CAD model generation method and device based on industrial large model, equipment and medium

Abstract

The invention belongs to the technical field of computers, and particularly relates to a CAD model generation method and device based on an industrial large model, equipment and a medium, and the method comprises the steps that request information of a user side is acquired, and the request information is used for indicating demand information for generating or processing a CAD model by a user; according to the request information, determining design steps corresponding to the request information and an API calling instruction corresponding to each design step through a large model; generating target CAD model data according to the design steps and corresponding API calling instructions; and displaying the design steps on a display interface of the user side, and rendering the target CAD model based on the target CAD model data. According to the method, automatic modeling is achieved through the generalization ability of a large model, the manual burden is relieved, complex design tasks are automated, and the accuracy and efficiency of workflow are improved. The user interface and natural language interaction are simplified, and the user experience and the processing efficiency are improved.

Description

Technical Field

[0001] The present application belongs to the field of computer technology, and specifically relates to a method, device, equipment and medium for generating a CAD model based on an industrial large model. Background Art

[0002] The creation of traditional CAD models (Computer-Aided Design) usually requires manual operation by professional designers, which not only requires the operator to have professional skills and experience, but also limits the efficiency and innovation of the design. Since it relies on manual design step by step, it takes a long time from conceptual design to the generation of the final model. At the same time, manual operation increases the risk of errors in the design process, especially when carrying out complex or detailed designs.

[0003] In addition, the user interfaces of some traditional CAD software are not intuitive enough, which makes it difficult for new users to get started, and the user experience needs to be improved, while hiring professional designers increases the cost burden of the enterprise. Summary of the invention

[0004] In response to the above problems, the present application provides a CAD model generation method, device, equipment and medium based on industrial large models.

[0005] In a first aspect, the present application provides a method for generating a CAD model based on an industrial large model, the method comprising:

[0006] Obtaining request information from a user, where the request information is used to indicate the user's requirement information for generating or processing a CAD model;

[0007] According to the request information, the design steps corresponding to the request information and the API call instructions corresponding to each design step are determined by the big model; wherein there is an association relationship between the request information and the design steps, and the big model is pre-trained based on the association relationship, the preset API call instructions and the corresponding function description;

[0008] Generate target CAD model data according to the design steps and corresponding API call instructions;

[0009] The design steps are displayed on a display interface of the user terminal, and the target CAD model is rendered based on the target CAD model data.

[0010] In a possible implementation, obtaining the design steps corresponding to the request information and the API call instructions corresponding to each design step through the big model according to the request information includes:

[0011] According to the request information and the preset step generation prompt template, a corresponding step generation prompt is obtained, wherein the step generation prompt template is used to prompt the large model to generate design steps for user interaction, and provide an API call instruction corresponding to each design step;

[0012] The step generation prompt is input into the large model to obtain the design steps output by the large model, as well as the single or sequence combination API call instructions corresponding to each design step.

[0013] In a possible implementation, the pre-training of the large model includes:

[0014] Acquire sample data, wherein the sample data includes a requirement example input by a user, a design step related to each design requirement, a single or sequence-combined API call instruction corresponding to each of the design steps, and a functional description of the API call instruction;

[0015] Based on the sample data, the initial large model is fine-tuned so that the initial large model updates the model parameters through back propagation to minimize the loss parameters, thereby obtaining a trained large model; wherein the initial large model is used to indicate a model that is pre-trained using a CAD design-related corpus.

[0016] In a possible implementation, obtaining sample data includes:

[0017] Obtain user input requirement examples, design steps related to each design requirement, and corresponding single or sequence combination API call instructions;

[0018] Obtaining a function description corresponding to each API in the API call instruction, wherein the function description includes a function description, parameter information, and usage scenario of the corresponding API;

[0019] Annotate and associate the requirements input by the user with the corresponding design steps, and annotate and associate each design step with the corresponding API call instruction;

[0020] The sample data is obtained based on the associated data and the functional description.

