Aircraft concept scheme design method and system based on generative artificial intelligence

Abstract

The invention relates to an aircraft concept scheme design method and system based on generative artificial intelligence. The method comprises the following steps: selecting a large language model as a system basic model; constructing an aircraft collaborative design cloud platform; driving the large language model to generate a design scheme in a local calling mode, wherein the design scheme comprises model type selection and cue word selection; weight, pneumatic, performance and modeling professional analysis tools are packaged into a capability unit of a design system, and related modeling, evaluation and verification links comprise appearance, power, pneumatic, weight, RCS and flight performance parts and are deployed and operated in a distributed mode. On the basis of a design project in a collaborative design cloud platform, an aircraft CATIA model, overall parameters, aerodynamic parameters, weight parameters and performance data formed by a design scheme are extracted, and elements therein are displayed. According to the system, the aircraft design efficiency is remarkably improved, and the design cost is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of aircraft design, and in particular to a generative artificial intelligence-based aircraft concept scheme design method and system, which are used for rapid design, evaluation and display of aircraft concept schemes. Background Art

[0002] Existing aircraft design methods involve complex multidisciplinary collaboration, requiring designers to have extensive experience and high-level professional skills. The traditional design process is usually time-consuming and costly, especially when faced with complex new aircraft designs, the efficiency problem is more prominent. Although existing neural network technology has made some progress in aerodynamic and structural design, its application in the conceptual design stage is still limited due to the large amount of data required and the difficulty of model migration.

[0003] The inventors of this application have found that the development of generative artificial intelligence, especially the widespread application of large language models, provides a new method for improving the efficiency of aircraft concept design. Large language models have powerful natural language processing capabilities and can generate high-quality design solutions through prompt words without a large amount of training data. Summary of the invention

[0004] In order to improve the design efficiency and reduce the design cost in the aircraft concept design stage, the present invention provides an aircraft concept scheme design method and system based on generative artificial intelligence, aiming to realize the automated design, evaluation and display of aircraft concept schemes through the powerful natural language processing capabilities of a large language model.

[0005] The aircraft concept design method based on generative artificial intelligence includes: Step S1: Select a large language model as the basic model of the system; Step S2: Build an aircraft collaborative design cloud platform as a data integration, management, and call platform for the system; Step S3: driving the large language model to generate a design solution through local calls, including model type selection and prompt word selection; Step S4: Encapsulate the weight, aerodynamics, performance and modeling professional analysis tools into the capability unit of the design system. The modeling, evaluation and verification involved include the shape, power, aerodynamics, weight, RCS and flight performance parts, which are deployed and operated in a distributed manner; the design system refers to the system that executes the aircraft concept design method based on generative artificial intelligence; Step S5: Based on the design project in the collaborative design cloud platform, the aircraft CATIA model, overall parameters, aerodynamic parameters, weight parameters, and performance data formed by the design scheme are extracted, and the elements therein are displayed.

[0006] According to another aspect of the present invention, there is also provided an aircraft concept design system based on generative artificial intelligence, the system comprising: Solution generation module, used to select a large language model as the basic model of the system; The data integration module is used to build an aircraft collaborative design cloud platform, which serves as the data integration, management, and call platform of the system; The solution evaluation module is used to drive the large language model to generate design solutions through local calls, including model type selection and prompt word selection; The professional analysis tools for weight, aerodynamics, performance and modeling are encapsulated into the capability units of the design system. The modeling, evaluation and verification involved include the shape, power, aerodynamics, weight, RCS and flight performance parts, which are deployed and operated in a distributed manner. Among them, the design system refers to the system that executes the aircraft concept design method based on generative artificial intelligence; The solution display module is used to extract the aircraft CATIA model, overall parameters, aerodynamic parameters, weight parameters, and performance data formed by the design solution based on the design project in the collaborative design cloud platform, and display the elements therein.

[0007] The system architecture, implementation method, beneficial effects and application scenarios of the present invention are described below.

[0008] 1. System architecture The core of the system of the present invention lies in its modular design, which specifically includes: Solution generation module: This module is first deployed locally based on the general open source large language model. The selected model types include multiple parameter versions of Llama3.1 and Qwen2. Then, according to the general prompt word framework of the large language model and the aircraft concept design scheme template, the aircraft concept design prompt word design is carried out. By introducing the Few-Shot method, three types of prompt words are formed: Zero-Shot, 1-Shot, and 5-Shot. By running this module, users can select appropriate models and prompt words according to design requirements to carry out intelligent design of aircraft concept schemes.

