Intelligent naval vessel structure construction method and system based on AI large model

Abstract

The invention belongs to the technical field of ship structure design, and particularly discloses a naval vessel structure intelligent construction method and system based on an AI large model, and the method comprises the steps: obtaining naval vessel structure design demand content; the naval vessel structure design requirement content is input to a naval vessel structure design professional large model, a naval vessel structure design scheme output by the large model is obtained, and the naval vessel structure design scheme is used for constructing a naval vessel structure matched with the naval vessel structure design requirement content; wherein the naval vessel structure design major large model is obtained based on the naval vessel structure design major knowledge base training large model, and the naval vessel structure design major knowledge base comprises a naval vessel structure element data set and a naval vessel structure design specification base. The naval vessel structural element data set is constructed based on a naval vessel structural design scheme sample and naval vessel structural element data, and the naval vessel structural element data is used for marking key features in the naval vessel structural design scheme sample. According to the invention, intelligent design of a naval vessel structure can be realized.

Description

Technical Field

[0001] The present application belongs to the technical field of ship structure design, and more specifically, to a method and system for intelligently constructing a ship structure based on an AI large model. Background Art

[0002] The traditional ship structure design industry has the following problems, which need to be solved urgently to fully release the potential of ship design and manufacturing and meet actual needs: 1. Ship design is a complex system engineering with multiple disciplines, involving multiple fields such as structure, hydrostatics, hydrodynamics, control theory, and system design. The traditional design process does not fully consider the interaction and integration between these disciplines. 2. The traditional tools and methods relied on by the existing design process limit design innovation and efficiency. 3. The ship design and construction specifications referenced by traditional ship structure design are difficult to apply to the design requirements of the new generation of ships. 4. As mission requirements change, ships may need to quickly adapt to new use environments and tasks. The traditional design process may not take into account this flexibility, resulting in large-scale modifications of ships when facing new tasks, which increases costs and time. 5. Since the design process is decentralized, multiple revisions may be required during the design process, which not only increases the workload, but may also lead to an extension of the design cycle and an increase in costs. How to achieve intelligent design of ship structures is a technical problem that needs to be solved in this field. Summary of the invention

[0003] In view of the defects of the prior art, the purpose of this application is to realize the intelligent design of ship structure.

[0004] To achieve the above objectives, in a first aspect, the present application provides a method for intelligently constructing a ship structure based on an AI large model, the method comprising: Obtain the requirements for ship structure design; Input the ship structure design requirements into the ship structure design professional large model, obtain the ship structure design plan output by the ship structure design professional large model, and the ship structure design plan is used to construct a ship structure that is compatible with the ship structure design requirements; Among them, the big model of ship structure design is obtained by training the big language model (LLM, referred to as the big model) based on the ship structure design professional knowledge base. The ship structure design professional knowledge base includes a ship structure element data set and a ship structure design specification library. The ship structure element data set is based on ship structure design scheme samples (including electronic design drawings, technical documents, historical data (simulation data, measured data), etc.) and ship structure element data. The ship structure element data is used to mark the key features in the ship structure design scheme samples (to ensure that the big model can understand the structure of the design scheme samples and the functions of different parts of the samples). The ship structure design specification library is constructed based on the ship structure design specification knowledge.

[0005] It can be understood that in the model training stage, a ship structure element data set can be constructed based on the ship structure design plan samples and the ship structure element data. The ship structure design plan samples may include electronic design drawings, technical documents, historical data (simulation data, measured data), etc. The ship structure element data is used to mark the key features in the ship structure design plan samples, which can ensure that the large model can understand the structure of the design plan samples and the functions of different parts of the samples during the training process. Based on the ship structure design specification knowledge, a ship structure design specification library can be constructed. The ship structure element data set and the ship structure design specification library are used to train the large model. The obtained ship structure design professional large model has the ability to construct ship structures, that is, the ability to generate ship structure design plans that are compatible with the content of ship structure design requirements under the constraints of ship design specification knowledge. During the model application stage, by inputting the ship structure design requirements into the ship structure design professional large model, the ship structure design plan output by the ship structure design professional large model can be obtained. This design plan can assist designers to efficiently construct a ship structure that is compatible with the ship structure design requirements and realize the intelligent design of the ship structure.

