Apparatus and method for complex system multi-solution generation and trade-off analysis based on sysml model

Abstract

The application provides a complex system multi-scheme generation and trade-off analysis device and method based on a SysML model, the device comprises a generation module and an analysis module; the generation module comprises an algorithm parameter design submodule, an optimization design submodule, a scheme generation and data pushing submodule; the analysis module comprises a design variable definition submodule, a device driving and parameter transmission submodule, a scheme library submodule, a trade-off analysis configuration submodule and a TradeStudy module. The application solves the problems of long time consumption, low efficiency and low interpretability caused by artificial construction and analysis of schemes during multi-scheme design and trade-off analysis based on the SysML model, and optimizes the overall process of complex system design.

Description

Technical Field

[0001] This invention belongs to the field of computer system technology based on specific computational models, and specifically relates to a device and method for generating and weighing multiple options for complex systems based on the SysML model. Background Technology

[0002] With the increasing complexity of system design (such as aerospace and unmanned systems), traditional document-based systems engineering methods have shown significant limitations in requirements management, design collaboration, system integration, and verification. To address this challenge, Model-Based Systems Engineering (MBSE), as an emerging systems engineering paradigm, is gradually becoming the mainstream approach for complex system design and development. MBSE significantly improves the efficiency and quality of system development by constructing unified, traceable, and executable system models, enabling full lifecycle management from requirements to design, verification, and validation. Within the MBSE technology framework, Systems Modeling Language (SysML), as a standardized modeling language, is widely used for requirements modeling, behavioral modeling, structural modeling, and parametric modeling of complex systems. SysML provides various standard iconographic tools, supporting multi-dimensional system expression from function, structure, behavior to parameters, providing a technical foundation for model-driven design throughout the system lifecycle. However, despite SysML's powerful expressive capabilities in system modeling, current SysML-based complex system design methods still face several key bottlenecks, particularly in the generation and trade-off analysis of multiple system solutions.

[0003] SysML, as a standard system architecture modeling language, provides a complete view of complex system design, from top-level requirements to bottom-level implementation. However, in the early stages of system design, requirements are often unclear, boundaries are ambiguous, and constraints are dynamically changing. Traditional design methods rely on engineers' experience and limited trial and error, making it difficult to systematically explore the vast and non-intuitive design space, thus missing more innovative or higher-performing solutions. Furthermore, when multi-scheme analysis is involved, engineers typically need to manually create and configure each design scheme to be analyzed. This process is time-consuming, labor-intensive, error-prone, and heavily reliant on the engineer's personal experience and breadth of knowledge, resulting in long design iteration cycles and low efficiency. There is an urgent need for a method and tool that can automatically generate massive amounts of design schemes and perform multi-scheme trade-off analysis. Summary of the Invention

[0004] This invention aims to solve one of the above-mentioned technical problems by providing a device and method for generating and analyzing multiple schemes for complex systems based on the SysML model, thereby solving the technical problem of generating massive design schemes and conducting multi-scheme trade-off analysis for complex systems.

[0005] The technical solution adopted by the present invention to solve the above-mentioned technical problems is as follows:

[0006] This invention provides a device for generating and weighing multiple options for complex systems based on a SysML model, including a generation module and an analysis module;

[0007] The generation module includes:

[0008] The algorithm parameter design submodule is used to read design variables and design constraints from the analysis module, as well as modify design variables and set algorithm parameters.

[0009] The optimization design submodule is used to construct an optimization model based on the received algorithm parameters, design variables, and design constraints. It is used to store multi-objective optimization algorithms and generate a large number of undecoded solutions for complex systems based on the optimization model and multi-objective optimization algorithms.

[0010] The scheme generation and data push submodule is used to receive massive amounts of undecoded schemes, obtain massive amounts of decoded schemes through decoding, and push the massive amounts of decoded schemes to the analysis module.

[0011] The analysis module includes:

[0012] The design variable definition submodule is used to define design variables and design constraints for complex systems.

[0013] The device drive and parameter transmission submodule is used to output design variables and design constraints, as well as the drive generation module;

[0014] The scheme library submodule is used to obtain a massive number of schemes decoded by the generation module and map the massive number of schemes into a scheme library that exists as instance elements.

[0015] The trade-off analysis configuration submodule is used to build the encapsulation module. The encapsulation module inherits from the TradeStudy module through generalization, encapsulates the complex system through the composition of associations, sets the trade-off analysis objectives and the criteria for evaluating each solution, generates auxiliary parameter schemes for trade-off analysis through parameter scanning, and obtains the solutions from the solution library and outputs them to the TradeStudy module.

