Complex equipment digital twin model construction method based on system engineering principle

Abstract

The invention relates to a method for constructing a digital twinborn model of complex equipment based on a system engineering principle, which comprises the following steps of: constructing a functional model {a system architecture model} of the complex equipment by using a system engineering method, comprehensively expressing system architecture, module unit composition and interactive relationship information of the complex equipment, and aiming at the application requirements of a digital twinborn technology, establishing a digital twinborn model of the complex equipment. The method comprises the following steps: constructing an fmu format performance model and a performance model resolving network, carrying out data fusion on external sensing data, generating parameter data capable of reflecting the working condition state of complex equipment, driving the generation of a dynamic visualization effect of a digital twin model, and directly identifying and processing the dynamic visualization effect by a computer system; on the basis of the function model {the system architecture model}, physical models and performance models of all module units of the equipment are automatically linked and integrated, and efficient construction of the digital twin model of the complex equipment is achieved; and finally, the working condition state of the equipment is consistent with the external form of the digital twin application effect.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computer data fusion and visualization, and particularly relates to a complex equipment digital twin model construction method based on system engineering principles. BACKGROUND

[0002] Digital twin technology is a data interaction and fusion technology that has emerged in recent years. This technology focuses on the information interaction between the physical entity of equipment and the digital model, realizes the synchronous mapping of the physical entity state to the virtual simulation, and generates entity equipment operation state prediction data and optimal control signals through data fusion methods with virtual model data, to achieve various digital twin application effects such as "virtual mapping real", "virtual simulation real", "virtual control real", etc. The emergence of digital twin technology is due to the rapid development of hardware device performance, intelligent algorithm technology and digital technology in recent years. In terms of hardware devices, the increase in the number of sensors and the improvement in performance enable the complete and comprehensive acquisition and transmission of equipment operation state data, and the improvement in computer performance provides sufficient computing power support for data processing. In terms of intelligent algorithm technology, various intelligent algorithms represented by deep neural networks have been developed, providing a powerful means for data fitting and prediction. In addition, the widespread application of digital technology in the field of industrial design and manufacturing makes digital models a necessary accompanying data throughout the life cycle of equipment, making digital models a requirement for equipment design, development and delivery. The above factors have jointly promoted the rapid development of digital twin technology.

[0003] The digital twin model is a complete mapping form of the physical equipment entity in the virtual space, and is the most core component element in the application of digital twin technology. The digital twin model data contains information such as physical structure, working condition performance parameters, external data interface, performance parameter calculation relationship and dynamic operation effect. The physical structure is the geometric appearance of the equipment, which is the direct source of the visualization effect of the digital twin model; the working condition performance parameters are a set of performance parameters that can represent the current running state of the equipment, which are the objects of data fusion; the external data interface is the entrance of the sensor data of the physical equipment to the digital twin model, which is the driving source of the operation of the digital twin; the performance parameter calculation relationship is the causal conversion relationship between the working condition performance parameters in the digital twin model, which is the abstract embodiment of the running mechanism of the physical equipment in the virtual model; the dynamic operation effect is the conversion relationship between the performance parameters and the simulated dynamic visualization effect, which is a necessary condition for the synchronous operation of the digital twin. The digital twin model containing the above information continuously interacts with the data information of the equipment entity in the real world, realizing the digital twin application effect.

[0004] The aforementioned digital twin model construction methods primarily focus on the hierarchical composition of the equipment's three-dimensional model and the establishment of related binding data sets. These methods only enable visualization of equipment operation but fail to reflect the equipment's internal functional performance, operating mechanisms, or operating conditions. Furthermore, for digital twin model construction tasks involving complex equipment, these methods are inefficient and involve extensive manual data binding and linking, which can easily lead to operational errors and compromise the quality of the constructed digital twin model.

[0005] The construction of a digital twin model is a prerequisite for the application of digital twins. In order to build a digital twin model, it is necessary to obtain various information about the equipment, including the overall system architecture of the equipment, the functional composition of each component, the interaction relationship, geometry, assembly relationship, and the mechanical, heat transfer, electrical conductivity, control and other subject performance related to the component function; the geometry and assembly in the above information are concentrated in the physical model of the equipment, and the performance of each subject is mainly reflected in the performance model of the equipment; therefore, based on a full understanding of the overall system architecture, functional composition, and component interaction relationship of the equipment, the construction of the digital twin model can be achieved by orderly linking and integrating the physical models of each component and the performance models of related subjects.