[0021] In a possible implementation, obtaining the request information of the user terminal includes:

[0022] Acquire original request information input by the user end, where the original request information is a text, a document, or a CAD model file to be processed and a requirement description of the user end based on the CAD model file to be processed;

[0023] If the original request information is text, then the text is the request information;

[0024] If the original request information is a document, extracting the target text in the document and using the target text as the request information;

[0025] If the original request information is a CAD model file to be processed and the user terminal has a requirement description based on the CAD model file to be processed, the API instruction for parsing the CAD model file to be processed in the preset API call instruction is called to obtain the CAD model information to be processed, and the CAD model information to be processed and the requirement description are used as the request information.

[0026] In a possible implementation, before obtaining the request information from the user terminal, the method further includes:

[0027] In response to a click operation of a start control by an operator on a development platform, a corresponding component list is obtained, wherein the development platform is used by the operator to arrange a workflow generated by a CAD model, and the component list includes a conditional branch and a plurality of functional components;

[0028] In response to the operator's operation of adding components based on the component list, the classification of the user's original request information is defined, and the text, document, or CAD model file to be processed and the user's requirement description based on the CAD model file to be processed are identified and processed to obtain the request information, and the design steps corresponding to the request information are obtained through the large model, as well as the API call instructions corresponding to each design step, to obtain the workflow of the target CAD model.

[0029] In a possible implementation manner, rendering the target CAD model based on the target CAD model data includes:

[0030] The target CAD model data is rendered into an HTML file by a WebGL-based 3D library;

[0031] According to the HTML file, when a page displaying the target CAD model in the user-end interface is loaded, a three-dimensional view of the target CAD model is generated and displayed.

[0032] In a second aspect, the present application provides a CAD model generation device based on an industrial large model, the device comprising:

[0033] An acquisition module, used to acquire request information from a user, where the request information is used to indicate the user's requirement information for generating or processing a CAD model;

[0034] In a possible implementation, the acquisition module is specifically used to:

[0035] Acquire original request information input by the user end, where the original request information is a text, a document, or a CAD model file to be processed and a requirement description of the user end based on the CAD model file to be processed;

[0036] If the original request information is text, then the text is the request information;

[0037] If the original request information is a document, extracting the target text in the document and using the target text as the request information;

[0038] If the original request information is a CAD model file to be processed and the user terminal has a requirement description based on the CAD model file to be processed, the API instruction for parsing the CAD model file to be processed in the preset API call instruction is called to obtain the CAD model information to be processed, and the CAD model information to be processed and the requirement description are used as the request information.

[0039] In a possible implementation, the acquisition module is specifically used to:

[0040] In response to a click operation of a start control by an operator on a development platform, a corresponding component list is obtained, wherein the development platform is used by the operator to arrange a workflow generated by a CAD model, and the component list includes a conditional branch and a plurality of functional components;

[0041] In response to the operator's operation of adding components based on the component list, the classification of the user's original request information is defined, and the text, document, or CAD model file to be processed and the user's requirement description based on the CAD model file to be processed are identified and processed to obtain the request information, and the design steps corresponding to the request information are obtained through the large model, as well as the API call instructions corresponding to each design step, to obtain the workflow of the target CAD model.

[0042] A processing module, configured to determine, according to the request information, the design steps corresponding to the request information and the API call instructions corresponding to each design step through a large model; wherein there is an association relationship between the request information and the design steps, and the large model is obtained by training based on the association relationship, preset API call instructions and corresponding function descriptions;

[0043] In a possible implementation manner, the processing module is specifically configured to:

[0044] According to the request information and the preset step generation prompt template, a corresponding step generation prompt is obtained, wherein the step generation prompt template is used to prompt the large model to generate design steps for user interaction, and provide an API call instruction corresponding to each design step;

[0045] The step generation prompt is input into the large model to obtain the design steps output by the large model, as well as the single or sequence combination API call instructions corresponding to each design step.

[0046] In a possible implementation manner, the processing module is specifically configured to:

[0047] Acquire sample data, wherein the sample data includes a requirement example input by a user, a design step related to each design requirement, a single or sequence-combined API call instruction corresponding to each of the design steps, and a functional description of the API call instruction;

[0048] Based on the sample data, the initial large model is fine-tuned so that the initial large model updates the model parameters through back propagation to minimize the loss parameters, thereby obtaining a trained large model; wherein the initial large model is used to indicate a model that is pre-trained using a CAD design-related corpus.