[0016] The TradeStudy module has a built-in TradeStudy mechanism. The TradeStudy module is used to receive solutions from the solution library and auxiliary parameter solutions for trade-off analysis. Based on the trade-off analysis objectives and standards, and combined with the encapsulated complex system, it uses the SysML model to simulate the configuration diagram to perform multi-solution trade-off analysis of the complex system and outputs the optimal solution design from the solution library.

[0017] Furthermore, the generation module also includes a data transmission channel, which is used to connect the generation module and the analysis module to transmit data in a standard format;

[0018] The algorithm parameter design submodule reads the design variables and design constraints of the design variable definition submodule in the analysis module through the data transmission channel;

[0019] The scheme generation and data push submodule pushes the decoded massive number of schemes to the scheme library submodule in the analysis module through the data transmission channel.

[0020] Furthermore, the design variables include structural parameters and performance parameters;

[0021] The algorithm parameters include at least the maximum number of generations, the population size, and the number of optimization targets;

[0022] The optimization model includes at least a solution space, an objective function, and design constraints. The solution space includes all structural parameters and performance parameters. The objective function includes at least two optimization objectives. The design constraints include at least equilibrium conditions, size ratios, and resource constraints.

[0023] The data transmission channel is also used to output data to local data storage;

[0024] The algorithm parameter setting submodule displays algorithm parameters and allows for custom input in a tree structure.

[0025] Furthermore, the design variable definition submodule uses the module definition diagram of the SysML model to define the design variables and design constraints of the complex system;

[0026] The device driving and parameter transmission submodule creates a parameter diagram using a SysML model based on the complex system design variables and design constraints, and uses a Python language driver generation module embedded in the SysML model constraint module.

[0027] The tradeoff analysis configuration submodule retrieves solutions from the solution library using reference elements in the internal module graph of the SysML model and outputs them to the TradeStudy module.

[0028] The tradeoff analysis configuration submodule generates a tradeoff analysis auxiliary parameter scheme by scanning the value attribute elements of the internal module graph in the SysML model.

[0029] The TradeStudy module includes several extended value attributes, which are used for storing values ​​during the trade-off analysis process.

[0030] Furthermore, the TradeStudy module's extended value attributes include:

[0031] OutOfSpec is used to store the proportion of schemes that do not meet the conditions of the trade-off analysis;

[0032] N is used to statistically calculate the number of options involved in the trade-off analysis;

[0033] Score is used to store the target value of the tradeoff analysis criteria;

[0034] Winner is used to store the winning solution that best meets the trade-off analysis criteria among a large number of solutions.

[0035] Furthermore, the tradeoff analysis configuration submodule has built-in BoundReference and designVariable constructor types, which are used to extend the bindingPath, boundEnd, min, max, and step feature attributes;

[0036] The bindingPath is the relative position of the value attribute linked by the binding linker, which is automatically obtained through the binding linker;

[0037] The boundEnd is the value attribute linked by the binding linker, which is automatically obtained through the binding connector;

[0038] The min value is set as the minimum value of the value attribute.

[0039] The `max` parameter is set to the maximum value of the attribute.

[0040] The step is the parameter generation step size between the maximum and minimum values ​​of the value attribute;

[0041] By using the BoundReference and designVariable constructors, the values ​​of min, max, and step are set to generate the value attributes to be analyzed in batches. The value attributes are then connected by a binding connector to complete the input of the auxiliary parameter scheme for the trade-off analysis.

[0042] Furthermore, the scheme generation and data push submodule is also used for the visualization of massive schemes decoded through tables and scatter plots;

[0043] The massive decoding scheme includes the values ​​of design variables;

[0044] The standards include at least cost standards and performance standards.

[0045] This invention also provides a method for generating and weighing multiple options for complex systems based on SysML models, the method comprising the following steps:

[0046] Define the design variables and design constraints for complex systems;

[0047] Set the algorithm parameters for the multi-objective optimization algorithm;

[0048] Construct an optimization model for a complex system based on algorithm parameters, design variables, and design constraints;

[0049] Based on the optimization model of complex systems, a massive number of undecoded solutions for complex systems are generated through a multi-objective optimization algorithm.

[0050] Decode to obtain a massive number of decoded solutions;

[0051] Create a SysML model instance table, and use the decoded massive number of solutions to create instance elements to build a solution library;

[0052] Create encapsulation modules, which inherit the TradeStudy mechanism through generalization relationships, and encapsulate complex systems through association relationships.

[0053] Set the objectives of the trade-off analysis and the criteria for evaluating each option;

[0054] Auxiliary parameter schemes for tradeoff analysis are generated through parameter scanning;

[0055] The system retrieves solutions from the solution library and auxiliary parameters for trade-off analysis. Based on the TradeStudy mechanism, and according to the trade-off analysis objectives and standards, combined with the encapsulated complex system, it uses a SysML model to simulate the configuration diagram and performs multi-solution trade-off analysis of the complex system, outputting the optimal solution design from the solution library.