[0006] The present invention uses the method of system engineering to construct a functional model {system architecture model} of complex equipment, comprehensively expresses the system architecture, module unit composition and interaction relationship of the complex equipment, and can be directly recognized and processed by the computer system. Based on the above functional model {system architecture model}, the physical model and performance model of each module unit of the equipment are automatically linked and integrated to achieve efficient construction of a digital twin model of complex equipment; in response to the application requirements of digital twin technology, by constructing an fmu format performance model and a performance model solution network, data fusion is performed on external sensor data to generate parameter data that can reflect the working status of complex equipment, and drive the generation of dynamic visualization effects of the digital twin model, and finally achieve a digital twin application effect in which the working status of the equipment is consistent with the external form; through this method, the multi-scale physical model, multi-disciplinary performance model and equipment functional model formed in the design process of complex equipment can be linked and integrated, thereby realizing the construction of a digital twin model of complex equipment and supporting the application of digital twin technology such as complex equipment operation visualization, working condition monitoring, fault diagnosis, and fault prediction. Summary of the Invention

[0007] This paper provides a method for constructing a digital twin model of complex equipment based on systems engineering principles. This method builds a digital twin model of complex equipment by linking and integrating physical models of different scales and performance models from different disciplines, based on the interpretation of functional models (architecture models).

[0008] System engineering is an organizational management technology applied to complex systems, and its research object is mainly the architecture, composition and interaction relationship of complex systems. Through the system engineering method, the system architecture, component function composition, interaction relationship, assembly relationship and other information in the complex equipment can be completely represented, and the function model (also known as the system architecture model) of the equipment is formed; at the same time, with the continuous development of digital twin technology in recent years, the construction of the digital twin model of complex equipment has become one of the problems concerned in the field. Due to the complexity of the equipment, the traditional digital twin model construction method based on manual link integration gradually cannot meet the needs of engineering practice, the manpower and time cost of model construction is high, and the correctness of model construction is difficult to guarantee. In order to solve the above problems, the present application proposes a complex equipment digital twin construction method based on system engineering.

[0009] The method comprises the following steps:

[0010] S1: Analyze the digital twin model construction requirements. The target equipment of the digital twin model construction is determined, and the module unit composition, module unit working principle and overall operation mechanism of the target equipment are analyzed. On this basis, the following several basic settings of the digital twin model construction are determined:

[0011] · Which external sensing data will be obtained by the constructed digital twin as the entity equipment to the digital twin model of the number

[0012] According to the flow input port;

[0013] · Which equipment working condition data needs to be obtained through data fusion, that is, the target parameters of digital twin application;

[0014] · What kind of control signal will the digital twin model output to the entity equipment;

[0015] After the above settings are determined, the external sensing data parameters, the target parameters of digital twin application and the control signal should be combined with the equipment operation mechanism and working principle to confirm that a unique mapping relationship can be established between them. If a unique mapping relationship cannot be established, additional external sensing data input should be added, or the target parameters of digital twin application and the control signal should be reduced.

[0016] S2: Construct the function model (system architecture model). According to the objective actuality of the target equipment of the digital twin model construction in system architecture, module unit composition, module unit interaction relationship, and combining with the application requirements of the digital twin model, the function model (system architecture model) conforming to the system engineering standard is constructed. The function model (system architecture model) should contain the following information:

[0017] ID of all module units: identify each module unit by number coding or string, and the ID should be unique;

[0018] · Hierarchical organization of all module units: the containing and contained relationship between module units, such as multiple basic components constitute a component, multiple components constitute a subsystem, multiple subsystems constitute a complex equipment whole, etc.

[0019]

[0020] · Working condition state parameters of all module units: a set of parameters that can represent the current working condition state of the module unit;

[0021] · Physical pose state parameters of all module units: parameters representing the position and attitude of the geometric structure of the module unit in space, including three coordinate values and three attitude angles.

[0022]

[0023] · Input and output interfaces of all module units: the abstract representation of the interaction between module units, i.e. what kind of parameters the module unit receives from the outside and outputs to the outside. Each input or output interface should include the parameter identification name and the module unit identification ID interacting.