[0006] Here we explain the requirements for ship structure design, which refers to the specific requirements and standards for specific ship types and uses during the ship design process. These requirements cover the functionality, performance indicators, size specifications, material selection, load capacity, safety and environmental adaptability of the ship.

[0007] Exemplarily, the ship structure design plan includes design documents and electronic design drawings. The design document is the text description part of the ship structure design plan, which records in detail the process, principles, calculation basis, design key points and related technical requirements of the ship structure design. The electronic design drawing is an important part of the ship structure design plan. It is a visual expression of the design document, including but not limited to the following contents: general layout drawing, structural layout drawing, detailed design drawing, construction drawing, etc.

[0008] Here, the process of generating electronic design drawings from a large model is exemplified. The large model of ship structure design can extract key features from the generated design documents, including but not limited to: hull shape, main hull dimensions, hull frame form, shape, material, position coordinates of various components that make up the hull frame, and connection forms between components. Then, the large model can use the extracted key features to generate graphic elements (such as lines, shapes, symbols, etc., which can represent different parts of the ship structure), coordinate positions of graphic elements, text annotation content (including annotations, dimension annotations, material descriptions, etc.) and the correspondence between text annotation content and graphic elements, and then call the API of computer-aided design (CAD) software to draw various geometric figures (such as lines, circles, polygons, etc.) according to the graphic elements and the corresponding coordinate positions, set attributes (such as color, line type, fill style, etc.) for the drawn graphic elements according to the text annotation content and the correspondence between the text annotation content and the graphic elements, and add text annotations (such as dimension annotations) on the drawings, and then obtain electronic design drawings.

[0009] In a possible implementation, the large model of ship structure design is obtained through the following steps: Based on the ship structure design scheme sample, ship structure element data and ship design specification knowledge, a question-answer pair sample is constructed. The question-answer pair sample is used to describe the ship structure design scheme that is compatible with the ship structure design requirements under the constraints of ship design specification knowledge. Based on the question-answer pair samples, a large model is trained to obtain a professional large model of ship structure design.

[0010] Exemplarily, the question-answer pair sample includes a question part and an answer part, wherein the question part is used to describe the content of the ship structure design requirements, and the answer part is used to provide a corresponding ship structure design plan under the constraints of ship design specification knowledge.

[0011] It can be understood that by using question-answer pair samples to train the big model, the professional big model of ship structure design obtained has the ability to construct ship structures, that is, the ability to generate ship structure design plans that are compatible with the content of ship structure design requirements under the constraints of ship design specification knowledge.

[0012] In a possible implementation, the method further includes: Perform graphic recognition and text recognition on the drawing image to obtain graphic information and text information corresponding to the drawing image; Based on the graphic information and text information corresponding to the drawing image, the CAD format data corresponding to the drawing image is obtained by calling the API of the CAD software; Based on the preset verification configuration, the CAD format data corresponding to the drawing image is verified, and the CAD format data that passes the verification is used as the electronic design drawing corresponding to the drawing image; Update the ship structure element dataset based on the electronic design drawings corresponding to the drawing images.

[0013] The drawing image is an image of a drawing. The drawing image may be an image obtained by photographing a paper drawing. The drawing image may also be extracted from various electronic documents (such as a PDF document containing a drawing image).

[0014] Specifically, the drawing image can be subjected to graphic recognition and text recognition through the machine learning model to obtain the graphic information and text information corresponding to the drawing image. The graphic information includes graphic elements (such as lines, shapes, symbols, etc., which can represent different parts of the ship structure) and the coordinate positions of the graphic elements. The text information includes text annotation content (including annotations, dimension annotations, material descriptions, etc.) and the correspondence between the text annotation content and the graphic elements. Then, based on the graphic information and text information corresponding to the drawing image, various geometric figures are drawn according to the graphic elements and the corresponding coordinate positions by calling the API of the CAD software. According to the text annotation content and the correspondence between the text annotation content and the graphic elements, the attributes of the drawn graphic elements are set and text annotations are added to the drawings, thereby obtaining the CAD format data corresponding to the drawing image. Then, based on the preset verification configuration, the CAD format data corresponding to the drawing image can be verified to ensure that the generated CAD format data meets the preset verification configuration. The verified CAD format data can then be used as the electronic design drawing corresponding to the drawing image, and the ship structure element data set can be updated based on the electronic design drawing corresponding to the drawing image. As the sample size of the ship structure element data set gradually increases, the large model can be regularly fine-tuned and trained to improve the accuracy of the large model output solution.