[0056] Furthermore, the multi-objective optimization algorithm is NSGA-II;

[0057] The optimization model includes at least a solution space, an objective function, and design constraints. The solution space includes all structural parameters and performance parameters. The objective function includes at least two optimization objectives. The design constraints include at least equilibrium conditions, size ratios, and resource limitations.

[0058] The optimization objectives include at least optimal structural design and minimum cost consumption.

[0059] Furthermore, the complex system is an aircraft, and the design variables of the aircraft are wingspan, mean chord length, wing material thickness, and cruise speed;

[0060] The objective function is: ,in, Let be the objective function. To be the minimum value, For aerodynamic drag, For manufacturing costs;

[0061] The aerodynamic resistance and manufacturing cost are respectively: , ,in, Where S is the air density and S is the wing area. denoted as zero-lift drag coefficient, e as start-up efficiency factor, k as induced drag correction coefficient, and W as total aircraft weight. For wingspan, For the average chord length, For the thickness of the wing material, Cruise speed;

[0062] The design constraints include:

[0063] Lift balance constraints: ;

[0064] Aspect ratio constraint: ;

[0065] Thickness-chord ratio constraint: ;

[0066] Process feasibility constraints: ;

[0067] Velocity-wingspan coupling constraints: ;in, The maximum lift coefficient, Each sets a threshold.

[0068] The beneficial effects of this invention compared to the prior art are as follows:

[0069] This invention discloses a device and method for generating and weighing multiple solutions for complex systems based on the SysML model. On one hand, it employs a multi-objective optimization algorithm to rapidly generate massive design solutions for complex systems, addressing the problems of insufficient design space exploration in the early stages of complex system design, low solution generation efficiency, and reliance on manual intervention. On the other hand, it utilizes a multi-solution trade-off analysis method based on the TradeStudy mechanism to address the issues of strong subjectivity, insufficient decision-making basis, and poor interpretability in current multi-solution analysis processes. This invention overcomes the bottleneck of the "last mile" from requirements to parametric analysis in MBSE (Multi-Solution Design), significantly improving the overall process efficiency, solution quality, and scientific rigor of decision-making in complex system design. Attached Figure Description

[0070] The accompanying drawings, which form part of this specification, are used to provide a further understanding of the embodiments of the invention and illustrate the principles of the invention together with the textual description. It is obvious that the drawings described below are merely some embodiments of the invention, and those skilled in the art can obtain other drawings based on these drawings without any creative effort.

[0071] Figure 1 A schematic diagram of a complex system multi-scheme generation and trade-off analysis device based on the SysML model provided for a specific embodiment of the present invention;

[0072] Figure 2 A schematic diagram illustrating the principle of a complex system massive solution generation module based on NSGA-II provided for a specific embodiment of the present invention;

[0073] Figure 3 A schematic diagram illustrating the invocation of a module for generating massive solutions for complex systems based on parameter graphs, provided for a specific embodiment of the present invention.

[0074] Figure 4 A schematic diagram of the composition of a complex system massive solution generation module based on NSGA-II provided for a specific embodiment of the present invention;

[0075] Figure 5 A flowchart illustrating the principle of the NSGA-II multi-objective optimization algorithm provided in a specific embodiment of the present invention;

[0076] Figure 6 A schematic diagram illustrating the principle of a multi-option tradeoff analysis module based on TradeStudy provided for a specific embodiment of the present invention;

[0077] Figure 7 This is a schematic diagram illustrating the acquisition and parameter binding of a multi-scheme trade-off analysis based on an internal module diagram, provided for a specific embodiment of the present invention. Detailed Implementation

[0078] Specific embodiments of the present invention will now be described in detail. In the following description, specific details are set forth for purposes of explanation and not limitation, in order to aid in a thorough understanding of the invention. However, it will be apparent to those skilled in the art that the invention may be practiced in other embodiments departing from these specific details.

[0079] It should be noted that, in order to avoid obscuring the invention with unnecessary details, only the device structure and / or processing steps closely related to the solution of the invention are shown in the accompanying drawings, while other details that are not closely related to the invention are omitted.

[0080] This invention belongs to the field of computer system technology based on specific computational models, and relates to a complex system design method that automatically generates massive design schemes based on multi-objective optimization algorithms and performs multi-scheme trade-off analysis when development requirements and design constraints are unclear. The complex systems in this invention are not limited to aerospace systems, unmanned systems, or marine systems.

[0081] As one aspect of this invention, a device for generating and analyzing multiple schemes for complex systems based on the SysML model is proposed. Under the premise of considering the structural parameters and performance parameters of complex systems and the comprehensive system design constraints, the device generates a large number of design schemes for complex systems with optimal structure and cost, and performs multiple scheme trade-off analysis. The device mainly includes a module for generating a large number of complex system schemes based on the NSGA-II (non-dominated sorting genetic algorithm II with elitist strategy) multi-objective optimization algorithm and a module for analyzing multiple schemes based on the TradeStudy mechanism. The two modules are integrated by constructing a standard data transmission channel between them, and together they constitute a device for generating and analyzing multiple schemes for complex systems.

