1008-1542 2019 40 1 68 78 10.7535/hbkd.2019yx01011 article 數字孿生的模型、問題與進展研究 Research on Digital Twin: Model, problem and progress 近年來,關于數字孿生的研究方興未艾。數字孿生作為一個新的范式或者方法體現出了巨大的潛力,但是,這一概念的內涵和范圍尚不確定,尤其是對數字孿生模型概念的界定很不清晰。根據模式類別可以將其分為通用模型和專用模型,其中專用模型仍是當前的研究熱點,研究內容主要體現為對具體項目使用數字孿生方法進行建模,也包括對專用模型進行開發。這些具體項目除了傳統制造業所涉及的零件測量和質量控制,增材制造,設計和工作過程,以及系統管理外,還包括在生物醫藥、石油工程領域的應用等。開發專用模型的工具和技術呈現多元化,有通用工業軟件、專用工業軟件、仿真平臺和自研二次開發工具等等。數字孿生通用模型的研究對象不針對某一具體項目,而是研究如何將模型受控元素表示為一組通用的對象以及這些對象之間的關系,從而在不同的環境之間為受控元素的管理和通信提供一種一致的方法。數字孿生通用模型的研究主要分為概念研究和通用模型的實現方法,兩者的研究熱度相當。其中概念研究從宏觀角度的產品生命周期管理,到描述系統行為,如一般系統行為和系統重新配置,再到具體工作流,如設計方法、產品構型管理、制造系統、制造過程等,研究內容較為發散,尚沒有出現特別突出的熱點。關于數字孿生通用模型實現方面,主要研究了建模語言的構建、模型開發方法的探索、具體工具的使用、元模型理念的植入和模型算法的探索。數字孿生模型是數字孿生研究的核心領域之一,其未來的研究重點是如何將不斷涌現的各不相同的數字孿生體的外部特征和內在屬性歸納為可集成、可交互、可擴展的模型,便于更高效地實現信息在物理世界和數字世界之間流動,從而實現數字孿生的普遍應用,繼而支持CPS(網絡物理空間)和CPPS(網絡物理生產系統)的建設。因此,數字孿生模型研究下一步需要解決的問題是如何對接標準參考架構,如德國提出的工業4.0參考架構模型RAMI4.0和中國的智能制造系統架構IMSA等;關于數字孿生模型需要建立統一的描述方法并確立一致的結論,以規范各自獨立發展建立起來的模型,從而改善模型的互操作性和可擴展性,否則,隨著系統規模的擴大模型效能會顯著下降;中國數字孿生模型的研究急需國產專業工業軟件和建模軟件的支持,以便中國學者深入開展更加符合國情的深入研究。 In recent years, the research on Digital Twin is in the ascendant. As a new paradigm or concept, it shows great potential. However, the connotation and scope of the Digital Twin concept is still uncertain, especially the Digital Twin Model definition is not clear.According to the pattern category, the Digital Twin Model can be divided into general model and special model, in which the special model is still the focus of current research, and the research content is mainly embodied in the use of Digital Twin method to model specific projects. It also includes concept for developing specialized models. These specific projects in addition to the traditional manufacturing related to parts measurement and quality control, manufacturing, design and work processes, as well as system management, but also in the field of biomedical applications and applications for petroleum engineering and so on. There are many tools and techniques for developing special models, such as general industrial software, special industrial software, simulation platform and self-developed secondary development tools, etc.The research object of the Digital Twin general model is not specific to a specific project, but how to represent the controlled elements of the model as a group of common objects and the relationships between these objects. This provides a consistent approach to the management and communication of controlled elements between different environments. The research on the general model is mainly divided into the conceptual research and the model implementation method; the research heat of the two directions is almost the same. Conceptual research ranges from product lifecycle management to system behavior description, such as general system behavior and system reconfiguration, and to product configuration management, to specific workflow, such as design methods, manufacturing systems and manufacturing processes. The research content is relatively divergent, and there is no particularly prominent hot spot. The research of Digital Twin general model implementation is mainly reflected in the modeling language construction, the model development methods exploration, the specific tools usage, the Meta-model concept implantation and the model algorithm exploration.Digital Twin Model is one of the core areas of Digital Twin research. Its future research focuses on how to integrate the external features and intrinsic properties from different Digital Twin artifacts into a model with interoperability, interactivity and scalability for more efficiently realizing the information flow between the physical world and the digital world, thus achieving the universal Digital Twin application, and then supporting the CPS (Cyber Physical Space) and CPPS (Cyber Physical Production System) construction. To this end, the next problem in the Digital Twin Model needing to be solved first is how to dock the standard reference architecture, such as the RAMI4.0 (Reference Architecture Model Industrial 4.0) proposed by Germany and the IMSA (Intelligent Manufacturing System Architecture) by China, etc. Secondly, the Digital Twin Model needs a unified method to describe and it also needs consistent conclusions, in order to standardize the models established by independent development, thus improving the interoperability and scalability of the model. Otherwise, the performance of the model will decrease significantly as the system scales raise. Thirdly, the research on China's Digital Twin Model requires the support of domestic professional industrial software and modeling software, so that the Chinese scholars can carry out in-depth research that is more in line with national conditions. 系統建模;數字孿生模型;網絡物理空間;網絡物理生產系統;工業4.0 system modeling; Digital Twin Model; cyber-physical space; cyber-physical production system; industry 4.0 劉 青,劉 濱,王 冠,張 宸,梁知行,張 鵬 LIU Qing, LIU Bin, WANG Guan, ZHANG Chen, LIANG Zhixing and ZHANG Peng hbkjdx/article/abstract/b201901010
彩票时时乐