[0015] In a possible implementation, the ship structure design scheme output by the ship structure design professional large model includes: design documents and electronic design drawings; After obtaining the ship structure design plan output by the large model of the ship structure design major, it also includes: Based on the design documents and electronic design drawings, simulate and analyze the ship structure and obtain the ship structure simulation data; Based on the ship structure design requirements and ship design specification knowledge, the ship structure simulation data is verified.

[0016] Specifically, a three-dimensional simulation model of the ship structure can be constructed based on the acquired design documents and electronic design drawings. The model should accurately reflect all the details of the design, including the geometry, material properties, boundary conditions, etc. of the structure. For example, professional simulation software such as ANSYS, ABAQUS, etc. can be used to establish a finite element model of the ship structure based on the design documents and electronic design drawings. This model can accurately reflect the geometry, material properties and boundary conditions of the actual ship structure. After the simulation model is constructed, the simulation parameters are set: according to the technical requirements in the design documents, the relevant parameters of the simulation analysis are set, including load conditions, environmental factors, material properties, and analysis types (such as static analysis, dynamic analysis, fatigue analysis, etc.). Then, simulation calculations are performed: the simulation software is run to calculate the ship structure model. This process may include multiple steps, such as meshing, solver settings, calculation execution, etc. Then, simulation data is obtained: after the simulation calculation is completed, a series of simulation result data are output, including stress, strain, displacement, vibration frequency, fatigue life, etc.

[0017] The verification process includes the following steps: (1) Data comparison: comparing the simulation data with the design requirements item by item to check whether there are any deviations; (2) Specification review: reviewing whether the simulation data meets the specified safety factor, material strength, structural stability and other requirements according to the ship design specifications.

[0018] In a possible implementation, after verifying the ship structure simulation data, the following is further included: When the ship structure simulation data is verified and passed, the ship structure design scheme is optimized according to the preset optimization goals based on the ship structure simulation data and the ship structure design requirements and under the constraints of the ship design specification knowledge.

[0019] Specifically, the process of optimizing the ship structure design scheme can be divided into the optimization preparation stage and the optimization execution stage.

[0020] The optimization preparation stage includes: (1) determining the optimization objective, (2) determining the design variables, (3) setting constraints, and (4) selecting the optimization algorithm.

[0021] (1) Determine the optimization goal (objective function): According to the requirements of the ship structure design, clarify the optimization goal, which may include reducing the structure weight, improving the structure strength, reducing the manufacturing cost, improving the seakeeping performance, etc.

[0022] (2) Determine design variables: Select design parameters that have a greater impact on the ship's structural performance as optimization variables, such as material thickness, profile size, material type, structural layout, etc.

[0023] (3) Constraint setting: Based on the knowledge of ship design specifications, set the constraints in the optimization process, including structural strength, stability, fatigue life, safety standards, etc.

[0024] (4) Select an optimization algorithm: According to the characteristics of the optimization problem, select an appropriate optimization algorithm, such as genetic algorithm, particle swarm optimization, simulated annealing, gradient descent, etc.

[0025] The optimization execution phase includes: (1) establishing the optimization model, (2) initializing the design parameters, (3) iterative optimization, (4) evaluating the optimization results, and (5) verifying the results.

[0026] (1) Establishing an optimization model: Building an optimization model based on optimization objectives, design variables, and constraints.

[0027] (2) Initialize design parameters: Set initial values ​​for design variables as the starting point for optimization iterations.

[0028] (3) Iterative optimization: Perform simulation analysis on the design scheme in each iterative step to obtain the simulation data of the ship structure; evaluate the performance of the current design scheme based on the simulation results and calculate the value of the objective function; check whether the design scheme meets all constraints. If not, adjustments need to be made; update the values ​​of the design variables according to the optimization algorithm and generate a new design scheme.

[0029] (4) Evaluation of optimization results: After each iteration, evaluate whether the optimization results have converged, that is, whether the objective function value meets expectations or the improvement is less than the preset threshold.

[0030] (5) Result verification: Verify the final optimized design solution to ensure that it meets all design specifications and design requirements.