[0082] (1) Module for generating massive solutions for complex systems based on NSGA-II: Based on the design variables and design constraints of complex systems defined by the SysML model, the module obtains the design variables and design constraints of complex systems through the data transmission channel, constructs the optimization model and multi-objective optimization algorithm, generates massive solutions using the multi-objective optimization algorithm, and synchronizes them to the multi-solution trade-off analysis module through the data transmission channel for solution trade-off analysis.

[0083] The module for generating massive solutions for this complex system mainly includes:

[0084] The algorithm parameter design submodule reads the design variables and constraints defined in the multi-scheme trade-off analysis module via a data transmission channel for model optimization. It also allows designers to manually modify design variables and set algorithm parameters through an interactive interface, directly contributing to model optimization. In this invention, design variables include structural parameters and performance parameters.

[0085] The optimization design submodule based on the optimization model: This submodule has a built-in multi-objective optimization algorithm. By receiving the algorithm parameters, design variables and design constraints passed by the algorithm parameter design submodule, it automatically constructs an optimization model. The optimization model includes at least a solution space, objective function and design constraints. It generates a large number of solutions for complex systems through the multi-objective optimization algorithm and outputs a large number of undecoded solutions.

[0086] The scheme generation and data push submodule receives a large number of undecoded schemes, obtains decoded schemes through decoding, and the decoded schemes include the values ​​of design variables. The schemes are visualized through tables and scatter plots, and the large number of schemes are pushed to the multi-scheme trade-off analysis module through the data transmission channel.

[0087] Data transmission channel: used to connect the complex system's massive solution generation module and multi-solution trade-off analysis module, to transmit data and output data to local data storage in a standard format, which is an Excel or CSV format agreed upon by both modules.

[0088] (2) Multi-scheme trade-off analysis module based on TradeStudy: Based on the massive schemes generated from the massive scheme generation module of complex system, the scheme library is generated by mapping the data to the multi-scheme trade-off analysis module, and the simulation data configuration is realized through the trade-off analysis configuration sub-module. By setting the trade-off analysis criteria, the optimal design of multiple schemes of complex system is realized.

[0089] The multi-option trade-off analysis module includes:

[0090] The design variable definition submodule uses a module definition diagram based on the SysML model to define the design variables and constraints of the complex system. By capturing the designer's knowledge, it outputs the composition structure and structural parameters of the complex system.

[0091] Device driving and parameter transmission submodule: Based on the design variables and constraints of the complex system, a parameter diagram is created using a SysML model. The SysML model constraint module is embedded with Python language to drive the generation module of massive solutions for complex systems.

[0092] The scheme library acquisition and generation submodule acquires a massive number of decoded schemes through the data transmission channel and maps these massive numbers of schemes into a system scheme library that exists as instance elements.

[0093] The tradeoff analysis configuration submodule builds an encapsulation module that inherits from the TradeStudy module through generalization. It encapsulates the complex system through association, sets the tradeoff analysis objectives and the criteria for evaluating each solution, generates other tradeoff analysis auxiliary parameter solutions through parameter scanning, and retrieves solutions and parameters from the solution library using reference elements in the internal module graph of the SysML model and outputs them to the TradeStudy module.

[0094] The TradeStudy module has a built-in TradeStudy mechanism that receives solutions and parameters from a solution library. Based on the tradeoff analysis objectives and standards, and combined with the encapsulated complex system, it uses a SysML model to simulate the configuration diagram and perform multi-solution tradeoff analysis of the complex system, outputting the optimal solution design from the solution library. The TradeStudy module includes extended value attributes "OutOfSpec:Real", "N:Integer", "score:Real", and "winner:String". These extended value attributes are used to store values ​​during the tradeoff analysis process. The TradeStudy module uses the constructed type "Analysis" for calculation and analysis.

[0095] This invention proposes a device for generating and weighing multiple solutions for complex systems based on SysML models. It breaks through the bottleneck in the process of designing solutions from requirements to solutions based on SysML models, and solves the problems of long time consumption, low efficiency and low interpretability caused by the need for manual construction and analysis of solutions when designing and weighing multiple solutions based on SysML models. This optimizes the overall process of complex system design.

[0096] As another aspect of the present invention, a method for generating and weighing multiple options for complex systems based on the SysML model is proposed, which specifically includes the following steps:

[0097] S1: The module for generating massive complex system solutions based on NSGA-II is invoked through the parametric graph in the SysML model. The design variables and constraints of the complex system defined in the SysML model are obtained via a data transmission channel. Then, the multi-objective optimization algorithm built into the module is used to set optimization objectives and generate a massive number of design solutions. The specific process is as follows: Figure 2 As shown.

[0098] S1.1 Define Design Variables and Constraints: Based on the design requirements of complex system design, determine the design variables and constraints of the complex system. Define the design variables and constraints based on the module definition diagram in the SysML model. For example, determine parameters such as wingspan, mean chord length, and wing material thickness in aircraft system design as design variables for optimization design, and define the design space.

[0099] S1.2 Module Call for Massive Solution Optimization Design of Complex Systems Based on Parametric Graphs: Based on design variables and constraints defined in the SysML model, a parametric graph is created using the SysML model, simultaneously displaying the corresponding design variables and constraints. A constraint module is created, and the Python language is embedded within the constraint module to implement the module call for the massive solution optimization design system for complex systems based on NSGA-II. The storage and transfer of design variables and constraints are achieved through binding connectors, such as... Figure 3 As shown.

[0100] S1.3 Multi-objective optimization algorithm parameter setting based on a visual interface: In the algorithm parameter setting submodule, algorithm parameters are displayed and customized input in a tree structure. Corresponding algorithm parameters, such as the maximum number of generations, population size, and the number of optimization objectives, are obtained through dynamic interaction. Figure 4 As shown.

[0101] S1.4 Acquisition of Design Constraint Parameters for Complex Systems: The algorithm parameter setting submodule reads the design variables and constraints already defined in the SysML model. The acquired design variables and constraints are then displayed, stored, and set in the algorithm parameter design submodule, such as... Figure 4 As shown.

[0102] S1.5 Solution Space Determination Based on Complex System Design Variables: The results of the complex system massive solution generation module should include the structural parameters and performance parameters of the complex system. The dimension of the solution is the sum of all structural and performance parameters. The solution space takes the following values: As shown below, where n is the sum of the number of structural parameters and performance parameters. Taking the optimized design of an aircraft system as a specific example, the design variables are the structural parameter wingspan ( ), mean chord length ( ), wing material thickness ( ) and performance parameters cruise speed ( ).

[0103] S1.6 Construction of the Objective Function for Complex System Design: The design process of a complex system is essentially a multi-objective, multi-constraint, and multi-level trade-off decision-making process. System design must achieve an optimal balance between structural rationality and resource utilization efficiency while meeting core indicators such as functionality and performance. Complex system design must minimize costs while ensuring system structural integrity and performance compliance. Therefore, the system optimization objective function needs to cover two interrelated but potentially conflicting main directions: optimal structural design. With minimum cost consumption . in, Let be the objective function. It is the minimum value.

[0104] Using an aircraft as a specific example, this paper constructs an objective function model for optimizing complex systems and optimizes the objective function. For aerodynamic drag, Manufacturing cost: , ,in, Where S is the air density and S is the wing area. denoted as zero-lift drag coefficient, e as start-up efficiency factor, k as induced drag correction coefficient, and W as total aircraft weight. ,in, For the cost of wing skin, The cost of machining the wing beams and ribs.

[0105] S1.7 Constraint Construction for Complex System Optimization Design: The design of complex systems must strictly adhere to key parameter constraints determined by the system's physical principles, operating environment, and engineering implementation, such as balance conditions, size ratios, and resource limitations, while meeting functional requirements, to ensure the feasibility of the system architecture. Using an aircraft as a specific example, constraints are set for lift balance, aspect ratio, thickness-chord ratio, manufacturing feasibility, and velocity-span coupling.

[0106] Lift balance constraints: ,

[0107] Aspect ratio constraint: ,

[0108] Thickness-chord ratio constraint: ,

[0109] Process feasibility constraints: ,

[0110] Velocity-wingspan coupling constraints: ,in, The maximum lift coefficient, Each sets a threshold.

[0111] S1.8, System Structure Optimization Calculation Based on Complex System Optimization Model: The complex system massive solution generation module uses the NSGA-II multi-objective optimization algorithm to generate massive solutions for complex systems. The algorithm principle of NSGA-II is as follows: Figure 5As shown, in the t-th iteration of the population, the parent population Pt contains N individuals. After crossover and mutation operations, N offspring are generated, forming the offspring population Qt. Then, the parent population Pt and the offspring population Qt are merged to form a merged population Rt. Individuals that successfully enter the next generation are selected from the merged population Rt. The individuals in the merged population Rt are divided into multiple levels using a non-dominated sorting algorithm. Then, the individuals in the merged population Rt are selected into the next generation according to the order of their assigned levels. The biggest advantage of NSGA-II is the addition of non-dominated sorting and crowding calculation, which reduces algorithm complexity and fully preserves the characteristics of the population. Individuals with a larger crowding distance will be given priority to enter the next generation population Pt+1.