[0009] Data integration module: This module integrates the generated design scheme with the collaborative design platform. Design data is stored in a structured manner and can be seamlessly connected with professional design tools through the collaborative platform, supporting multi-user design data sharing and editing.

[0010] Scheme evaluation module: This module conducts a comprehensive evaluation of the generated design schemes. Through the distributed operation of various capability unit components, it realizes the modeling, analysis and evaluation of the preliminary design scheme of the large model. It includes sub-modules such as shape modeling, aerodynamic estimation, weight estimation, power performance calculation, flight performance calculation, and RCS calculation.

[0011] Solution display module: This module obtains design solution data, analyzes and evaluates result data through the data integration module, and makes comprehensive use of 3D models, curves, pie charts, tables and other components to achieve dynamic and real-time display of design solutions and evaluation results, helping users make design decisions and optimizations.

[0012] 2. Implementation method Base model deployment and operation: Localized deployment based on the general open source large language model, establishing a virtual environment and project structure, providing a controlled and isolated environment for running applications, ensuring that its dependencies are separated from other system-wide packages, simplifying dependency management, and helping to maintain consistency between different environments. The selected model types include multiple parameter versions of Llama3.1 and Qwen2.

[0013] Demand analysis and prompt word generation: Based on the general prompt word framework of the large language model and the aircraft concept design template, the aircraft concept design prompt words are designed. By introducing the Few-Shot method, three types of prompt words are formed: Zero-Shot, 1-Shot, and 5-Shot. By running this module, users can select appropriate models and prompt words according to design requirements to perform intelligent design of aircraft concept plans. According to the design requirements input by the user, the prompt word generation module converts the requirements into prompt words suitable for large language model processing. The prompt word content includes design goals, background information, style requirements, etc.

[0014] Design solution generation: Various professional analysis tools need to be encapsulated into capability units. The modeling, evaluation, and verification involved mainly include appearance, power, aerodynamics, weight, RCS, flight performance, etc., which are deployed and operated in a distributed manner. The large language model module automatically generates design solutions such as the aircraft's overall layout, power system selection, three-dimensional modeling, and performance analysis based on the generated prompt words.

[0015] Data integration and evaluation: The design requirements of the large model and the generated overall aircraft design standard plan are stored as a data package in the form of structured data. Each design indicator is regarded as a data item. Each distributed terminal can edit, modify, and upload data in units of data packages, and download and call data in units of data items. The data packages and data items are accompanied by version information. In addition, the performance requirements of the design plan, performance calculation intermediate data, result data, CAD models and other downstream data and intermediate parameters can all be defined, viewed, edited, and downloaded through the platform. The generated design plan is uploaded to the collaborative design platform through the data integration module and integrated with professional design tools to automatically evaluate aerodynamic characteristics, weight estimation, power characteristics, RCS characteristics and flight performance.

[0016] Design scheme display: The visualization display module is based on the design projects in the collaborative design platform, extracts the models and data formed by the design scheme, and displays the elements therein, supporting multiple formats such as pictures, audio, video, curves, time series data, CAD models, CAE models, etc. The scheme evaluation module analyzes the design schemes generated in multiple rounds, selects the best scheme through the optimization algorithm, and outputs the final design results. Through this module, designers can clearly and intuitively obtain the design status of the entire design scheme for comparison, decision-making, and subsequent optimization and detailed design.

[0017] 3. Beneficial effects The system of the present invention significantly improves the efficiency of aircraft concept design through automated design process and multi-module collaboration. The system can generate multiple high-quality design solutions in a short time and supports multi-user collaborative design, which is suitable for rapid concept design of complex aircraft. Through integration with the collaborative design platform, the system can achieve data consistency and integrity throughout the design process, greatly reducing the design cycle and cost.