[0024]

[0025] The constructed functional model {architecture model} should comply with the relevant standards of system engineering modeling, such as using SysML language to construct the functional model {architecture model}. According to the specific application requirements of the digital twin model, the granularity of the functional model {architecture model} can be appropriately controlled, such as combining multiple components into a single component and considering it as a basic unit in the model, or omitting the state parameters or input and output parameters that have less impact on the function and performance of the module. In order to reduce the workload of constructing the functional model {architecture model}, it is recommended to refer to the functional model {architecture model} file generated in the equipment design process that complies with the system engineering modeling standards, and modify and edit the existing model file to complete the functional model {architecture model} construction step.

[0026] S3: Generate digital twin model framework. Use system engineering related degrees in computer to check the constructed complex equipment functional model {architecture model}, and determine whether it meets the standards. After the above functional model {architecture model} is interpreted, the digital twin model framework is generated. The model framework contains the hierarchical organization information of all module units, the parameter interaction relationship information between module units, and the sensing data input port and target parameter output port. Figure 2 ​​​The digital twin model framework of a certain equipment is shown. The equipment includes components 1-3, component 2 includes parts 1-2, sensor data is input by component 1, and two target parameters are output.

[0027] S4a: Constructing module unit performance model. According to the input and output parameters of each module unit in the digital twin model framework, construct the performance model of each module unit one by one. The input and output of the performance model should be consistent with the corresponding interaction parameter name in the digital twin model framework. The construction of the performance model should be based on relevant data reflecting the main function of the module unit, such as physical test data, finite element simulation data, logic simulation data, analytical formula, etc. The model construction methods include but are not limited to interpolation, fitting, neural network, analytical formula editing, etc. Finally, the fmu format performance model file is constructed.

[0028] S4b: Linking module unit performance model. Interpreting the performance model in fmu format by computer. According to the module unit name, input parameter name, output parameter name and other information of the performance model, search the matching position in the digital twin model framework, and fill it into the network framework. Interpret the performance model of all module units one by one, until all the input and output parameters in the digital twin model framework are linked, forming a performance model calculation network. Figure 3 The performance model calculation network formed by linking the digital twin model framework based on Figure 2 The performance model calculation network formed by linking the digital twin model framework based on

[0029] S5a: Constructing module unit physical model. Construct the physical model of each module unit based on three-dimensional geometric structure information. The physical model should reflect the main geometric characteristics of the module unit. At the same time, in order to ensure the smooth running of the digital twin model, part of the physical structure details can be ignored according to the application requirements.

[0030] S5b: Linking module unit physical model: According to the assembly relationship information of the entity equipment, establish the transformation relationship between the pose state parameters of each module unit. According to the order from high to low level, define the transformation relationship between the pose state parameters of the lower level module unit and the pose state parameters of the higher level module unit one by one, realize the physical model linking of the lower level module unit to the higher level module unit, and form a physical assembly relationship network. Among them, for the part of the equipment that can perform mechanism action, the corresponding free pose state parameters {such as the relative displacement corresponding to the translation, the attitude angle corresponding to the rotation} should be kept as the free state that is not linked. Figure 4 The physical assembly relationship network formed by linking the digital twin model framework based on Figure 2The physical assembly relationship network formed by the digital twin model framework shown in the figure includes the physical models of component 1, component 2-part 1, component 2-part 2, and component 3. Among them, component 2-part 1 and component 2-part 2 have corresponding free posture state parameters defined because they have mechanism action characteristics.

[0031] S6: Performance model - physical model link: Based on the operating principle and mechanism of the equipment, define the mapping relationship between specific parameters in the performance model solution network and free posture state parameters in the physical assembly relationship network. Among them, the above-mentioned performance model parameters can be either output parameters in the performance model solution network or state parameters of a certain module unit. The constructed mapping relationship includes mathematical analytical equations, conditional judgments, intelligent algorithms, physical engine solutions, etc. Through the above-mentioned mapping relationship, the association between the equipment performance model parameters and the physical model posture is realized, thereby supporting the dynamic visualization of the equipment physical model driven by performance solution, and meeting the application requirements of digital twin technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] This manual has 4 drawings.

[0033] Figure 1 Flowchart of steps for building a digital twin model;

[0034] Figure 2 Schematic diagram of the digital twin model framework;

[0035] Figure 3 Performance model link diagram;

[0036] Figure 4 Schematic diagram of physical model links. DETAILED DESCRIPTION

[0037] The present invention will be described below in conjunction with the accompanying drawings:

[0038] S1: Analyze the requirements for building a digital twin model. Identify the target equipment for building the digital twin model and analyze its modular unit composition, working principles, and overall operating mechanism. Based on this, define the following basic settings for building the digital twin model:

[0039] What sensor data will the constructed digital twin obtain from the outside world as data from the physical equipment to the digital twin model?