[0112] S1.9, Generation of Massive Solutions for Complex Systems: The multi-objective optimization algorithm embedded in the optimization design submodule based on the optimization model in the massive solution generation module for complex systems performs the optimization model calculations in S1.5-S1.7 above, generating solutions that satisfy multiple optimization objectives in all solution spaces. After decoding the solutions output by the obtained design schemes, the values ​​of the complex system design variables are output and displayed in the optimization design submodule based on the optimization model, thus completing the optimization design of massive solutions for complex systems based on the NSGA-II optimization algorithm.

[0113] S1.10, Massive Solution Push Based on Complex System Optimization Model: To achieve deep integration of optimization results and system models, this invention further constructs a data-model mapping channel, i.e., a data transmission channel. This data transmission channel automatically maps the structured optimal parameter combination schemes output by the optimization calculation back to the solution library in the original SysML model, transmitting the schemes and objective function values, generating a massive number of optional complex system solutions that meet design requirements, to support subsequent multi-scheme trade-off analysis.

[0114] S2 stores and expresses optimized design schemes output by the module generating massive solutions for complex systems using a solution library. Other auxiliary parameters for trade-off analysis are generated and stored through parameter scanning. The internal module diagram enables the binding of parameters for massive solutions and the setting of multi-solution trade-off analysis. The TradeStudy mechanism performs multi-solution trade-off analysis calculations, and based on the trade-off analysis, a system solution that meets the design requirements is derived. The specific process is as follows: Figure 6 As shown.

[0115] S2.1 Setting up the TradeStudy module: The TradeStudy module has a built-in TradeStudy mechanism. The TradeStudy module extends the value attributes "OutOfSpec:Real", "N:Integer", "score:Real", and "winner:String", and applies the constructor type "Analysis" for calculation and analysis. Among these:

[0116] OutOfSpec: Used to store the percentage of schemes that do not meet the conditions of the tradeoff analysis;

[0117] N: Used for statistical calculation of the number of options involved in the trade-off analysis;

[0118] score: Used to store the target value (maximum value) of the tradeoff analysis criteria;

[0119] winner: Used to store the winning solution that best meets the trade-off analysis criteria among a massive number of solutions.

[0120] S2.2 Inter-module data synchronization based on standard data format: Import standard format data through data transmission channels to directly obtain massive solutions generated by the complex system massive solution generation module that meet design constraints and requirements.

[0121] S2.3 System Multi-Solution Generation Based on SysML Model Instance Table: Create an instance table and set the classifier and range by dragging and dropping model elements from the structure tree. In this instance table, create instance elements based on the massive number of solutions obtained from the complex system massive solution generation module, and build a solution library for multi-solution trade-off analysis using all instance elements.

[0122] S2.4 Setting up multi-option trade-off analysis for complex systems based on SysML model module definition graph: The encapsulation module inherits from the TradeStudy module through generalization relationships, encapsulating the complex system through compositional relationships. The SysML model constraint module determines the objectives of the trade-off analysis and the criteria used to evaluate each option. Criteria may include multiple dimensions such as cost and performance.

[0123] S2.5 Parameter Scanning-Based Trade-off Analysis Auxiliary Parameter Generation: The parameters involved in multi-scheme trade-off analysis are more extensive than those generated by the massive scheme generation module. Parameters that cannot be fully considered and optimized during scheme generation are further trade-off analyzed in this part. The trade-off analysis configuration submodule has built-in "BoundReference" and "designVariable" constructor types, used to extend the "bindingPath", "boundEnd", and "min", "max", and "step" feature attributes, respectively.

[0124] bindingPath: The relative position of the value property linked by the binding linker, which is automatically obtained through the binding linker;

[0125] boundEnd: The value attribute that the linker binds to, which is automatically obtained through the binding linker;

[0126] min: Sets the minimum value for a value attribute;

[0127] max: Sets the maximum value of the value attribute;

[0128] step: The parameter generation step size between the maximum and minimum values ​​of the value attribute.

[0129] By creating value attribute elements from the internal module diagram in the SysML model, applying the "BoundReference" and "designVariable" constructor types, and setting the values ​​of "min," "max," and "step," the value attributes to be analyzed in batches are generated. Connecting these value attributes via binding connectors completes the input of auxiliary parameter schemes for the trade-off analysis. Figure 7 As shown.

[0130] S2.6 Configuration of Trade-off Analysis Scheme Based on Internal Module Graph: An internal module graph is created under the encapsulated module. Parameters from the solution library are obtained using reference attributes in the internal module graph. The reference attributes are constructed using the "alternatives" constructor type. After applying this constructor type, the reference attributes are extended with the "Soures" and "Kind" feature attributes. In "Soures," the corresponding solution library from the instance table is selected, and "Kind" is set as a Table. Parameters from a massive number of solutions in the solution library are obtained through binding connectors, such as... Figure 7 As shown.