[0049] In a possible implementation manner, the processing module is specifically configured to:

[0050] Obtain user input requirement examples, design steps related to each design requirement, and corresponding single or sequence combination API call instructions;

[0051] Obtaining a function description corresponding to each API in the API call instruction, wherein the function description includes a function description, parameter information, and usage scenario of the corresponding API;

[0052] Annotate and associate the requirements input by the user with the corresponding design steps, and annotate and associate each design step with the corresponding API call instruction;

[0053] The sample data is obtained based on the associated data and the functional description.

[0054] A generation module, used to generate target CAD model data according to the design steps and corresponding API call instructions;

[0055] A display module is used to display the design steps on the display interface of the user terminal and render the target CAD model based on the target CAD model data.

[0056] In a possible implementation, the display module is specifically used for:

[0057] The target CAD model data is rendered into an HTML file by a WebGL-based 3D library;

[0058] According to the HTML file, when a page displaying the target CAD model in the user-end interface is loaded, a three-dimensional view of the target CAD model is generated and displayed.

[0059] In a third aspect, the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer-executable instructions, and when the computer-executable instructions are executed by a processor, they are used to implement the method as described in any one of the first aspects.

[0060] In a fourth aspect, the present application provides an electronic device, comprising: at least one processor and a memory; wherein:

[0061] The memory stores computer-executable instructions;

[0062] The at least one processor executes the computer-executable instructions stored in the memory, so that the at least one processor performs the method as described in any one of the first aspects.

[0063] In a fifth aspect, the present application also provides a computer program product, including a computer program, which, when executed by a processor, can implement the steps of the method described in any one of the first aspects.

[0064] The CAD model generation method, device, equipment and medium based on industrial large models provided by the present application, by determining the design steps corresponding to the request information based on the large model, and the API call instructions corresponding to each design step according to the request information of the user end, that is, the user's demand information for generating or processing the CAD model. The API call instructions can be single or in a sequence combination. Among them, the large model is obtained through fine-tuning training with a large corpus and related background. According to the design steps and the corresponding API call instructions, the target CAD model data is generated. Finally, the design steps and the rendered target CAD model are displayed on the display interface of the user end. This method realizes automatic modeling through the generalization ability of the large model, reduces the manual burden, automates complex design tasks, and improves the accuracy and efficiency of the workflow. It simplifies the user interface and natural language interaction, and improves the user experience and processing efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0066] Figure 1 A schematic diagram of a workflow provided according to an embodiment of the present application;

[0067] Figure 2 The process of the CAD model generation method based on the large model provided in the embodiment of the present application Figure 1 ;

[0068] Figure 3 is a schematic diagram of the development platform interface according to this embodiment;

[0069] Figure 4 The process of the CAD model generation method based on the large model provided in the embodiment of the present application Figure 2 ;

[0070] Figure 5 A CAD model generation device diagram based on an industrial large model provided in an embodiment of the present invention;

[0071] Figure 6 A hardware schematic diagram of an electronic device provided by an embodiment of the present invention.

[0072] The above drawings have shown clear embodiments of the present application, which will be described in more detail later. These drawings and text descriptions are not intended to limit the scope of the present application in any way, but to illustrate the concept of the present application to those skilled in the art by referring to specific embodiments. DETAILED DESCRIPTION

[0073] In order to make the purpose, technical solutions and advantages of this application clearer, the technical solutions in this application will be clearly and completely described below in conjunction with the drawings in this application. Obviously, the described embodiments are part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0074] The terms "first", "second", "third", "fourth", etc. (if any) in the description and claims of the present invention and the above drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the numbers used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in sequences other than those illustrated or described herein, for example.

[0075] In the embodiments of the present application, the words "exemplary" or "for example" are used to indicate examples, illustrations or descriptions. Any embodiment or design described as "exemplary" or "for example" in 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.

[0076] With the popularization of AIGC generative artificial intelligence technology, the model generation technology of text-generated images and image-generated images has become well-known, but this image generation technology has not been widely and effectively promoted and applied in the field of industrial CAD design. The shortcomings of existing CAD generation technology are mainly as follows:

[0077] Reliance on manual operation, long generation time, and high risk of errors: The creation of traditional CAD models usually requires manual operation by professional designers, which not only requires the operator to have professional skills and experience, but also limits the efficiency and innovation of the design. Due to the reliance on manual design step by step, it takes a long time from conceptual design to the generation of the final model. At the same time, manual operation increases the risk of errors in the design process, especially when performing complex or detailed designs.

[0078] The user interface is difficult to operate, the interaction is not friendly and concise enough, and the technical threshold is limited: the user interface of some traditional CAD software may not be intuitive enough, which makes it difficult for new users to get started, and the user experience needs to be improved. Hiring professional designers increases the cost burden of the enterprise.

[0079] Based on the above problems, this application hopes to realize the automatic generation of CAD models through generative artificial intelligence and intelligent workflows. Through automated design, manual operations and error risks are reduced, thereby improving design efficiency. AI is used to decompose tasks and automatically generate required CAD models, improve the accuracy and explainability of decisions, and enhance intelligence. Natural language prompts and intuitive interactions are introduced to reduce learning difficulty and optimize user experience.

[0080] Therefore, the present application provides a CAD model generation method based on an industrial large model, by determining the design steps corresponding to the request information based on the large model, and the API call instructions corresponding to each design step according to the request information of the user end, that is, the demand information of the user to generate or process the CAD model. The API call instruction can be a single one or a sequence combination. Among them, the industrial large model (hereinafter referred to as the large model) in the present application is obtained by fine-tuning training under the condition of having a large corpus and related background in the industrial field. According to the design steps and the corresponding API call instructions, the target CAD model data is generated. Finally, the design steps and the rendered target CAD model are displayed on the display interface of the user end. In addition, the process of realizing the automation can be combined with an intelligent development platform to arrange the workflow, and the workflow is controlled by an open visualization platform to realize the intelligence and automation of the processing flow. This method realizes automatic modeling through the generalization ability of the large model, reduces the manual burden, automates complex design tasks, and improves the accuracy and efficiency of the workflow. The user interface and natural language interaction are simplified, and the user experience and processing efficiency are improved.

[0081] Figure 1Schematic diagram of a workflow provided according to an embodiment of the present application. Figure 1 As shown, after the operator performs workflow arrangement processing through the development platform, the intelligent workflow includes: the problem classifier can perform classification processing based on the content entered by the user on the user-side display interface. Exemplarily, the user can enter text, documents, CAD models to be processed and related text requirements of the CAD models to be processed. When it is text, it is directly sent to the prompt word generator for workflow decomposition; when it is a document containing text, it is necessary to first extract the text content in the document and then send it to the prompt word generator; when it is a CAD model to be processed, it is necessary to first load the CAD model to be processed, extract features from the CAD model, and combine the processing requirements of the CAD model to be processed, and provide it to the prompt word generator for processing. When the CAD model to be processed is loaded, the relevant API instructions in the Open API instruction set can be called to implement it. The instructions in the Open API instruction set are preset so that they can be directly called when performing related processing. For text or documents, the demand is usually to create a CAD model. At this time, the prompt word generator is disassembled into specific design execution steps according to the demand, and the design execution steps can also be displayed on the user-side display interface. For the CAD model to be processed, it is usually required to perform relevant processing operations on the CAD model to be processed, such as checking whether there are CAD model components similar to the CAD model to be processed in the existing inventory, that is, similarity sorting; or determining whether two CAD models to be processed can be assembled together, that is, interference checking. At the same time, the disassembled design execution steps will match the single API call instructions or combined sequence of API instructions for the relevant calls to execute these API instructions and return the execution results. That is, the created model is displayed on the user-side display interface, or the processing results of the CAD model to be processed are displayed.

[0082] The technical solution of the present application and how the technical solution of the present application solves the above-mentioned technical problems are described in detail below with specific embodiments. The following specific embodiments may exist independently or in combination with each other, and the same or similar concepts or processes may not be described in detail in some embodiments. The embodiments of the present application will be described below in conjunction with the accompanying drawings.

