于贝叶斯和遗传算法的知识融合方法将分散的关联知识元融合成知识单元,实现知识重用.以覆盖件模具结构设计为案例验证了所提出方法的可行性和有效性.

Research on automobile panel die design methodology based on knowledge fusion

Aiming at the problems of low retrieve efficiency and low reuse efficiencyof distributed knowledge in the product design process, a knowledge fusion framework was proposed. The knowledge demand model was used to convey the require-ment information, and the knowledge element was used to express the heterogeneous knowledge in a structured form. The semantic ontology was used to organize the semantic concepts and form a mapping relationship with the knowledge element. Also,the candidate knowledge elements was obtained based on the semantic distance calculation. The knowledge fusion methodology based on Bayesian and genetic algorithm was used to fuse the distributed knowledge elements into knowledge u-nits to realize knowledge reuse. Finally, the feasibility and effectiveness of the proposed method are verified in the die de-sign process.