[0018] 4. Application scenarios The system of the present invention is suitable for the early conceptual design stage of an aircraft, and shows significant advantages, especially when multiple design schemes need to be quickly generated and evaluated. The modular design of the system allows it to flexibly adapt to design tasks of different scales and complexities, and can be seamlessly integrated with existing design tools and processes to ensure design quality and efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 The overall design workflow diagram of the aircraft of the present invention; Figure 2 It is the main tool and interface of the aircraft concept design method based on generative artificial intelligence of the present invention; Figure 3 A list of aircraft design program indicators and requirements of the present invention; Figure 4 This is an example of the overall design prompt word scheme of the aircraft of the present invention; Figure 5 Co-designing a cloud platform data packet interface for the aircraft of the present invention; Figure 6 Generate a front-end window for the solution module of the present invention; Figure 7 The same state diagram of each module system of the collaborative design cloud platform of the present invention; Figure 8 Data integration and evaluation of the overall aircraft design scheme of the present invention; Fig. 9 It is the interface for displaying the design results of the present invention. DETAILED DESCRIPTION

[0020] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0021] The overall design idea of ​​the present invention is described below: according to the textual design requirement description provided by the user, parameter determination, modeling, demand analysis and other tasks are automatically completed to provide an overall aircraft design plan that meets the user's needs. The system includes four major components: a solution generation module, a data integration module, a solution evaluation module, and a solution display module. Among them, the solution generation module calls the locally deployed open source large language model through a Python script, converts the design requirements given by the user into large language model prompts in a specific format, performs overall solution design, and outputs the design results in Json format to complete solution generation. The large language model prompts include structured design requirements from user requirements and 27 aircraft overall solution design parameters, and respond to format constraints in Json format. The large language model prompts include three types: Zero-Shot, 1-Shot, and 5-Shot. Among them, 1-Shot and 5-Shot use classic aircraft design examples as design samples in the form of standard design cases, and embed the prompts of the solution generation module. The data integration module can upload and download data in a distributed manner through a digital bus deployed in the cloud, serving as a carrier and conversion center for data such as user design requirements, large model design solutions, and design solution evaluation results. The scheme evaluation module can obtain design scheme data from the data integration module according to the component-based deployment and automated operation tool software, and perform shape modeling, aerodynamic analysis, weight analysis, power performance calculation, RCS calculation, flight performance evaluation, and return the calculation results to the data integration module. Shape modeling is to use Python and CATIA integration to complete the overall layout and overall arrangement of the aircraft through predefined parametric aircraft wings, fuselage, horizontal tail, vertical tail, canard and other components, and receive the design parameters of the scheme generation module to achieve rapid three-dimensional modeling of the shape. Among them, aerodynamic estimation, weight estimation, power performance, and flight performance calculations use Raymer engineering calculation methods, and Python is used to develop calculation scripts to achieve characteristic data estimation. RCS calculation uses numerical calculation methods such as physical optics, combined with the three-dimensional model of the aircraft, to achieve automatic calculation of the arithmetic mean and geometric mean of RCS in the head direction, circumferential direction, and omnidirectional direction. The design scheme display module can visualize the design scheme and evaluation results through three-dimensional models, curves, pie charts, tables, and other forms to achieve the display of solutions. The visualization data of this module comes from the data integration module and is updated in real time according to the data version. After the design is completed, the design results are returned for the next round of iterative optimization design.

[0022] like Figure 1 , Figure 2 As shown, the present invention provides an aircraft concept design method based on generative artificial intelligence, comprising: Step S1: Select a suitable large language model as the basic model of the system; Step S2: Build an aircraft collaborative design cloud platform as a data integration, management, and call platform for the system; Step S3: driving the large language model to generate a design solution through local calls, including model type selection and prompt word selection; Step S4: encapsulate professional analysis tools such as weight, aerodynamics, performance and modeling into capability units of the design system. The modeling, evaluation and verification involved mainly include shape, power, aerodynamics, weight, RCS, flight performance and other parts, which are deployed and operated in a distributed manner; the design system refers to a system that executes the aircraft concept design method based on generative artificial intelligence; Step S5: Based on the design project in the collaborative design cloud platform, the aircraft CATIA model, overall parameters, aerodynamic parameters, weight parameters, and performance data formed by the design scheme are extracted and the elements thereof are displayed. In its implementation, the elements can be displayed in various formats such as pictures, audio, video, curves, time series data, CAD models, CAE models, etc.

[0023] In step S1, a suitable large language model is selected as the basic model of the system, including: Step S1.1: Select Llama3.1 and Qwen2 models as the system base models: In terms of model parameter quantity, Llama3.1 and Qwen2 provide multiple versions with large and small parameters for local deployment. The demand for local deployment mainly comes from the high confidentiality requirements of the aircraft design task. Considering the limitations of hardware conditions, the system provides two versions with smaller parameter quantities, Llama3.1-8B and Qwen2-7B, and two versions with larger parameter quantities, Llama3.1-405B / Qwen2-72B. The diverse model selection can fully adapt to the system operation requirements under different hardware conditions.

[0024] Step S1.2: Based on the general prompt word framework, design the overall design standard scheme of the aircraft, which is specifically implemented as follows: The aircraft overall design standard scheme prompt word framework adopts the general prompt word framework. Figure 3 The design scheme indicators shown are as follows: Figure 4 shown.

[0025] Given the complexity of the overall aircraft design, only Figure 4The prompt words shown may not achieve a good enough design effect, so the few-shot method is used to optimize the prompt words. Few-shot is a common method in the application field of large language models. It can significantly improve the model task capability by adding a small number of samples to the prompt words. Therefore, the prompt word framework is expanded using classic aircraft design cases, and the sample (SHOT) part is added to enhance the aircraft design capability of the general large language model. Due to the limitation of input tokens, two cases, 1-Shot and 5-Shot, are designed.

[0026] In step S2, an aircraft collaborative design cloud platform is constructed as a data integration, management, and call platform for the system, including: Step S2.1: Build an aircraft collaborative design cloud platform, which is specifically implemented as follows: Since the present invention needs to realize the overall aircraft design based on the large language model, it not only includes the generation of the overall plan based on the large language model, but also covers upstream and downstream links such as demand analysis and plan evaluation. In order to ensure the availability and consistency of data, an aircraft collaborative design cloud platform is constructed as the system's data integration, management, and calling platform.

[0027] The core idea is to define and save all design products in each stage of aircraft design in the form of structured data, and to store all data and models in the form of project-viewpoint-model and data-element. Correspondingly, the data models generated by each distributed terminal will also interact with the aircraft design collaboration platform in the form of project-data package-data item-element.

[0028] Step S2.2: storing the aircraft overall design standard plan as a data package in the form of structured data, which is specifically implemented as follows: like Figure 5 As shown, using the aircraft collaborative design cloud platform, the aircraft overall design standard plan defined in step S1.2 is stored as a data packet in the form of structured data, where each design indicator is treated as a data item. Each distributed terminal can edit, modify, and upload data in units of data packets, and download and call data in units of data items. The data packets and data items are accompanied by version information.

[0029] In addition, the performance requirements of the design plan, performance calculation intermediate data, result data, CAD models and other downstream data and intermediate parameters can all be defined, viewed, edited and downloaded through the aircraft collaborative design cloud platform.

[0030] In step S3, the large language model is driven by local calls to generate solutions, including model type selection and prompt word selection, including: Step S3.1: Call the large language model selected in step S1.1. The specific implementation is as follows: In this aircraft concept design method, a self-written Python script is used to call the model, and it is packaged into an application module in the self-developed integrated development environment (IDE) for functional integration. The interface after integration is as follows Figure 6 shown.

[0031] Step S3.2: Run the solution generation module, make a data request through the data interface of the aircraft collaborative design cloud platform, and obtain the latest version of the design requirements.

[0032] After analyzing the design requirements, the prompt word splicing is completed according to the prompt word type selected by the designer, and then the corresponding large language model is called to complete the design scheme. After the design is completed, the design results output by the large language model will be uploaded to the aircraft collaborative design cloud platform in the form of structured data, such as Figure 8 As shown in the figure, the design solution update is completed and awaits subsequent modeling, evaluation, and verification.

[0033] In step S4, various professional analysis tools are encapsulated into capability units of the design system. The modeling, evaluation, and verification involved mainly include appearance, power, aerodynamics, weight, RCS, flight performance, etc., which are deployed and operated in a distributed manner, such as Figure 7 As shown, including: Step S4.1: Define the aircraft shape using a parametric model. The specific implementation is as follows: In order to realize the automatic operation of the system, the parametric model is used to define the aircraft shape. Through manual modeling, several parametric models of typical aircraft configurations are predefined in CATIA. After the design results are given in the large model, the parametric model is instantiated in the form of parameter update to realize the automatic modeling of the aircraft shape.

[0034] Step S4.2: Determine the overall dimensions and weight characteristics of the engine by parameter matching in a predefined database.

[0035] In the traditional aircraft design process, the power system is generally not designed separately, but the engine is selected according to the overall performance requirements. Therefore, the engine is not designed in detail in the present invention, and only the maximum static thrust at sea level required by the aircraft is given. In this case, the engine's overall dimensions and weight characteristics are determined by matching parameters in a predefined database. Thrust characteristics, fuel consumption characteristics, etc. are derived from public data. For engines with undisclosed data, they are estimated through engineering algorithms.

[0036] Step S4.3: Perform pneumatic engineering estimation using pneumatic engineering estimation software.

[0037] like Figure 8 As shown in the upper left figure, the software can quickly calculate the lift-drag characteristic curve of the aircraft at different Mach numbers by inputting the main aerodynamic shape parameters of the aircraft. However, the software only has a good estimation effect for normal layout aircraft and canard layout aircraft, and because it ignores the aerodynamic shape details of the aircraft, the differences between different design schemes are relatively small.

[0038] Step S4.4: Estimate the weight of the structure and the weight of the load.

[0039] The weight of an aircraft mainly includes the structure weight, fuel weight, load weight, etc. Since the large language model has given the expected fuel weight that meets the range requirements when designing the scheme, it is only necessary to estimate the structure weight and load weight. In this invention, the classic Raymer method is used for weight estimation, such as Figure 8 As shown in the upper right figure, the algorithm is encapsulated as a component in the form of a Python script and can be directly called in the design process.

[0040] Step S4.5: Calculate the RCS using numerical simulation method.

[0041] RCS calculation adopts numerical simulation method and is calculated in the software RCSPlus, such as Figure 8 As shown in the lower left figure. By setting the number of grids, radar wavelength, and angle calculation range, the software can automatically perform grid division, RCS calculation, and output the omnidirectional RCS curve of the whole aircraft, the arithmetic mean and geometric mean of RCS within the selected angle range, etc. Since the calculation process requires a 3D model in STL format as input, this submodule needs to be downstream of the 3D modeling submodule in the entire solution evaluation module.

[0042] Step S4.6: Calculate the flight performance using engineering estimation methods.

[0043] Flight performance is the part that is most closely related to the required parameters, including range, maximum flight speed, cruising speed, maximum climb rate, take-off and landing distance, etc. The input parameters required to calculate flight performance include aerodynamics, weight, and power characteristics, etc. Therefore, this submodule needs to be the downstream of the aerodynamic, weight, and power characteristics evaluation submodules, and the corresponding flight performance is calculated through a python script. The calculation results are as follows: Figure 8 It should be pointed out that under engineering estimation conditions, it is extremely difficult to obtain accurate results of flight performance, but it can still be used as an evaluation standard to provide a relative reference for the solutions generated by different models.

[0044] In step S5, based on the design project in the collaborative design platform, the model and data formed by the design solution are extracted, and the elements therein are displayed, including: like Fig. 9 As shown in the figure, the visualization display module is based on the design project in the aircraft collaborative design cloud platform, extracts the model and data formed by the design scheme, and displays the elements therein, supporting multiple formats such as pictures, audio, video, curves, time series data, CAD models, CAE models, etc. Through this module, designers can clearly and intuitively obtain the design status of the entire design scheme for comparison, decision-making, and subsequent optimization and detailed design.

[0045] According to an embodiment of the present invention, there is also provided an aircraft concept scheme design system based on generative artificial intelligence, comprising: Solution generation module, used to select a large language model as the basic model of the system; The data integration module is used to build an aircraft collaborative design cloud platform, which serves as the data integration, management, and call platform of the system; The solution evaluation module is used to drive the large language model to generate design solutions through local calls, including model type selection and prompt word selection; The professional analysis tools for weight, aerodynamics, performance and modeling are encapsulated into the capability units of the design system. The modeling, evaluation and verification involved include the shape, power, aerodynamics, weight, RCS and flight performance parts, which are deployed and operated in a distributed manner. Among them, the design system refers to the system that executes the aircraft concept design method based on generative artificial intelligence; The solution display module is used to extract the aircraft CATIA model, overall parameters, aerodynamic parameters, weight parameters, and performance data formed by the design solution based on the design project in the collaborative design cloud platform, and display the elements therein.

[0046] For the parts of the above system not described in detail, please refer to the description of the method implementation method.

[0047] The parts not described in detail in the present invention belong to the well-known technology in the art.

[0048] The above descriptions are only some specific implementation methods of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by any person familiar with the art within the technical scope disclosed in the present invention should be covered within the protection scope of the present invention.

