航空学报 > 2026, Vol. 47 Issue (8): 432623-432623   doi: 10.7527/S1000-6893.2025.32623

基于知识-规则融合与动态置信度量化的双机器人钻铆智能参数推荐方法

陶永1,2(), 王潇桐2, 刘亚醉2, 刘海涛1, 张宇帆3, 薛雷4, 郭瑞军5, 任帆6, 魏洪兴1   

  1. 1.北京航空航天大学 机械工程及自动化学院,北京 102206
    2.北京航空航天大学 航空发动机研究院,北京 102206
    3.北京航空航天大学 大型飞机高级人才班,北京 102206
    4.中国商飞上海飞机制造有限公司,上海 201324
    5.北京华航唯实机器人科技股份有限公司,北京 100094
    6.南开大学 人工智能学院,天津 300350
  • 收稿日期:2025-07-24 修回日期:2025-08-14 接受日期:2025-10-09 出版日期:2025-12-09 发布日期:2025-12-08
  • 通讯作者: 陶永 E-mail:taoy@buaa.edu.cn
  • 基金资助:
    国家重点研发计划(2022YFB4700400)

Intelligent parameter recommendation method for dual-robot drilling and riveting with knowledge-rule fusion and dynamic confidence quantification

Yong TAO1,2(), Xiaotong WANG2, Yazui LIU2, Haitao LIU1, Yufan ZHANG3, Lei XUE4, Ruijun GUO5, Fan REN6, Hongxing WEI1   

  1. 1.School of Mechanical Engineering & Automation,Beihang University,Beijing 102206,China
    2.Research Institute of Aero-Engine,Beihang University,Beijing 102206,China
    3.School of Large Aircraft Advanced Training Center,Beihang University,Beijing 102206,China
    4.COMAC Shanghai Aircraft Manufacturing Co. ,Ltd. ,Shanghai 201324,China
    5.Beijing C. H. L. Robotics Co. ,Ltd. ,Beijing 100094,China
    6.College of Artificial Intelligence,Nankai University,Tianjin 300350,China
  • Received:2025-07-24 Revised:2025-08-14 Accepted:2025-10-09 Online:2025-12-09 Published:2025-12-08
  • Contact: Yong TAO E-mail:taoy@buaa.edu.cn
  • Supported by:
    National Key Research and Development Program of China(2022YFB4700400)

摘要:

针对双机器人钻铆工艺参数依赖人工经验、工艺知识碎片化导致的智能化决策难题,提出一种融合知识图谱与规则推理的双机器人钻铆智能参数推荐方法。首先,构建双机器人钻铆工艺多层知识图谱,融合工具、工件、工艺、特征、实例层5层的292个实体与711条关系,实现钻铆材料属性、工艺参数、质量指标的多模态语义关联;接着,提出一种两阶段推理方法,通过层次分析法定权的加权余弦相似度算法匹配相似案例,结合基于优先级的规则引擎进行钻铆机器人参数约束与冲突消解,提升双机器人钻铆推荐参数的准确性与一致性;进而,提出一种基于数据与规则约束的动态置信度计算模型,构建双机器人钻铆的数据规模、质量、分布和规则冲突四维评估方法,基于动态权重置信度量化评估实现双机器人钻铆智能参数推荐结果的可解释性。最后,在中国商飞上海飞机制造有限公司的双机器人钻铆工作站的实验表明,该方法推荐参数综合置信度达0.93,且生成参数符合工艺物理约束,满足工艺要求。所提出的方法为航空制造机器人场景的工艺知识沉淀与智能化决策提供了可行的解决方案。

关键词: 参数推荐, 知识图谱, 钻铆工艺, 知识推理, 工艺知识建模

Abstract:

To address the challenge of intelligent decision-making in dual-robot drilling and riveting, which is often hindered by reliance on manual expertise and fragmented process knowledge, this study proposes an intelligent parameter recommendation method that integrates knowledge graphs with rule reasoning. Initially, a multi-layer knowledge graph for dual-robot drilling and riveting is constructed. This graph comprises 292 entities and 711 relationships, organized across five layers: tool, workpiece, process, feature, and instance. It enables multimodal semantic associations among material properties, process parameters, and quality indicators. Based on this knowledge graph, a two-stage reasoning framework is introduced. In the first stage, similar cases are retrieved using a weighted cosine similarity algorithm, where the weights are determined via the analytic hierarchy process. In the second stage, a priority-based rule engine enforces parameter constraints and resolves conflicts. This framework enhances both the accuracy and consistency of parameter recommendations. Furthermore, a dynamic confidence evaluation model is developed by incorporating data- and rule-based constraints, establishing a four-dimensional assessment along the axes of data scale, data quality, data distribution, and rule conflicts. This dynamic weighting mechanism enhances the interpretability of the recommendation results. Experimental validation on a dual-robot drilling and riveting workstation at COMAC Shanghai Aircraft Manufacturing Co., Ltd. demonstrates that the proposed method achieves a comprehensive confidence level of 0.93, with the recommended parameters satisfying physical process constraints. Overall, this method provides an effective approach for capturing and reusing process knowledge while supporting intelligent decision-making in complex manufacturing scenarios.

Key words: parameter recommendation, knowledge graph, drilling and riveting process, knowledge reasoning, knowledge modeling

中图分类号: