• CN:11-2187/TH
  • ISSN:0577-6686

机械工程学报 ›› 2024, Vol. 60 ›› Issue (6): 1-10,20.doi: 10.3901/JME.2024.06.001

• 特邀专栏:数据-知识混合驱动的智能制造系统 • 上一篇    下一篇

扫码分享

数据-知识混合驱动的离散制造系统智能控制体系构架研究

张党1, 赵永宣1,2, 王振军3, 张映锋1   

  1. 1. 西北工业大学机电学院 西安 710072;
    2. 中国航空发动机研究院 北京 101300;
    3. 南昌航空大学航空制造工程学院 南昌 330034
  • 收稿日期:2023-09-30 修回日期:2024-01-31 出版日期:2024-03-20 发布日期:2024-06-07
  • 通讯作者: 张映锋,男,1979年出生,博士,教授,博士研究生导师。主要研究方向为制造物联网、制造系统智能化、数字孪生等。E-mail:zhangyf@nwpu.edu.cn
  • 作者简介:张党,男,1993年出生,博士研究生。主要研究方向为制造系统智能化、复杂产品设计-运维一体化。E-mail:188114409@qq.com;赵永宣,男,1985年出生,博士研究生,高级工程师。主要研究方向为设计制造协同、企业信息化等。E-mail:416171106@qq.com;王振军,男,1974年出生,博士,教授,硕士研究生导师。主要研究方向为复合材料结构/性能一体化设计与高性能制造等。E-mail:wanzhj@nchu.edu.cn
  • 基金资助:
    国家自然科学基金资助项目(U2001201)。

Research on Data-knowledge Driven Intelligent Control Architecture for Discrete Manufacturing System

ZHANG Dang1, ZHAO Yongxuan1,2, WANG Zhenjun3, ZHANG Yingfeng1   

  1. 1. School of Mechanical Engineering, Northwestern Polytechnical University, Xi'an 710072;
    2. China Aviation Engine Research Institute, Beijing 101300;
    3. School of Aeronautical Manufacturing Engineering, Nanchang Hangkong University, Nanchang 330034
  • Received:2023-09-30 Revised:2024-01-31 Online:2024-03-20 Published:2024-06-07

摘要: 在分析当前离散制造系统智能控制、基于数据知识化的制造资源智能化配置、数据与知识驱动算法迭代演进方面所面临挑战的基础上,提出一种数据-知识混合驱动的离散制造系统智能控制体系构架,并进一步分析数据-知识混合驱动在制造系统的多维运作机理。进而,针对架构操作层制造资源端,结合信息物理系统和工业物联技术,设计一种基于操作层与算法层动态交互的制造资源端配置方法。针对架构算法层中数据驱动与知识驱动算法固有优缺点,提炼出数据-知识混合驱动算法在离散制造场景的三种应用模式。所提体系和策略可为新一代离散制造系统的智能控制提供一种理论参考。

关键词: 智能制造, 离散制造系统, 数据-知识混合驱动, 工业物联

Abstract: Current discrete manufacturing systems have challenges in intelligent control, manufacturing resource configuration, and iterative evolution of data and knowledge-driven algorithms. To solve these challenges, a kind of architecture of intelligent control for discrete manufacturing system is proposed in the data and knowledge-driven environment. Based on this, a multi-dimensional operation mechanism of data and knowledge in manufacturing systems is analyzed. Furthermore, for the underlying manufacturing resources in operation layer, combined with CPS and industrial IoT technology, a manufacturing resource configuration method based on the dynamic interaction between the operation layer and the algorithm layer is designed. In view of the inherent advantages and disadvantages of data-driven algorithms and knowledge-driven algorithms in the architecture, three data and knowledge-driven algorithm modes in the algorithm layer are refined. The proposed system and strategy can be a theoretical reference for the research and application of intelligent control in the next generation of smart factory.

Key words: smart manufacturing, discrete manufacturing system, data and knowledge, industrial internet of things

中图分类号: