Computer Integrated Manufacturing System ›› 2024, Vol. 30 ›› Issue (11): 3755-3770.DOI: 10.13196/j.cims.2024.0462

    Next Articles

Cloud-edge-device system architecture and key technologies of industrial digital twin system based on multi-agent

LI Hao1,XING Zhiyuan1,LI Linli1+,LU Xiaoping2,YANG Wenchao1,WANG Haoqi1,LIU Gen1,WANG Xiaocong3,LIU Xiaojun4,WEN Xiaoyu1,WU Chunlong5,XING Hongwen6,DONG Liyang2,WANG Zhiqiang7   

  1. 1.Henan Provincial Key Laboratory of Intelligent Manufacturing of Mechanical Equipment,Zhengzhou University of Light Industry
    2.State Key Laboratory of Massive Personalized Customization System and Technology
    3.School of Mechanical Engineering,Dalian University of Technology
    4.School of Mechanical Engineering,Southeast University
    5.National Engineering Research Center for Technological Innovation Method and Tool,Hebei University of Technology
    6.Shanghai Aircraft Manufacturing Limited Company
    7.High-Tech Standardization Research Institute,China National Institute of Standardization
  • Online:2024-11-30 Published:2024-11-27
  • Supported by:
    Project supported by the National Natural Science Foundation,China (No.52175256),the Henan Provincial Science and Technology Research and Development Plan Joint Fund,China(No.225200810029),and the State Key Laboratory of Massive Personalized Customization System and Technology,China(No.H&C-MPC-2023-02-02).

基于多智能体的工业数字孪生系统云边端架构与关键技术

李浩1,邢志远1,李琳利1+,鲁效平2,杨文超1,王昊琪1,刘根1,王晓丛3,刘晓军4,文笑雨1,武春龙5,邢宏文6,董李扬2,王志强7   

  1. 1.郑州轻工业大学河南省机械装备智能制造重点实验室
    2.大规模个性化定制系统与技术全国重点实验室
    3.大连理工大学机械工程学院
    4.东南大学机械工程学院
    5.河北工业大学国家技术创新方法与实施工具工程技术研究中心
    6.上海飞机制造有限公司
    7.中国标准化研究院高新技术标准化研究所
  • 作者简介:
    李浩(1981-),男,河南唐河人,教授,博士,博士生导师,研究方向:工业数字孪生、产品设计方法学、智能制造服务等,E-mail:lihao@zzuli.edu.cn;

    邢志远(1999-),男,河南舞钢人,硕士研究生,研究方向:工业数字孪生、分布式计算,E-mail:1941353506@qq.com;

    +李琳利(1982-),男,河南伊川人,高级工程师,博士,研究方向:工业数字孪生、分布式智能制造等,通讯作者,E-mail:lilinli163@163.com;

    鲁效平(1981-),男,山东泰安人,教授级高级工程师,副院长,研究方向:智能制造、工业互联网,E-mail:luxiaoping@haier.com;

    杨文超(1988-),男,河南郑州人,博士,讲师,研究方向:资源建模与仿真、实时调度等,E-mail:yangwc6532@gmail.com;

    王昊琪(1989-),男,河南郑州人,研究员,博士,硕士生导师,研究方向:工业数字孪生、产品设计方法学、基于模型的系统工程等,E-mail:haoqiwang0218@163.com;

    刘根(1989-),男,河南平顶山人,助理实验师,研究方向:工业数字孪生、模型智能化技术,E-mail:1003198364@qq.com;

    王晓丛(1994-),女,河南信阳人,博士研究生,研究方向:制造系统的智能化技术、智能网联协同控制,E-mail:xiaocongwang@foxmail.com;

    刘晓军(1980-),男,河北张家口人,教授,博士,博士生导师,研究方向:工业数字孪生、数字化设计与制造、高性能装配等,E-mail:liuxiaojun@seu.edu.cn;

    文笑雨(1988-),女,河南唐河人,副教授,博士,硕士生导师,研究方向:车间调度、智能优化算法、工业数字孪生,E-mail:xiaoyuup@gmail.com;

    武春龙(1986-),男,山西大同人,副教授,博士,硕士生导师,研究方向:数字孪生、智能产品服务系统、产品设计方法学、创新设计方法等,E-mail:wuchunlong@hebut.edu.cn;

    邢宏文(1984-),男,湖北襄阳人,正高级工程师,硕士,研究方向:飞机数字化装配、柔性装配产线规划集成、数字孪生等,E-mail:xinghongwen@comac.cc;

    董李扬(1986-),女,河北邢台人,副高级工程师,智能制造实验室主任,研究方向:智能制造、工业互联网等,E-mail:dongliyang@haier.com;

    王志强(1975-),男,青海西宁人,副研究员,研究方向:工业数据、智能制造、工业数字孪生等,E-mail:wangzhq@cnis.ac.cn。
  • 基金资助:
    国家自然科学基金面上项目(52175256);河南省科技研发计划联合基金资助项目(225200810029);大规模个性化定制系统与技术全国重点实验室开放课题(H&C-MPC-2023-02-02)。

Abstract: The multi-agent-based industrial Digital Twin System (iDTS) is composed of multiple agents,including humans,machines,environments,digital twin simulation systems and computing systems.It is a new generation of iDTS system driven by the integration of multiple agents,Artificial Intelligence (AI) algorithms and digital twin systems.In view of the real-time,intelligence and load balancing needs of the iDTS system,the typical characteristics,cloud-edge-device system architecture and key technologies of the iDTS system were presented based on multiple agents,which mainly included the methods of agents' self-learning and optimization driven by Large Language Models (LLMs),the methods of clock synchronization and data sharing for the interaction of iDTS system based on multiple agents,the distributed control methods for the collaboration of multiple agents in the iDTS system,the collaborative scheduling methods for multiple manufacturing resources in the iDTS system and the load balancing methods for the iDTS system with a cloud-edge-device architecture.Finally,the application effect of multi-agent-based iDTS system was demonstrated through a typical case.The proposed system architecture and key technologies provide important reference for the in-depth application of AI and digital twin technology in intelligent manufacturing processes.

Key words: industrial digital twin, multi-agent, artificial intelligence, cloud-edge-end collaboration, distributed computing

摘要: 基于多智能体的工业数字孪生系统(iDTS)由人、设备、环境、孪生仿真系统和计算系统等多智能体构成,是一个多智能体、AI算法和数字孪生融合驱动的新一代iDTS系统。针对工业数字孪生系统在实时性、智能性和负载均衡等方面的需求,提出了基于多智能体的iDTS系统的典型特征、云边端架构和关键技术体系,主要包括模型驱动智能体自学习优化方法、多智能体交互的同步与共享方法、多智能体协同的分布式控制方法、iDTS系统制造资源协同调度方法和iDTS系统的云边端负载均衡方法等。最后,通过典型案例展示了基于多智能体的iDTS系统的应用效果。所提出的系统架构与关键技术为推动人工智能和数字孪生技术在智能制造过程中的深入应用提供了重要参考。

关键词: 工业数字孪生, 多智能体, 人工智能, 云边端协同, 分布式计算

CLC Number: