Method for automatically generating mass simulation model clusters based on system architecture reproduction, component generalization and parameter variation

Abstract

The invention provides a mass simulation model cluster automatic generation method based on system architecture reproduction, component generalization and parameter variation, and relates to the field of system simulation model cluster automatic generation methods. The method for automatically generating the mass simulation model clusters based on system architecture reproduction, component generalization and parameter variation is provided. According to the invention, the design space can be covered in the aspects of technical state combination coverage and simulation precision; the massive simulation model cluster automatic generation method based on system architecture reproduction, component generalization and parameter variation comprises five implementation steps of architecture derivation, architecture verification, component derivation, parameter derivation and formation of a large-scale system simulation model set capable of covering a design space. The model generation efficiency is greatly improved, massive system simulation model clusters covering the design space can be quickly formed, and a large amount of manual modeling time and energy are saved.

Description

Technical Field

[0001] The present invention relates to the field of automatic generation methods of system simulation model clusters, and in particular to an automatic generation method of massive simulation model clusters based on system architecture propagation, component generalization and parameter deformation. Background Art

[0002] Model-based systems engineering is a model-centric approach for system design, analysis, verification and validation. It promotes cross-disciplinary collaboration, ensures consistency and traceability, and enables automated analysis and simulation through tool support. Compared with traditional document-driven methods, model-based systems engineering can improve the efficiency and quality of system development, and meet specification requirements by automatically generating documents to ensure that the system can effectively meet the set requirements.

[0003] A system simulation model cluster is based on a certain system simulation model. By arranging and combining model components and their connection relationships, a variety of model connection schemes are derived. Based on this, all possible design value states of key parameters are combined to form a large-scale parameter scheme set, which is integrated into the simulation model. All possible technical states and capabilities of the design space are seamlessly combined to generate a large-scale system simulation model set that can cover the design space. A model cluster with more than 10,000 system simulation models is called a massive simulation model cluster.

[0004] Model-based system engineering usually has difficulty in achieving full design space coverage in terms of technical state combination coverage and simulation sophistication during complex system analysis, specification, design, verification and validation, resulting in insufficient scheme demonstration and weak integration in overall engineering scheme demonstration and optimization design. The main manifestations are: 1) The overall and subsystem linkage is not strong, and each subsystem is designed separately. The subsystem performance in the resulting scheme may be optimal rather than the overall performance; 2) Designers are constrained by existing design concepts and design principles, which limits the possibility of scheme innovation and seeking the optimal solution; 3) For the design of new equipment, designers find it difficult to adapt to the new requirements of fast-paced and disruptive innovation. The automatic generation of massive system simulation model clusters can solve the problem that the above-mentioned system simulation models are difficult to cover the entire design space.

[0005] To realize the automatic generation of system simulation model clusters, it is necessary to derive schemes for system models from three dimensions: architecture, components, and parameters. In the architecture dimension, the possible connection modes of component interfaces, the number and type of interfaces, and the types and numbers of key components are used as variables; in the component dimension, the replaceable components are used as variables; in the parameter dimension, the key design parameters are used as variables. Within the range of the design space, all possible design value states are combined to form a massive system simulation model set that can cover the design space.

[0006] The present invention proposes a method to realize scheme derivation of a system model in the dimensions of architecture, components and parameters, and automatically generate massive simulation model clusters to cover the design space in terms of technical state combination coverage and simulation precision. Summary of the invention

[0007] 1. Technical issues to be solved

[0008] In view of the deficiencies in the prior art, the present invention provides a method for automatically generating massive simulation model clusters based on system architecture propagation, component generalization and parameter variation. By deriving solutions in architecture, component and parameter dimensions, comprehensive coverage of the system design space is achieved, ensuring the coverage of technical state combinations and simulation precision, avoiding possible design omissions in traditional methods, and greatly improving the efficiency of model generation through an automated generation process. It is able to quickly form massive system simulation model clusters covering the design space, saving a large amount of manual modeling time and effort. In addition, it supports the automatic generation of a large number of model solutions, and also provides visualization tools and manual intervention mechanisms. Designers can flexibly screen and optimize the generated architecture connection solutions to ensure that the final design solution meets business needs.

[0009] (II) Technical solution

[0010] To achieve the above objectives, the present invention is implemented through the following technical solutions: a method for automatically generating a massive simulation model cluster based on system architecture propagation, component generalization and parameter variation, which realizes the scheme derivation of the system model in the dimensions of architecture, components and parameters, and automatically generates a massive simulation model cluster to cover the design space in terms of technical state combination coverage and simulation precision. The implementation steps of the method for automatically generating a massive simulation model cluster based on system architecture propagation, component generalization and parameter variation are as follows:

[0011] S1: Architecture derivation. The architecture derivation stage requires understanding the system composition and operation principle, building the main architecture model based on the SysML-BDD view, and deriving multiple architecture connection solutions;

[0012] S2: Architecture verification. Among the various architecture connection relationships derived from the architecture derivation phase, some architecture connection solutions are unreasonable. In the architecture verification phase, all architecture connection solutions are manually checked through a visual interface to filter out unreasonable solutions that do not conform to business logic, and the architecture model connection solution in the Modelica language is generated using SysML-Modelica technology.

[0013] S3: Component derivation. In the actual engineering development process, the system components often have multiple models and parameters. Component derivation is based on several specific types of components selected in the main architecture model, setting multiple sets of parameters for them to form multiple replaceable components, and generating multiple component combination schemes through permutations and combinations;

[0014] S4: Parameter derivation: Parameter derivation is based on the key input parameters and output variables that affect the simulation results selected in the main architecture model. Through batch simulation design, experimental design or Monte Carlo design, different parameter assignments are performed on the input parameters according to different business scenarios, so as to observe the impact of input parameter changes on output variables, so as to support the selection of the optimal solution after batch simulation of the model cluster, form a variety of parameter combination solutions in the parameter dimension, and achieve comprehensive coverage of the design space;

[0015] S5: Finally, all possible design value states of the schemes derived from the three dimensions of architecture connection scheme, component combination scheme, and parameter combination scheme are combined to form a large-scale system simulation model set that can cover the design space. The model cluster with more than 10,000 system simulation models is called a massive simulation model cluster.

[0016] Preferably, in the architecture derivation step, all possible architecture connection schemes based on the SysML-IBD view are automatically derived. The architecture master model clarifies which components the system is composed of and which interfaces and types each component has, but does not clarify the connection relationship between components. The architecture connection scheme is based on the architecture master model and adds the connection relationship between components. Only interfaces of the same type can be connected between different components. And because the number of each component in the architecture model is different, and the number and type of interfaces of each component are different, an architecture model can usually derive multiple different architecture connection relationships, and the number of certain types of components between each architecture connection relationship is different, and the interface connection relationship between components is also different;

[0017] Through the above technical solution, taking the winch system as an example, the system usually includes a winch bracket, a rope press, a reducer, a drum, a hydraulic motor, a solenoid valve, a wire rope, etc. In the process of building the main model of the system architecture, you can create it from scratch or pre-build an architecture metamodel library. The architecture metamodel refers to a general architecture model template in a specific field built on the basis of the SysML-BDD view, which is convenient for users to design and make architecture models based on the template. In the architecture metamodel library, select and import an architecture metamodel as a template, and use the drag and connection method to configure the system architecture composition relationship based on the SysML-BDD view. For example, the winch system is configured to be composed of a winch bracket, a rope press, a reducer, a drum, a hydraulic motor, a solenoid valve, and a wire rope. According to the composition and working principle of the winch system, determine and set the possible and prohibited connection relationships of each interface, the number and type of interfaces, and the type and quantity range of key components.

[0018] Preferably, in the architecture verification step, the corresponding Modelica model is selected for each component in the architecture connection scheme to be filled and improved in principle, and the model filled in principle comes from a pre-built component metamodel library based on the Modelica language or is newly created from scratch. Then, the corresponding Modelica model is generated using the SysML-conversion Modelica technology. After the architecture verification stage, the architecture dimension achieves full coverage of the design space.

[0019] Preferably, in the component derivation step, for some components in the architecture model, multiple sets of parameters of a certain component are configured to form multiple replaceable components, and then these replaceable components are arranged and combined by a full arrangement or ordered arrangement method to generate all possible component configuration schemes;

[0020] Through the above technical solution, taking the winch as an example, 5 sets of parameter values ​​are set for the length, width and height of the solenoid valve and drum components in the winch system, corresponding to 5 replaceable components, and finally 5*5=25 sets of component combination solutions are formed through full arrangement. After the component derivation stage, the design space is fully covered in the component dimension.

[0021] (III) Beneficial effects

[0022] The present invention provides a method for automatically generating a massive simulation model cluster based on system architecture propagation, component generalization and parameter variation. It has the following beneficial effects:

[0023] 1. The present invention provides a method for automatically generating a large number of simulation model clusters based on system architecture propagation, component generalization and parameter variation. The method adopts the standard SysML language and does not modify the syntax of SysML. SysML is a system modeling language in the field of system engineering. The design purpose of SysML is to solve the modeling problems faced in system engineering and provide system designers with a simple, easy-to-learn and powerful modeling language. SysML is particularly effective for describing system requirements analysis, structural analysis, behavior description, parameter allocation and attribute constraints in system design analysis. It supports multiple structured and object-oriented methods and processes.

[0024] 2. The present invention provides a method for automatically generating a large number of simulation model clusters based on system architecture propagation, component generalization and parameter variation, using the standard Modelica language without modifying the syntax of Modelica. Modelica language is an object-oriented, equation-based, non-causal, multi-domain unified modeling language, which has a natural advantage in the construction of large-scale complex heterogeneous models involving multiple disciplines and multi-professional couplings such as mechanics, electronics, control, hydraulics, pneumatics, and heat.

[0025] 3. The present invention provides a method for automatically generating massive simulation model clusters based on system architecture propagation, component generalization and parameter variation. By deriving solutions in architecture, component and parameter dimensions, it achieves comprehensive coverage of the system design space, ensures the coverage of technical state combinations and simulation precision, and avoids possible design omissions in traditional methods. Through the automated generation process, the efficiency of model generation is greatly improved, and a massive system simulation model cluster covering the design space can be quickly formed, saving a lot of manual modeling time and effort. It not only supports the automatic generation of a large number of model solutions, but also provides visualization tools and manual intervention mechanisms. Designers can flexibly screen and optimize the generated architecture connection solutions to ensure that the final design solution meets business needs. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 It is a system composition diagram of the present invention;

[0027] Figure 2 A flowchart of an embodiment of the method of the present invention;

[0028] Figure 3 It is a structural diagram of the architecture metamodel during the implementation steps of the present invention;

[0029] Figure 4 It is a design diagram of the simulation main model in the implementation steps of the present invention;

[0030] Figure 5 The architecture model connection scheme 1 in the implementation steps of the present invention;

[0031] Figure 6 The architecture model connection scheme 2 in the implementation steps of the present invention;

[0032] Figure 7 The architecture model connection scheme 3 in the implementation steps of the present invention;

[0033] Figure 8 The architecture model connection scheme 4 in the implementation steps of the present invention;

[0034] Fig. 9 It is a design component diagram during the implementation steps of the present invention;

[0035] Fig.10 A diagram for deriving a configuration of components during the implementation steps of the present invention;

[0036] Fig.11 Generate various component combination scheme diagrams during the implementation steps of the present invention;

[0037] Fig.12 It is a diagram of parameter combination scheme in the implementation steps of the present invention. DETAILED DESCRIPTION

[0038] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. In the description of this application, it should be noted that the terms used here are only for describing specific implementations, and are not intended to limit the exemplary implementations according to the present application. For ease of description, the sizes of the various parts shown in the drawings are not drawn according to the actual proportional relationship. The techniques, methods and devices known to ordinary technicians in the relevant fields may not be discussed in detail, but where appropriate, the techniques, methods and devices should be regarded as part of the authorization specification. In all examples shown and discussed here, any specific value should be interpreted as merely exemplary, not as a limitation. Therefore, other examples of the exemplary embodiments may have different values. It should be noted that similar reference numerals and letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further discussed in subsequent drawings.

[0039] Embodiment 1:

[0040] like Figure 1-12As shown, an embodiment of the present invention provides a method for automatically generating a massive simulation model cluster based on system architecture propagation, component generalization and parameter variation. The method realizes scheme derivation of the system model in the dimensions of architecture, component and parameter, and automatically generates a massive simulation model cluster to cover the design space in terms of technical state combination coverage and simulation precision. The implementation steps of the method for automatically generating a massive simulation model cluster based on system architecture propagation, component generalization and parameter variation are as follows:

[0041] S1: Architecture derivation. The architecture derivation stage requires understanding the system composition and operation principle, building the main architecture model based on the SysML-BDD view, and deriving multiple architecture connection solutions;

[0042] S2: Architecture verification. Among the various architecture connection relationships derived from the architecture derivation phase, some architecture connection solutions are unreasonable. In the architecture verification phase, all architecture connection solutions are manually checked through a visual interface to filter out unreasonable solutions that do not conform to business logic, and the architecture model connection solution in the Modelica language is generated using SysML-Modelica technology.

[0043] S3: Component derivation. In the actual engineering development process, the system components often have multiple models and parameters. Component derivation is based on several specific types of components selected in the main architecture model, setting multiple sets of parameters for them to form multiple replaceable components, and generating multiple component combination schemes through permutations and combinations;

[0044] S4: Parameter derivation: Parameter derivation is based on the key input parameters and output variables that affect the simulation results selected in the main architecture model. Through batch simulation design, experimental design or Monte Carlo design, different parameter assignments are performed on the input parameters according to different business scenarios, so as to observe the impact of input parameter changes on output variables, so as to support the selection of the optimal solution after batch simulation of the model cluster, form a variety of parameter combination solutions in the parameter dimension, and achieve comprehensive coverage of the design space;

[0045] S5: Finally, all possible design value states of the schemes derived from the three dimensions of architecture connection scheme, component combination scheme, and parameter combination scheme are combined to form a large-scale system simulation model set that can cover the design space. The model cluster with more than 10,000 system simulation models is called a massive simulation model cluster.

[0046] Embodiment 2:

[0047] like Figure 1-12 As shown, the embodiment of the present invention provides a method for automatically generating a massive simulation model cluster based on system architecture propagation, component generalization and parameter deformation. The specific implementation steps of the method are as follows:

[0048] Step 1: Build an architecture metamodel library based on SysML-BDD view and a component metamodel library based on Modelica, and provide a multi-domain system simulation master model and a selection component metamodel as a design template for subsequent steps;

[0049] Step 2: Create a new SysML-BDD view architecture model from scratch, or select and import an architecture metamodel from the architecture metamodel library as an architecture model template, such as Figure 3 As shown;

[0050] Step 3: Based on the metamodel in step 2, modify the composition of the main architecture model according to business needs, such as Figure 4 As shown;

[0051] Step 4: Set architecture derivation related variables, including possible connection methods of component interfaces, number and type of interfaces, key component types and quantity ranges, etc.

[0052] Step 5: Based on the settings in step 4, derive all possible architecture model connection schemes based on SysML-IBD views;

[0053] Step 6: Visualize all the architecture model connection solutions in step 5, and manually filter out unreasonable solutions that do not conform to business logic, such as Figure 5 , Figure 6 , Figure 7 , Figure 8 As shown;

[0054] Step 7: Fill in the Modelica principles of the components and use SysML-Modelica technology to generate the architecture model connection solution of the Modelica language;

[0055] Step 8: Based on business requirements, select the components that need to be derived from the main architecture model in step 3;

[0056] Step 9: For each component selected in step 8 that needs to be derived, multiple sets of replaceable components of the component are configured in the component derivation module. The replaceable components come from manually configuring multiple sets of parameters of the component or selecting replaceable component metamodels from the metamodel library, such as Fig. 9 As shown;

[0057] Step 10: Based on the configuration information in step 9, select the full arrangement or ordered arrangement method to derive all possible component combination solutions, such as Fig.10 As shown;

[0058] Step 11: Based on the simulation main model in step 3, identify key parameters and set key input parameters and output variables to the input set and output set of the parameter derivation interface. According to the needs of the business scenario, choose to use batch simulation design (step 12), experimental design (step 13) or Monte Carlo design (step 14) to assign parameters to the input parameters in different ways, such as Fig.11 As shown;

[0059] Step 12: Batch simulation design: manually set the custom input parameter assignment combination;

[0060] Step 13: Experimental design: Set the response variable, factors, levels and experimental design method, and automatically generate multiple parameter combinations;

[0061] Step 14: Monte Carlo design: Set the number of random samples, random number seed, maximum and minimum values ​​of the input set, and distribution mode to automatically generate multiple parameter combinations;

[0062] Step 15: Generate multiple parameter combination schemes based on the method of step 12, step 13 or step 14, such as Fig.12 As shown;

[0063] Step 16: Based on the architecture connection scheme in step 7, the component combination scheme in step 10, and the parameter combination scheme in step 15, all possible design value states are combined to form a large-scale system simulation model set that can cover the design space. A model cluster with more than 10,000 system simulation models is called a massive simulation model cluster.

[0064] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for automatically generating a large number of simulation model clusters based on system architecture propagation, component generalization and parameter variation, characterized by: The method realizes the scheme derivation of the system model in the dimensions of architecture, components and parameters, and automatically generates a large number of simulation model clusters to cover the design space in terms of technical state combination coverage and simulation precision. The implementation steps of the method for automatically generating a large number of simulation model clusters based on system architecture propagation, component generalization and parameter variation are as follows: S1: Architecture derivation. The architecture derivation stage requires understanding the system composition and operation principle, building the main architecture model based on the SysML-BDD view, and deriving multiple architecture connection solutions; S2: Architecture verification. Among the various architecture connection relationships derived from the architecture derivation phase, some architecture connection solutions are unreasonable. In the architecture verification phase, all architecture connection solutions are manually checked through a visual interface to filter out unreasonable solutions that do not conform to business logic, and the architecture model connection solution in the Modelica language is generated using SysML-Modelica technology. S3: Component derivation. In the actual engineering development process, the system components often have multiple models and parameters. Component derivation is based on several specific types of components selected in the main architecture model, setting multiple sets of parameters for them to form multiple replaceable components, and generating multiple component combination schemes through permutations and combinations; S4: Parameter derivation: Parameter derivation is based on the key input parameters and output variables that affect the simulation results selected in the main architecture model. Through batch simulation design, experimental design or Monte Carlo design, different parameter assignments are performed on the input parameters according to different business scenarios, so as to observe the impact of input parameter changes on output variables, so as to support the selection of the optimal solution after batch simulation of the model cluster, form a variety of parameter combination solutions in the parameter dimension, and achieve comprehensive coverage of the design space; S5: Finally, all possible design value states of the schemes derived from the three dimensions of architecture connection scheme, component combination scheme, and parameter combination scheme are combined to form a large-scale system simulation model set that can cover the design space. The model cluster with more than 10,000 system simulation models is called a massive simulation model cluster.

2. The method for automatically generating a large number of simulation model clusters based on system architecture propagation, component generalization and parameter variation according to claim 1 is characterized by: In the architecture derivation step, all possible architecture connection schemes based on the SysML-IBD view are automatically derived. The architecture master model clarifies which components the system is composed of and which interfaces and types each component has, but does not clarify the connection relationship between components. The architecture connection scheme is based on the architecture master model and adds the connection relationship between components. Only interfaces of the same type can be connected between different components. And because the number of each component in the architecture model is different, and the number and type of interfaces of each component are different, an architecture model can usually derive multiple different architecture connection relationships, and the number of certain types of components between each architecture connection relationship is different, and the interface connection relationship between components is also different.

3. The method for automatically generating a large number of simulation model clusters based on system architecture propagation, component generalization and parameter variation according to claim 1, characterized in that: In the architecture verification step, the corresponding Modelica model is selected for each component in the architecture connection scheme to be filled and improved in principle. The model filled in principle comes from a pre-built component metamodel library based on the Modelica language or is newly created from scratch. Then the corresponding Modelica model is generated using the SysML-conversion Modelica technology. After the architecture verification phase, the architecture dimension achieves full coverage of the design space.

4. The method for automatically generating a massive simulation model cluster based on system architecture propagation, component generalization and parameter variation according to claim 1, characterized in that: In the component derivation step, for some components in the architecture model, multiple sets of parameters of a certain component are configured to form multiple replaceable components. Then, these replaceable components are arranged and combined by full arrangement or ordered arrangement to generate all possible component configuration solutions.

Images