[0031] In a possible implementation, after optimizing the ship structure design scheme, the following steps are further included: Based on the optimized ship structure design scheme, three-dimensional modeling is performed to obtain a three-dimensional simulation model; Based on the 3D simulation model, the 3D structure is visualized through visualization tools.

[0032] Specifically, 3D modeling software (such as AutoCAD, SolidWorks) can be used to perform 3D modeling to obtain a 3D simulation model corresponding to the optimized ship structure design scheme, and then visualization tools (such as Unity, Blender) can be used to display the appearance and content of the 3D simulation model. Performance data can also be combined with the 3D simulation model to display the performance of the ship using charts, animations, etc.

[0033] In the second aspect, the present application provides a ship structure intelligent construction system based on an AI large model, the system comprising: Design requirement acquisition module, used to obtain the requirements for ship structure design; A design scheme generation module is used to input the ship structure design requirements into the ship structure design professional large model, and obtain the ship structure design scheme output by the ship structure design professional large model. The ship structure design scheme is used to construct a ship structure that is compatible with the ship structure design requirements; Among them, the ship structure design professional big model is obtained based on the ship structure design professional knowledge base training big model. The ship structure design professional knowledge base includes a ship structure element data set and a ship structure design specification library. The ship structure element data set is constructed based on ship structure design scheme samples and ship structure element data. The ship structure element data is used to mark the key features in the ship structure design scheme samples. The ship structure design specification library is constructed based on the ship structure design specification knowledge.

[0034] In a possible implementation, it also includes: a training module and an intelligent drawing recognition module; The training module is used to: Based on the ship structure design scheme sample, ship structure element data and ship design specification knowledge, a question-answer pair sample is constructed. The question-answer pair sample is used to describe the ship structure design scheme that is compatible with the ship structure design requirements under the constraints of ship design specification knowledge. Based on the question-answer pair sample training big model, obtain the big model of ship structure design; The intelligent drawing recognition module is used for: Perform graphic recognition and text recognition on the drawing image to obtain graphic information and text information corresponding to the drawing image; Based on the graphic information and text information corresponding to the drawing image, the CAD format data corresponding to the drawing image is obtained by calling the API of the CAD software; Based on the preset verification configuration, the CAD format data corresponding to the drawing image is verified, and the CAD format data that passes the verification is used as the electronic design drawing corresponding to the drawing image; Update the ship structure element dataset based on the electronic design drawings corresponding to the drawing images.

[0035] In a possible implementation, the ship structure design scheme output by the ship structure design professional large model includes: design documents and electronic design drawings, and the system also includes: a ship structure data calculation and verification module and a design scheme intelligent optimization module; The ship structure data calculation and verification module is used for: Based on the design documents and electronic design drawings, simulate and analyze the ship structure and obtain the ship structure simulation data; Verify the ship structure simulation data based on the ship structure design requirements and ship design specification knowledge; The design scheme intelligent optimization module is used to: When the ship structure simulation data is verified and passed, the ship structure design scheme is optimized according to the preset optimization goals based on the ship structure simulation data and the ship structure design requirements and under the constraints of the ship design specification knowledge.

[0036] In a possible implementation, a design solution intelligent display module is also included, which is used to: Based on the optimized ship structure design scheme, three-dimensional modeling is performed to obtain a three-dimensional simulation model; Based on the 3D simulation model, the 3D structure is visualized through visualization tools.

[0037] In general, the above technical solutions conceived by this application have the following beneficial effects compared with the prior art: In the model training stage, the ship structure element data set and the ship structure design specification library are used to train the large model. The obtained ship structure design professional large model has the ability to construct ship structure. In the model application stage, by inputting the ship structure design requirement content into the ship structure design professional large model, the ship structure design plan output by the ship structure design professional large model can be obtained. This design plan can assist designers to efficiently construct a ship structure that is compatible with the ship structure design requirement content and realize the intelligent design of ship structure. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 It is one of the flow charts of the method for intelligently constructing a ship structure based on an AI large model provided in an embodiment of the present application; Figure 2 This is the second flow chart of the method for intelligently constructing a ship structure based on an AI large model provided in an embodiment of the present application; Figure 3 It is a schematic diagram of the architecture of the intelligent auxiliary system for ship structure design provided in an embodiment of the present application. DETAILED DESCRIPTION

[0039] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application 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 application and are not used to limit the present application.

