Abstract: Aiming at the problems of aerospace complex product assembly semantics without standard constraints,difficult reuse of assembly process and long assembly design cycle,a knowledge-driven assembly process reconstruction method for aerospace complex products is proposed.By analyzing the correlation between satellite assembly semantics and process elements,the logical relationship and multi-level ontology model of satellite product assembly process knowledge system are constructed,and the hierarchical association representation of satellite product assembly knowledge is realized.Secondly,the assembly process knowledge mining model of satellite products is designed.On this basis,the assembly process similarity calculation and improved AP clustering algorithm are used to mine the typical assembly process knowledge,and the effective reuse of historical assembly process knowledge is realized.Finally,taking a certain satellite product as an example,a rapid configuration generation system for assembly process is developed.The successful implementation of this technology in aerospace enterprises verifies the effectiveness of the proposed method,which shortens the generation cycle of assembly process description content and supporting information by 52%.

Key words: aerospace complex products, assembly process design, ontology model, knowledge mining and reuse, process rapid configuration

摘要： 针对航天复杂产品装配语义无标准约束、装配工艺难重用、装配设计周期长等问题,提出一种基于知识驱动的航天复杂产品装配工艺重构方法。通过分析卫星装配语义及工艺要素间的相关性,构建了卫星产品装配工艺知识体系逻辑关系和多级本体模型,实现了对卫星产品装配知识分层关联表示。其次,设计了卫星产品装配工艺知识挖掘模型,在此基础上,利用装配工艺过程相似度计算和改进AP聚类算法挖掘出典型装配工艺知识,实现历史装配工艺知识的有效重用。最后,以某型卫星产品为例,开发了装配工艺快速配置生成系统,该技术在航天企业的顺利实施验证了所提方法有效性,使得装配工艺配套信息生成效率提升,卫星产品装配周期缩短32%。

关键词: 航天复杂产品, 装配工艺设计, 本体模型, 知识挖掘与重用, 工艺快速配置