[0040] Data stream input port;

[0041] What equipment operating condition data needs to be obtained through data fusion, i.e., the target parameters for digital twin applications;

[0042] What control signals will the digital twin model output to the physical equipment?

[0043] After the above settings are clear, the external sensing data parameters and the target parameters and control signals used by the digital twin should be confirmed to establish a unique mapping relationship. If a unique mapping relationship cannot be established, additional external sensing data inputs should be added or the target parameters and control signals used by the digital twin should be reduced.

[0044] S2: Construct a functional model {system architecture model}. According to the objective reality of the target equipment constructed based on the digital twin model in terms of system architecture, module unit composition, and module unit interaction relationship, and in combination with the application requirements of the digital twin model, a functional model {system architecture model} that meets the system engineering standards is constructed. The functional model {system architecture model} should include the following information:

[0045] Identification ID of all module units: Each module unit is identified by a digital code or a string, and the identification ID should be unique;

[0046] ·Hierarchical organization relationship of all module units: The containing and contained relationship between module units, such as multiple basic components constituting a component, multiple components constituting a subsystem, and multiple subsystems constituting a complex equipment whole, etc.

[0047]

[0048] ·Working condition state parameters of all module units: A parameter set that can represent the current working condition state of the module unit;

[0049] ·Physical pose state parameters of all module units: Parameters representing the position and attitude of the geometric structure of the module unit in space, including three coordinate values and three attitude angles;

[0050]

[0051] ·Input and output interfaces of all module units: Abstract representation of the interaction between module units, i.e., what kind of parameters the module unit receives from the outside and what kind of parameters it outputs to the outside. Each input or output interface should include the parameter identification name and the module unit identification ID interacting.

[0052]

[0053] ​​​The constructed function model {architecture model} should comply with the relevant standards of system engineering modeling, such as using the SysML language to construct the function model {architecture model}. According to the specific application requirements of the digital twin model, the granularity of the function model (architecture model) can be appropriately controlled, such as combining multiple components into a single component and considering it as a basic unit in the model, or omitting state parameters or input and output parameters that have little effect on module function performance. In order to reduce the workload of constructing the function model {architecture model}, it is recommended to refer to the function model {architecture model} file generated in the equipment design process that complies with the system engineering modeling standards, and modify and edit the existing model file to complete the function model {architecture model} construction step.

[0054] S3: Generate a digital twin model framework. Use the system engineering related degree in the computer to check the constructed complex equipment function model {architecture model}, and determine whether it meets the standards. After the function model {architecture model} is interpreted, a digital twin model framework is generated. The model framework contains the hierarchical organization information of all module units, the parameter interaction relationship information between the modules, and the sensor data input port and target parameter output port; Figure 2 A digital twin model framework diagram of a certain equipment is shown, which includes components 1-3, component 2 includes parts 1-2, sensor data is input by component 1, and two target parameters are output.

[0055] S4a: Construct a module unit performance model. According to the input and output parameters of each module unit in the digital twin model framework, construct the performance model of each module unit one by one. The input and output of the performance model should be consistent with the corresponding interaction parameter name in the digital twin model framework. The construction of the performance model should be based on relevant data that reflects the main function of the module unit, such as physical test data, finite element simulation data, logic simulation data, analytical formula, etc. The model construction methods include but are not limited to interpolation, fitting, neural network, analytical formula editing, etc. Finally, the fmu format performance model file is constructed.

[0056] S4b: Link the module unit performance model. Interpret the fmu format performance model using a computer. According to the module unit name, input parameter name, output parameter name, etc. of the performance model, search for the matching position in the digital twin model framework, and fill it into the network framework. Interpret the performance model of all module units one by one until all input and output parameters in the digital twin model framework are linked to form a performance model calculation network. Figure 3 A performance model calculation network linked based on the digital twin model framework shown in FIG. 4 is shown, which includes the performance models of component 1, component 2-part 1, component 2-part 2, and component 3. Figure 2 A performance model calculation network linked based on the digital twin model framework shown in FIG. 4 is shown, which includes the performance models of component 1, component 2-part 1, component 2-part 2, and component 3.