[0040] In the embodiments of the present application, words such as "exemplary" or "for example" are used to indicate examples, illustrations or descriptions. Any embodiment or design described as "exemplary" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary" or "for example" is intended to present related concepts in a specific way.

[0041] In the description of the embodiments of the present application, unless otherwise specified, "multiple" means two or more than two. For example, multiple processing units refer to two or more processing units, etc.; multiple elements refer to two or more elements, etc.

[0042] The embodiments of the present application are described below in conjunction with the drawings in the embodiments of the present application.

[0043] Figure 1 This is one of the flow charts of the method for intelligently constructing a ship structure based on an AI large model provided in an embodiment of the present application, such as Figure 1 As shown, the method includes the following steps S101 and S102.

[0044] Step S101, obtaining the ship structure design requirements; Step S102, inputting the ship structure design requirement content into the ship structure design professional large model, obtaining the ship structure design scheme output by the ship structure design professional large model, and the ship structure design scheme is used to construct a ship structure that is compatible with the ship structure design requirement content; Among them, the big model of ship structure design is obtained by training the big language model (LLM, referred to as the big model) based on the ship structure design professional knowledge base. The ship structure design professional knowledge base includes a ship structure element data set and a ship structure design specification library. The ship structure element data set is based on ship structure design scheme samples (including electronic design drawings, technical documents, historical data (simulation data, measured data), etc.) and ship structure element data. The ship structure element data is used to mark the key features in the ship structure design scheme samples (to ensure that the big model can understand the structure of the design scheme samples and the functions of different parts of the samples). The ship structure design specification library is constructed based on the ship structure design specification knowledge.

[0045] It can be understood that, in the model training stage, the large model is trained using the ship structure element data set and the ship structure design specification library, and the obtained ship structure design professional large model has the ability to construct ship structures. In the model application stage, by inputting the ship structure design requirements into the ship structure design professional large model, the ship structure design plan output by the ship structure design professional large model can be obtained. This design plan can assist designers to efficiently construct a ship structure that is compatible with the ship structure design requirements and realize the intelligent design of ship structures.

[0046] Figure 2 This is the second flow chart of the method for intelligently constructing a ship structure based on an AI large model provided in an embodiment of the present application, such as Figure 2 As shown, the method includes: (1) Build a professional knowledge base of ship structure design based on AI big model (including ship structure element data set and ship structure design specification library) as the training knowledge base of the big model; (2) Use AI to intelligently identify ship structure drawings and obtain a data set of ship structure elements; (3) Carry out the selection, training and optimization of large AI models; (4) Construct a ship structure data calculation and verification module to calculate and verify the design scheme given by the large model; (5) Construct evaluation criteria and optimization strategies for ship structure design schemes, and intelligently optimize the design schemes given by the large model; (6) Build intelligent 3D modeling display and technical document output functions to facilitate users to view design plans.

[0047] Figure 3 Schematic diagram of the intelligent auxiliary system architecture for ship structure design provided by the embodiment of the present application, such as Figure 3 As shown in the figure, the architecture of the intelligent auxiliary system for ship structure design mainly includes the following seven modules: ship structure element data set, ship structure design specification library, intelligent drawing recognition module, AI large model, ship structure data calculation and verification module, design scheme intelligent optimization module, and design scheme intelligent display module.

[0048] (1) Ship structure element data set: The ship structure data set stores many element data related to the ship structure, including ship performance parameters (overall scale, tonnage, speed, seakeeping and seaworthiness, etc.) and hull structure parameters (hull shape, hull main scale, hull frame form, size, shape, position and connection relationship of each component of the hull, etc.). By extracting the key features of the ship structure as ship structure element data and combining it with the common data structure of the large model, the ship structure element data set is obtained. This data set is used to store a large amount of ship structure element data to provide to the AI ​​large model as input data and the expertise required for training the large model.

[0049] like Figure 3 As shown, the specific scheme of the ship structure element dataset is as follows.

[0050] (1-1) Data collection: Collect existing ship structure design plans. The design plans can be design drawings, design documents, historical data (simulation data, measured data), etc. The design plans include but are not limited to hull size, material properties, power system, speed, displacement, etc.

