One 、 Digital twin development background
“ twin ” The concept of "space" originated from NASA's “ Apollo project ”, That is to build two identical spacecraft , One of them was launched into space on a mission , The other is left on earth to reflect the operation of spacecraft in space during the mission period , And then assist engineers to analyze and deal with emergencies in space . Yes, of course , The two spacecraft here are all real physical entities .
2003 Around the year , On the digital twin （DigitalTwin） For the first time Grieves Professor of product lifecycle management at the University of Michigan . however , at that time “DigitalTwin” The word has not been formally put forward ,Grieves Call this idea “ConceptualIdealforPLM（ProductLifecycleManagement）”, As shown in the figure below . even so , In this conception, the basic idea of digital twin has been embodied , That is, the digital model built in the virtual space and the physical entity interactive mapping , Faithfully describe the trajectory of the life cycle of physical entities .
until 2010 year ,“DigitalTwin” The word in NASA In the technical report of , And is defined as “ Integrating multiple physical quantities 、 Multiscale 、 Multi probability system or aircraft simulation process ”.2011 year , The U.S. Air Force has explored the application of digital twins in aircraft health management , The technical challenges of implementing digital Twin are discussed in detail .2012 year , NASA and the air force have jointly published a paper on digital twins , It is pointed out that digital twin is one of the core technologies to drive the development of future aircraft . In the next few years , More and more researches are applying digital twin to aerospace field , Including fuselage design and maintenance , Aircraft capability assessment , Aircraft fault prediction, etc .
In recent years , Digital twins are spreading more and more widely . meanwhile , It depends on the Internet of things 、 big data 、 Cloud computing 、 The development of a new generation of information technology, such as artificial intelligence , The implementation of digital twin has gradually become possible . Today, , Except for Aerospace , Digital twins are also used in electricity 、 A ship 、 city management 、 Agriculture 、 Architecture 、 manufacture 、 Oil and gas 、 Health care 、 Environmental protection and other industries . Especially in the field of Intelligent Manufacturing , Digital twin is considered to be an effective means to realize the interactive integration of manufacturing information world and physical world . Many famous enterprises （ Like Airbus 、 Lockheed Martin 、 Siemens, etc ） And organization （ Such as Gartner、 Deloitte 、 Intelligent Manufacturing Association of China Association for science and technology ） Great importance has been attached to the digital twin , And began to explore a new mode of intelligent production based on digital twin .
Two 、 The concept of digital twin and its typical characteristics
Concepts in standardization organizations ： Digital twin is the digital expression of a specific physical entity or process with data connection , The data connection can ensure the same rate convergence between physical state and virtual state , And provide an integrated view of the entire life cycle of a physical entity or process , To optimize the overall performance .
The concept of Academia ： Digital twin is a virtual entity that constructs physical entities in a digital way , Using historical data 、 Real time data and algorithm model , simulation 、 verification 、 forecast 、 Technical means to control the life cycle process of physical entities .
essentially , Digital twin can be defined as the continuous development of digital information about the history and current behavior of physical objects or processes conducive to optimizing business performance . The digital twin model is based on large-scale , The cumulative , real time , Real world data measurement .
The definition of enterprise ： Digital twins are software representations of assets and processes , For understanding 、 Forecast and optimize performance to achieve improved business results . The digital twin consists of three parts ： Data model , A set of analyses or algorithms , And knowledge .
Digital twin has long been in the industry , It innovates the process in the whole value chain . As a product , Virtual representation of production process or performance , It makes the various process stages seamlessly linked . It's constantly improving efficiency , Minimize the failure rate , Reduce the development cycle , And open up new business opportunities ： In other words , It can create a lasting competitive advantage .
From the concept of digital twin, we can see that , The digital twin has the following typical characteristics ：
1、 Interoperability ： The physical objects and the digital space in the digital twin can be mapped Bi directionally 、 Dynamic interaction and real-time connectivity , Therefore, digital twins have the ability to map physical entities with various digital models , Have the ability to convert between different digital models 、 Merge and build “ expression ” It's the same as .
2、 Extensibility ： Digital twin technology has the ability to integrate 、 The ability to add and replace digital models , It can be targeted at multiple scales 、 Multi physics 、 Multi level model content is expanded .
3、 The real time ： Digital twin technology requires digitization , That is to manage the data in a way that can be recognized and processed by computer to represent the physical entities that change with the time axis . The objects represented contain appearances 、 state 、 attribute 、 Internal mechanism , Digital virtual body mapping that forms the real-time state of physical entities .
4、 Fidelity ”： The fidelity of digital twins refers to the proximity between the digital virtual model and the physical entity . It requires that the virtual body and the solid body not only keep the high simulation of the geometric structure , In state 、 We should also simulate the phase state and the temporal state . It's worth mentioning that in different digital twin scenarios , The simulation degree of the same virtual body may be different . For example, only the physical properties of the virtual body may be required to be described in the working condition scenario , There's no need to pay attention to the details of the chemical structure .