[0057] S5a: Construct a physical model of the module unit. Based on the 3D geometric structure information, construct a physical model of each module unit one by one. The physical model should reflect the key geometric features of the module unit. To ensure smooth operation of the digital twin model, some physical structure details can be omitted based on application requirements.

[0058] S5b: Linking the physical models of modular units: Based on the assembly relationship information of the physical equipment, establish the transformation relationship between the pose state parameters of each modular unit. In descending order, define the transformation relationship between the pose state parameters of the lower-level modular units and the pose state parameters of the higher-level modular units one by one, realize the physical model linking of the lower-level modular units to the higher-level modular units, and form a physical assembly relationship network. Among them, for the part of the equipment that can perform mechanical actions, the corresponding free pose state parameters {such as the relative displacement corresponding to the translation and the attitude angle corresponding to the rotation} should be kept in an unlinked free state. Figure 4 Demonstrated based on Figure 2 The physical assembly relationship network formed by the digital twin model framework shown in the figure includes the physical models of component 1, component 2-part 1, component 2-part 2, and component 3. Among them, component 2-part 1 and component 2-part 2 have corresponding free posture state parameters defined because they have mechanism action characteristics.

[0059] S6: Performance model - physical model link: Based on the operating principle and mechanism of the equipment, define the mapping relationship between specific parameters in the performance model solution network and free posture state parameters in the physical assembly relationship network. Among them, the above-mentioned performance model parameters can be either output parameters in the performance model solution network or state parameters of a certain module unit. The constructed mapping relationship includes mathematical analytical equations, conditional judgments, intelligent algorithms, physical engine solutions, etc. Through the above-mentioned mapping relationship, the association between the equipment performance model parameters and the physical model posture is realized, thereby supporting the dynamic visualization of the equipment physical model driven by performance solution, and meeting the application requirements of digital twin technology.

[0060] Explanation of relevant terms

[0061] Digital twin: A technology that enables equipment monitoring, operating condition prediction, and control strategy optimization through information and data exchange between physical entities and virtual digital models. It includes three technical elements: physical entities, digital twin models, and information data flows.

[0062] Physical entity: In the context of digital twin technology, it refers to a collection of objectively existing materials in the real physical world, such as parts, components, systems, equipment, etc.

[0063] Digital twin model: refers to the digital model containing the information of three-dimensional structure, system architecture, functional composition, running mechanism, etc. of a specific physical entity in computer virtual space, which has equivalent information with the real world physical entity, and is generally constructed by integrating and linking the digital models of specific physical entity function model, performance model, physical model, etc.

[0064] Function model: a digital model describing the internal module composition, hierarchical relationship, function allocation, integrated interface, and running sequence of complex physical entities such as equipment and systems, also known as architecture model. The main form is a set of multiple logical views, which can be constructed by system engineering modeling tools;

[0065] Performance parameter: a parameter describing the state of a target entity such as a component, assembly, or equipment, or the interaction with other entities, such as stress, deformation, temperature distribution, radiation intensity, current, electromotive force, logic state, digital signal, etc.

[0066] Performance model: a digital model describing the calculation relationship between the performance parameters associated with a specific object entity, including interpolation table, analytical formula, iterative analysis algorithm, etc.

[0067] Physical model: a model describing the three-dimensional geometric structure information of a target entity such as a component, assembly, or equipment, which is mainly constructed by specialized computer design software, including prt, stp, step, fbx, etc.

[0068] Assembly relationship: describes the constraint conditions of relative position, relative attitude, and relative motion between two or more physical entities.

Claims

1. A method for constructing a digital twin model of complex equipment based on system engineering principles, characterized by: By interpreting the functional model of complex equipment to generate a digital twin model framework, and by integrating and linking the physical model and performance model of complex equipment to build a digital twin model, it supports various digital twin applications such as complex equipment operation visualization, working condition monitoring, and fault prediction.

2. The method for constructing a digital twin model of complex equipment based on system engineering principles according to claim 1 is characterized by: The functional model is constructed based on system engineering methods and modeling languages, complies with relevant standards of system engineering models, includes the system architecture, functional composition, and interaction relationship information of complex equipment, and can be interpreted by computers.

3. The method for constructing a digital twin model of complex equipment based on system engineering principles according to claim 1 is characterized by: A performance model is a black box model that characterizes the solution relationship of performance parameters. It is obtained by interpolation fitting of experimental simulation data or intelligent algorithm training, or by modeling scientific formulas or logical processes.

Images