[0051] (1-2) Data collation: Classify, clean and standardize the collected data to ensure its consistency and accuracy.

[0052] (1-3) Feature extraction: Identify the key features of the ship structure (ship structure element data), such as the hull shape, main dimensions of the hull, hull frame form, the shape, material, position coordinates of the various components that make up the hull frame, and the connection form between the components, and extract these features as labels for the data set.

[0053] (1-4) Dataset construction: Based on the extracted features, a structured dataset is constructed. It can be in tabular form or in a more advanced data structure such as a knowledge graph.

[0054] (1-5) Dataset verification: Verify the accuracy and practicality of the dataset through expert review or actual case testing.

[0055] (1-6) Dataset update: Establish a data update mechanism and regularly add new ship structure data to maintain the timeliness of the dataset.

[0056] (2) Ship structure design specification library: By collecting the contents of ship structure design specifications, a ship structure design specification library is constructed, and the contents of the ship structure design specification library are organized and constructed in layers to meet the design needs of ships at different levels and categories. The specification library acts like a bookcase, where the stored knowledge is classified and placed in it, and reasonable storage rules and retrieval methods are formulated to be suitable for the call of the AI ​​large model. It serves as a training knowledge base during large model training and as a professional knowledge base when the system is officially running to ensure the normal operation of the system.

[0057] like Figure 3 As shown, the specific plan of the ship structure design specification library is as follows.

[0058] (2-1) Knowledge collection: Collect specifications related to ship design, such as design standards, design principles, historical cases, etc.

[0059] (2-2) Knowledge organization: Classify and structure the collected knowledge to form a knowledge base that is easy to retrieve and apply.

[0060] (2-3) Knowledge representation: Using natural language processing technology, the knowledge in the knowledge base is converted into a machine-readable format, such as semantic networks, ontologies, etc.

[0061] (2-4) Knowledge base construction: Develop a knowledge base management system to support the storage, retrieval, updating and maintenance of knowledge.

[0062] (2-5) Knowledge base verification: Verify the accuracy and effectiveness of the knowledge base by designing case tests.

[0063] (2-6) Knowledge base application: Integrate the knowledge base with the ship structure element dataset as training material to train the large model of ship structure design.

[0064] (3) Intelligent drawing recognition module: This module uses intelligent image recognition and text recognition technology to intelligently identify the components, composition and size of the hull in the ship design drawings, and intelligently generate CAD design drawings. The ship structure drawing intelligent recognition technology is used to intelligently identify and verify the existing ship structure drawings, convert paper drawings into electronic data, and include them in the ship structure element data set as input data for the structural design large model. After calculation and optimization of the large model, a design solution that meets the requirements is obtained. The existing ship design drawings and technical data are huge in number, and the training of the large model requires sufficient training data. Therefore, with the help of drawing intelligent recognition technology, high recognition accuracy and fast recognition speed can be achieved.

[0065] The intelligent drawing recognition module mainly uses the following key technologies: (3-1) intelligent image recognition; (3-2) intelligent text recognition; (3-3) intelligent generation of CAD design drawings; (3-4) intelligent auxiliary verification of CAD design drawings. It effectively combines the professional knowledge of ship design with intelligent recognition technology to improve the accuracy and speed of recognition and give full play to its advantages of efficiency and intelligence.

[0066] (4) AI big model: The big model of ship structure design is the core of the system. Whether it can effectively realize the expected functions depends on the training effect of the big model. Therefore, by formulating a reasonable training plan and accumulating sufficient training data, we can ensure the effective training of the big model.

[0067] like Figure 3 As shown, the training plan for the AI ​​large model is as follows.

[0068] (4-1) Technology selection: By analyzing the large models in the current artificial intelligence field, such as Transformer and BERT, we select models suitable for ship structure design.

[0069] (4-2) Model training: The selected model is trained using the ship structure design professional knowledge base (including the ship structure element data set and the ship structure design specification library) so that it can understand the ship structure design process and generate ship structure design solutions.

[0070] (4-3) Model optimization: Through experiments and feedback, the model parameters are continuously adjusted to improve the accuracy and efficiency of the model.

[0071] (4-4) System integration: Integrate the trained model into the ship structure optimization design process to achieve automated design.