[0131] S2.7 Trade-off Analysis Simulation Configuration: Based on the simulation configuration diagram, the target object for the system's multi-option trade-off analysis and the location for saving the simulation analysis results are determined. Simulation configuration elements are created in the SysML model simulation configuration diagram. The "execution Target" feature attribute of the simulation configuration element is set to the "encapsulation module" in S2.4, and the "ResultLocation" feature attribute is set to the pre-created instance table displaying the simulation results. The element classifier of this instance table is still selected as the encapsulation module in S2.4. The multi-option trade-off analysis results are stored in the instance table through simulation calculation.

[0132] This invention constructs a method and apparatus for generating and analyzing complex system solutions that integrates "model-driven, intelligent optimization, and quantitative trade-offs." It addresses existing technical problems in complex system design, such as reliance on human experience for solution generation, insufficient exploration of the design space, strong subjectivity in multi-solution trade-off analysis, poor interpretability, and difficulty in co-optimizing structure and cost. Based on fully considering system functional requirements, performance parameter requirements, and multi-dimensional design constraints (such as physical laws, geometric limitations, and resource boundaries), this invention aims to achieve synergistic optimization of system structural rationality and minimization of life-cycle costs, thereby enhancing the automation and rapid iteration capabilities of forward design for complex systems.

[0133] The features described and / or illustrated above with respect to one embodiment may be used in the same or similar manner in one or more other embodiments, and / or in combination with or in lieu of features in other embodiments.

[0134] It should be emphasized that the term "including / comprises" as used herein refers to the presence of a feature, whole, step, or component, but does not exclude the presence or addition of one or more other features, wholes, steps, components, or combinations thereof.

[0135] Many features and advantages of these embodiments are apparent from this detailed description, and therefore the appended claims are intended to cover all such features and advantages of these embodiments that fall within their true spirit and scope. Furthermore, since many modifications and alterations will readily occur to those skilled in the art, the embodiments of the invention are not intended to be limited to the precise structures and operations illustrated and described, but rather to encompass all suitable modifications and equivalents falling within their scope.

[0136] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

[0137] The parts of this invention not described in detail are techniques known to those skilled in the art.

Claims

1. A SysML model based complex system multi-scenario generation and trade-off analysis apparatus, characterized in that, The generation module comprises: The generation module comprises: The algorithm parameter design submodule is used for reading the design variables and design constraints in the analysis module, and modifying the design variables and setting the algorithm parameters; The optimization design submodule is used for constructing an optimization model according to the received algorithm parameters, design variables and design constraints, storing a multi-objective optimization algorithm, and generating a large number of undecoded schemes of a complex system through the multi-objective optimization algorithm according to the optimization model; The scheme generation and data pushing submodule is used for receiving the large number of undecoded schemes, obtaining decoded large number of schemes through decoding, and pushing the decoded large number of schemes to the analysis module; The analysis module comprises: The design variable definition submodule is used for defining the design variables and design constraints of the complex system; The device driver and parameter transmission submodule is used for outputting the design variables and design constraints, and driving the generation module; The scheme library submodule is used for obtaining the large number of decoded schemes of the generation module, and mapping the large number of schemes into a scheme library existing in the form of instance elements; The trade-off analysis configuration submodule is used for constructing an encapsulation module, the encapsulation module inherits a TradeStudy module through a generalization relationship, encapsulates the complex system through a composition association relationship, sets a trade-off analysis target and a standard for evaluating each scheme, generates a trade-off analysis auxiliary parameter scheme through parameter scanning, and obtains the schemes in the scheme library and outputs them to the TradeStudy module; The TradeStudy module is built-in a TradeStudy mechanism, and is used for receiving the schemes in the scheme library and the trade-off analysis auxiliary parameter scheme, performing complex system multi-scheme trade-off analysis on the encapsulated complex system according to the trade-off analysis target and the standard, and outputting an optimal scheme design in the scheme library.

2. The apparatus of claim 1, wherein, The generation module further comprises a data transmission channel, which is used for connecting the generation module and the analysis module to perform data transmission in a standard format; The algorithm parameter design submodule reads the design variables and design constraints of the design variable definition submodule in the analysis module through the data transmission channel; The scheme generation and data pushing submodule pushes the decoded large number of schemes to the scheme library submodule in the analysis module through the data transmission channel.

3. The apparatus of claim 2, wherein, The design variables comprise structure parameters and performance parameters; The algorithm parameters at least comprise a maximum evolution generation number, a population number and an optimization target number; The optimization model at least comprises a solution space, a target function and a design constraint, the solution space comprises all structure parameters and performance parameters, the target function at least comprises two optimization targets, and the design constraint at least comprises a balance condition, a size ratio and a resource limitation constraint; The data transmission channel is also used for outputting data to a local data storage; The algorithm parameter setting submodule displays and customizes the input of the algorithm parameters in a tree structure.