5、 Closed loop ： The virtual body of numbers in the twins of numbers , It is used to describe the visual model and internal mechanism of physical entity , In order to monitor the status data of physical entities 、 Analytical reasoning 、 Optimize process parameters and operation parameters , Realize the decision function , That is to give the digital virtual body and physical entity a brain . So digital twins are closed-loop .
3、 ... and 、 The difference between digital twin and other technologies
Digital twin and simulation （Simulation） The difference between ： Simulation technology is the use of simulation hardware and software through simulation experiments , With the help of some numerical calculation and problem solving , Modeling techniques that reflect the behavior or process of a system , It's a way to transform a model containing deterministic laws and complete mechanisms into software to simulate the physical world , The goal is to rely on the right model and complete information 、 Environmental data , Reflect the characteristics and parameters of the physical world . Simulation technology can only simulate the physical world offline , No analysis and optimization function , So it doesn't have the real-time property of digital twin 、 Closed loop and other characteristics .
Digital twins need to rely on include simulation 、 Actually measured 、 The state of physical entity is perceived by means of data analysis 、 Diagnosis and prediction , To optimize the physical entities , And evolve your own digital model . Simulation technology is the core technology of building and running digital twin , Digital twin is the foundation of data interaction and fusion . On this basis , Digital twins must rely on and integrate other new technologies , Together with the sensor online to ensure its authenticity 、 Real time and closed loop .
Digital twin and information physical system （CPS） The difference between ： Number twin and CPS They all use digital means to build systems to serve reality . In which the ,CPS It belongs to system implementation , The digital twin focuses on the construction of the model and other technical implementation .CPS It's by integrating advanced perception 、 Calculation 、 Communication and control information technology and automatic control technology , Build a physical space and virtual space of people 、 machine 、 matter 、 Environment and information are mapped to each other 、 Timely interaction 、 Efficient collaborative complex systems , Realize the on-demand response of resource configuration and operation in the system 、 Fast iteration and dynamic optimization .
Compared to the comprehensive calculation 、 The Internet 、 Multi dimensional complex system of physical environment CPS, The construction of digital twin is the construction CPS The enabling technology foundation of the system , yes CPS Concrete materialization . The use of digital twins has its own products 、 There are also production lines 、 Factories and workshops , Directly correspond to CPS The products we are facing 、 Equipment and systems, etc . The digital twin was established at the beginning of its establishment 、 Model is a system engineering based on model , It is more suitable to use new computing capabilities such as artificial intelligence or big data for data processing tasks .
Digital twin and digital main line （DigitalThread） The difference between ： Digital mainline is considered as the communication channel of product model evolution and application in all aspects , It's a business system based on the product life cycle , Include product ideas 、 Design 、 Supply chain 、 manufacture 、 After sales service and other aspects . Throughout the life cycle of the product , By providing access 、 Integration and will be different / The ability to transform decentralized data into actionable information to inform decision makers .
The digital mainline is also a communication framework that allows data streams to be connected , And provide an integrated view of the functions of each link in the life cycle . Digital mainline has the ability to provide access to digital twins of products 、 Integration and transformation capabilities , Its goal is to link product life cycle and value chain , Achieve full traceability 、 Information interaction and value chain collaboration . Therefore visible , The digital twin of products is the object 、 Models and data , And the digital thread is the method 、 passageway 、 Links and interfaces .
In short , In the generalized model of digital twins , There are small models associated with each other . Digital mainline can determine the relationship between these small models and provide support . therefore , In the broad sense of life cycle , Digital mainline is a life cycle oriented digital twin .
Digital twin and asset management shell （AssetadministrationShell） The difference between ： From industry 4.0 Our asset management shell , It's German self industry 4.0 Component start , A set of description language and modeling tools developed , So that the device 、 Every asset of an enterprise, such as components, can be interconnected and interoperated . With its modeling language 、 Tools and protocols , When an enterprise constructs a production line , It can have universal interface , I.e. implementation “ Plug and play ” sex , Greatly reduce the time of engineering configuration , Better interoperability between systems .
Since the advent of the digital twin and the asset management shell , More people think that the two are the embodiment of the different industrial cultures of the United States and Germany . in fact , Compared with the asset management shell, which plays a role of control and support “ The housekeeper ”, The digital twin is like a “ practitioners ”, From design 、 Start with models and data , Perceive and optimize physical entities , Push the sensor at the same time 、 Design software 、 The Internet of things 、 Update iteration of new technology . however , Based on the fact that the two are similar in the level of technical implementation , Germany is also trying to transform the asset management shell into the basic technology supporting the digital twin .