[0072] (4-5) Model evaluation: The performance of the model, including design quality, efficiency and innovation, is evaluated through actual design cases such as actual ship design schemes.

[0073] (5) Ship structure data calculation and verification module: It mainly calculates and verifies the design results of the large model, eliminates the schemes that do not meet the technical requirements and design specifications, and provides a basis for subsequent scheme optimization.

[0074] like Figure 3 As shown, the construction scheme of the ship structure data calculation and verification module is as follows.

[0075] (5-1) Collect and organize relevant calculation methods: The objects of calculation include ship structural strength (such as total longitudinal strength, local strength, structural stability, etc.) and ship performance (such as speed, seakeeping and seaworthiness, etc.). Therefore, it is necessary to call the ship structure design specification library to collect and organize relevant design specifications; (5-2) Develop a calculation and verification module for ship structure data with reference to the calculation methods and processes specified in the above documents; (5-3) Develop interfaces for certain industry-wide computing software, such as finite element structural analysis software, to facilitate the import of relevant data into industry-wide software for verification and calculation.

[0076] (6) Design scheme intelligent optimization module: The ship design scheme obtained by large-scale model design may not necessarily meet the technical requirements, so it needs to be further optimized. Selecting reasonable optimization criteria and optimization strategies is the key.

[0077] like Figure 3 As shown in the figure, the specific scheme of the design scheme intelligent optimization module is as follows. (6-1) Constructing ship structure design optimization criteria; (6-2) Constructing ship structure design optimization strategy; (6-3) Evaluation and optimization of design schemes: The ship structure design large model is designed according to the given requirements, and the obtained scheme is calculated and analyzed by the data calculation and verification module, and the schemes that do not meet the requirements are eliminated. The remaining schemes are optimized according to the optimization strategy to obtain further optimized schemes.

[0078] (7) Design scheme intelligent display module: Using the electronic data of the ship structure generated by the ship structure drawing intelligent recognition module, or the technical scheme obtained by the large model design, the three-dimensional modeling of the ship structure is automatically performed, so that the technical scheme can be intuitively presented in the form of three-dimensional graphics.

[0079] The design scheme intelligent display module uses the following key technologies: 3D engine; 3D model intelligent reconstruction.

[0080] Specifically, 3D modeling software (such as AutoCAD, SolidWorks) can be used to perform 3D modeling to obtain a 3D simulation model corresponding to the optimized ship structure design scheme, and then visualization tools (such as Unity, Blender) can be used to display the appearance and content of the 3D simulation model. Performance data can also be combined with the 3D simulation model to display the performance of the ship using charts, animations, etc.

[0081] It will be easily understood by those skilled in the art that the above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present application shall be included in the scope of protection of the present application.

Claims

1. A method for intelligently constructing a ship structure based on an AI large model, characterized in that: include: Obtain the requirements for ship structure design; Input the ship structure design requirements into the ship structure design professional large model, obtain the ship structure design plan output by the ship structure design professional large model, and the ship structure design plan is used to construct a ship structure that is compatible with the ship structure design requirements; Among them, the ship structure design professional big model is obtained based on the ship structure design professional knowledge base training big model. The ship structure design professional knowledge base includes a ship structure element data set and a ship structure design specification library. The ship structure element data set is constructed based on ship structure design scheme samples and ship structure element data. The ship structure element data is used to mark the key features in the ship structure design scheme samples. The ship structure design specification library is constructed based on the ship structure design specification knowledge.

2. The method for intelligently constructing a ship structure based on an AI large model according to claim 1 is characterized in that: The professional large model of ship structure design is obtained through the following steps: Based on the ship structure design scheme sample, ship structure element data and ship design specification knowledge, a question-answer pair sample is constructed. The question-answer pair sample is used to describe the ship structure design scheme that is compatible with the ship structure design requirements under the constraints of ship design specification knowledge. Based on the question-answer pair samples, a large model is trained to obtain a professional large model of ship structure design.

3. The method for intelligently constructing a ship structure based on an AI large model according to claim 1 is characterized in that: Also includes: Perform graphic recognition and text recognition on the drawing image to obtain graphic information and text information corresponding to the drawing image; Based on the graphic information and text information corresponding to the drawing image, the CAD format data corresponding to the drawing image is obtained by calling the API of the CAD software; Based on the preset verification configuration, the CAD format data corresponding to the drawing image is verified, and the CAD format data that passes the verification is used as the electronic design drawing corresponding to the drawing image; Update the ship structure element dataset based on the electronic design drawings corresponding to the drawing images.