Claims

1. A method for designing aircraft concepts based on generative artificial intelligence, characterized in that: include: Step S1: Select a large language model as the basic model of the system; Step S2: Build an aircraft collaborative design cloud platform as a data integration, management, and call platform for the system; Step S3: driving the large language model to generate a design solution through local calls, including model type selection and prompt word selection; Step S4: Encapsulate the weight, aerodynamics, performance and modeling professional analysis tools into the capability unit of the design system. The modeling, evaluation and verification involved include the shape, power, aerodynamics, weight, RCS and flight performance parts, which are deployed and operated in a distributed manner; the design system refers to the system that executes the aircraft concept design method based on generative artificial intelligence; Step S5: Based on the design project in the collaborative design cloud platform, the aircraft CATIA model, overall parameters, aerodynamic parameters, weight parameters, and performance data formed by the design scheme are extracted, and the elements therein are displayed.

2. The aircraft concept design method based on generative artificial intelligence according to claim 1 is characterized in that: In step S1, a large language model is selected as the basic model of the system, including: Step S1.1: Select Llama3.1 and Qwen2 models as the basic models of the system; Step S1.2: Based on the general prompt word framework, design a standard plan for the overall design of the aircraft.

3. The aircraft concept design method based on generative artificial intelligence according to claim 2 is characterized in that: In step S2, an aircraft collaborative design cloud platform is constructed as a data integration, management, and call platform for the system, including: Step S2.1: Build an aircraft collaborative design cloud platform: define and save all design products at all stages of aircraft design in the form of structured data, and store all data and models in the form of project-viewpoint-model and data-element; Step S2.2: Store the aircraft overall design standard plan in the form of structured data as a data package.

4. The aircraft concept design method based on generative artificial intelligence according to claim 3 is characterized in that: In step S3, the large language model is driven by local calls to generate solutions, including model type selection and prompt word selection, including: Step S3.1: Call the large language model selected in step S1.1; Step S3.2: Run the solution generation module, make data requests through the data interface of the aircraft collaborative design cloud platform, and obtain the latest version of the design requirements; after parsing the design requirements, complete the prompt word splicing according to the prompt word type selected by the designer, and then call the corresponding large language model to complete the solution design.

5. The aircraft concept design method based on generative artificial intelligence according to claim 4 is characterized in that: In step S4, the weight, aerodynamics, performance and modeling professional analysis tools are encapsulated into the capability unit of the design system. The modeling, evaluation and verification involved include the shape, power, aerodynamics, weight, RCS and flight performance parts, which are deployed and operated in a distributed manner, including: Step S4.1: Using a parametric model to define the aircraft shape: To realize the automatic operation of the system, the parametric model is used to define the aircraft shape. After the large model gives the design result, the parametric model is instantiated in the form of parameter update to realize automatic modeling of the aircraft shape. Step S4.2: Determine the overall dimensions and weight characteristics of the engine by parameter matching in a predefined database; Step S4.3: Performing pneumatic engineering estimation using pneumatic engineering estimation software; Step S4.4: Estimate the weight of the structure and the weight of the load; Step S4.5: Calculate RCS using numerical simulation method; Step S4.6: Calculate the flight performance using engineering estimation methods.

6. The aircraft concept design method based on generative artificial intelligence according to claim 5 is characterized in that: In step S5, based on the design project in the collaborative design platform, the model and data formed by the design solution are extracted, and the elements therein are displayed, including: The display formats include at least one of the following: pictures, audio, video, curves, time series data, CAD models, and CAE models.

7. An aircraft concept design system based on generative artificial intelligence, characterized in that: include: Solution generation module, used to select a large language model as the basic model of the system; The data integration module is used to build an aircraft collaborative design cloud platform, which serves as the data integration, management, and call platform of the system; The solution evaluation module is used to drive the large language model to generate design solutions through local calls, including model type selection and prompt word selection; The professional analysis tools for weight, aerodynamics, performance and modeling are encapsulated into the capability units of the design system. The modeling, evaluation and verification involved include the shape, power, aerodynamics, weight, RCS and flight performance parts, which are deployed and operated in a distributed manner. Among them, the design system refers to the system that executes the aircraft concept design method based on generative artificial intelligence; The solution display module is used to extract the aircraft CATIA model, overall parameters, aerodynamic parameters, weight parameters, and performance data formed by the design solution based on the design project in the collaborative design cloud platform, and display the elements therein.

Images