4. The apparatus of claim 1, wherein, The design variable definition submodule defines the design variables and design constraints of the complex system by using a module definition diagram of a SysML model; The device driving and parameter transmission submodule creates a parameter diagram by using a SysML model according to complex system design variables and design constraints, and generates a module by using a Python language in a SysML model constraint module; The trade-off analysis configuration submodule obtains a scheme output in a scheme library from a TradeStudy module by using a reference element of an internal module diagram in a SysML model; The trade-off analysis configuration submodule generates a trade-off analysis auxiliary parameter scheme by parameter scanning through a value attribute element of the internal module diagram in the SysML model; The trade-off analysis configuration submodule determines a trade-off analysis target and a standard for evaluating each scheme through the SysML model constraint module; The TradeStudy module includes several extended value attributes, which are used for storing values in a trade-off analysis process.

5. The apparatus of claim 4, wherein, The TradeStudy module extended value attributes include: OutOfSpec, which is used for storing a scheme ratio that does not meet a trade-off analysis condition; N, which is used for counting a number of schemes participating in the trade-off analysis; Score, which is used for storing a target value of a trade-off analysis standard; Winner, which is used for storing a winning scheme that meets the trade-off analysis standard most in a mass of schemes.

6. The apparatus of claim 4, wherein, The trade-off analysis configuration submodule includes a BoundReference and a designVariable construction type, which are used for extending a bindingPath, a boundEnd, a min, a max, and a step feature attribute; The bindingPath is a relative position of a value attribute linked by a binding connector, which is automatically obtained through the binding connector; The boundEnd is the value attribute linked by the binding connector, which is automatically obtained through the binding connector; The min is a minimum value setting of the value attribute; The max is a maximum value setting of the value attribute; The step is a parameter generation step between the maximum value and the minimum value of the value attribute; The BoundReference and the designVariable construction type are applied to complete batch generation of the value attribute that needs to be analyzed by setting the values of the min, the max, and the step, to complete input of a trade-off analysis auxiliary parameter scheme by connecting the trade-off analysis value attribute through the binding connector.

7. The apparatus of claim 1, wherein, The scheme generation and data pushing submodule is further used for visualizing a mass of decoded schemes by using a table and a scatter plot; The decoded mass of schemes includes values of design variables; The standard at least includes a cost standard and a performance standard.

8. A method for multi-solution generation and trade-off analysis of complex systems based on SysML models, characterized in that, The method includes the following steps: defining design variables and design constraints of a complex system; setting algorithm parameters of a multi-objective optimization algorithm; constructing a complex system optimization model according to the algorithm parameters, the design variables, and the design constraints; generating a mass of undecoded schemes of the complex system by using the multi-objective optimization algorithm according to the complex system optimization model; obtaining a mass of decoded schemes by decoding; creating a SysML model instance table to create instance elements by using the mass of decoded schemes, and constructing a scheme library; creating an encapsulation module, which inherits a TradeStudy mechanism through a generalization relationship and encapsulates the complex system through a composition association relationship; setting a trade-off analysis target and a standard for evaluating each scheme; generating a trade-off analysis auxiliary parameter scheme through parameter scanning; obtaining a scheme in a scheme library and the trade-off analysis auxiliary parameter scheme, and performing trade-off analysis on the complex system through a SysML model simulation configuration diagram based on a TradeStudy mechanism, according to the trade-off analysis target and the standard, in combination with the encapsulated complex system, and outputting an optimal scheme design in the scheme library.

9. The method of claim 8, wherein, The multi-objective optimization algorithm is NSGA-II. The optimization model at least includes a solution space, a target function and a design constraint, the solution space includes all structure parameters and performance parameters, the target function at least includes two optimization targets, and the design constraint at least includes a balance condition, a size ratio and a resource limit. The optimization targets at least include optimal structure design and minimum cost consumption.

10. The method of claim 9, wherein, The complex system is an aircraft, and the design variables of the aircraft are wingspan, average chord length, wing material thickness and cruising speed. The objective function is: wherein, is the objective function, is the minimum value, is the aerodynamic drag, is the manufacturing cost; the aerodynamic drag, manufacturing cost are respectively: , wherein, is the air density, S is the wing area, is the zero-lift drag coefficient, e is the start-up efficiency factor, k is the induced drag correction factor, W is the total weight of the aircraft, is the wingspan, is the mean chord length, is the wing material thickness, is the cruise speed; The design constraints include: Lift balance constraint: Aspect ratio constraint: Thickness-to-chord ratio constraint: Process feasibility constraint: Velocity-span coupling constraint: wherein is the maximum lift coefficient, are set thresholds, respectively.

Images