4. The method for intelligently constructing a ship structure based on an AI large model according to any one of claims 1 to 3, characterized in that: The ship structure design scheme output by the ship structure design professional large model includes: design documents and electronic design drawings; After obtaining the ship structure design plan output by the large model of the ship structure design major, it also includes: Based on the design documents and electronic design drawings, simulate and analyze the ship structure and obtain the ship structure simulation data; Based on the ship structure design requirements and ship design specification knowledge, the ship structure simulation data is verified.

5. The method for intelligently constructing a ship structure based on an AI large model according to claim 4 is characterized in that: After verifying the ship structure simulation data, it also includes: When the ship structure simulation data is verified and passed, the ship structure design scheme is optimized according to the preset optimization goals based on the ship structure simulation data and the ship structure design requirements and under the constraints of the ship design specification knowledge.

6. The method for intelligently constructing a ship structure based on an AI large model according to claim 5 is characterized in that: After optimizing the ship structure design, it also includes: Based on the optimized ship structure design scheme, three-dimensional modeling is performed to obtain a three-dimensional simulation model; Based on the 3D simulation model, the 3D structure is visualized through visualization tools.

7. An intelligent ship structure construction system based on AI large model, characterized in that: include: Design requirement acquisition module, used to obtain the requirements for ship structure design; A design scheme generation module is used to input the ship structure design requirements into the ship structure design professional large model, and obtain the ship structure design scheme output by the ship structure design professional large model. The ship structure design scheme is used to construct a ship structure that is compatible with the ship structure design requirements; Among them, the ship structure design professional big model is obtained based on the ship structure design professional knowledge base training big model. The ship structure design professional knowledge base includes a ship structure element data set and a ship structure design specification library. The ship structure element data set is constructed based on ship structure design scheme samples and ship structure element data. The ship structure element data is used to mark the key features in the ship structure design scheme samples. The ship structure design specification library is constructed based on the ship structure design specification knowledge.

8. The intelligent ship structure construction system based on AI big model according to claim 7 is characterized in that: Also includes: Training module and intelligent drawing recognition module; The training module is used to: Based on the ship structure design scheme sample, ship structure element data and ship design specification knowledge, a question-answer pair sample is constructed. The question-answer pair sample is used to describe the ship structure design scheme that is compatible with the ship structure design requirements under the constraints of ship design specification knowledge. Based on the question-answer pair sample training big model, obtain the big model of ship structure design; The intelligent drawing recognition module is used for: Perform graphic recognition and text recognition on the drawing image to obtain graphic information and text information corresponding to the drawing image; Based on the graphic information and text information corresponding to the drawing image, the CAD format data corresponding to the drawing image is obtained by calling the API of the CAD software; Based on the preset verification configuration, the CAD format data corresponding to the drawing image is verified, and the CAD format data that passes the verification is used as the electronic design drawing corresponding to the drawing image; Update the ship structure element dataset based on the electronic design drawings corresponding to the drawing images.

9. The intelligent ship structure construction system based on AI big model according to claim 7 or 8 is characterized in that: The ship structure design scheme output by the ship structure design professional large model includes: design documents and electronic design drawings, and the system also includes: a ship structure data calculation and verification module and a design scheme intelligent optimization module; The ship structure data calculation and verification module is used for: Based on the design documents and electronic design drawings, simulate and analyze the ship structure and obtain the ship structure simulation data; Verify the ship structure simulation data based on the ship structure design requirements and ship design specification knowledge; The design scheme intelligent optimization module is used to: When the ship structure simulation data is verified and passed, the ship structure design scheme is optimized according to the preset optimization goals based on the ship structure simulation data and the ship structure design requirements and under the constraints of the ship design specification knowledge.

10. The intelligent ship structure construction system based on AI big model according to claim 9 is characterized in that: It also includes a design solution intelligent display module for: Based on the optimized ship structure design scheme, three-dimensional modeling is performed to obtain a three-dimensional simulation model; Based on the 3D simulation model, the 3D structure is visualized through visualization tools.

Images