[0083] This embodiment provides a CAD model generation method based on an industrial large model. Figure 2 The process of the CAD model generation method based on the large model provided in the embodiment of the present application Figure 1 The method comprises:

[0084] S201. Obtain request information from a user, where the request information is used to indicate the user's requirement information for generating or processing a CAD model.

[0085] In this step, the user's specific requirements for designing or processing CAD models are clarified by obtaining the user's request information (text, document, or CAD model file and requirement description). In order to adapt to the user's diverse input methods (such as text description, document, or existing CAD model file), it is necessary to be able to process different forms of original request information to accurately obtain the requirement information and guide the subsequent steps.

[0086] Exemplarily, the obtaining of the request information of the user terminal includes:

[0087] Acquire original request information input by the user end, where the original request information is a text, a document, or a CAD model file to be processed and a requirement description of the user end based on the CAD model file to be processed;

[0088] If the original request information is text, then the text is the request information;

[0089] If the original request information is a document, extracting the target text in the document and using the target text as the request information;

[0090] If the original request information is a CAD model file to be processed and the user terminal has a requirement description based on the CAD model file to be processed, the API instruction for parsing the CAD model file to be processed in the preset API call instruction is called to obtain the CAD model information to be processed, and the CAD model information to be processed and the requirement description are used as the request information.

[0091] As shown in the above example, different parsing methods can be selected according to the type of information input by the user. Request information in text form is usually the design requirements directly described by the user, which contains the user's specific requirements or modification instructions for the model, receives text input, and directly stores it as request information for subsequent operations. Request information in document form may contain descriptions or instructions related to CAD model design requirements. In order to identify the requirement information, useful text content needs to be extracted from the document. The design-related parts can be identified and extracted in the document through a preset text extraction algorithm or keyword extraction method, and used as request information for subsequent design step generation. For existing CAD model files, user requirements are usually to modify, adjust or analyze existing models. In order to accurately understand these requirements, it is necessary to first obtain the structural information of the CAD model. The CAD model file can be parsed by calling preset API instructions to extract the geometric structure, size, feature information, etc. of the model. The parsed model information is combined with the requirement description provided by the user to form a complete request information for subsequent operations.

[0092] For example, in some specific application scenarios, the user is a designer who needs to create a new CAD model. He enters a text requirement description of "create a mechanical part with a bracket and a hole" through the input interface. The system obtains the text as request information and directly passes the user's text description to the next step. In some embodiments, the user is an engineer who records the design requirements in a project document. The document describes in detail the design specifications and parameters of the CAD model to be generated. The user uploads this document through the system input interface. The system detects that the input file is a document, automatically reads the document content and extracts the target text related to the CAD design. The system uses the extracted text as request information and passes the text description to the next step. In other embodiments, the user has designed a basic CAD model and uploaded the CAD model file, and at the same time describes the required modifications to the model through text (such as "add a bracket to this model"). The user hopes to adjust and improve the original design. The system detects that the input contains a CAD model file and related requirement descriptions, calls a preset API to parse the CAD model file, obtains key information such as the model's geometry, dimensions, and features, and combines the parsed CAD model information with the requirement description to form a complete request information, which is then passed to the next step.

[0093] S202. According to the request information, determine the design steps corresponding to the request information and the API call instructions corresponding to each design step through a large model.

[0094] There is an association relationship between the request information and the design steps, and the large model is pre-trained based on the association relationship, preset API call instructions and corresponding function descriptions.

[0095] In this step, the large model parses the user's request information, automatically generates the required design steps, and determines the appropriate API call instructions for each step to achieve the automation of CAD model generation or processing. The user's request information contains the key content of the design requirements. The large model learns the relationship between user requirements and design steps during training, so that it can automatically generate design steps that meet the requirements. Design steps usually need to be implemented through specific API call instructions. The large model learns the matching relationship between design steps and API instructions through pre-training.

[0096] The request information obtained in step S201 is passed to the large model as its input. The model first parses the request information through natural language processing technology to extract key information such as design goals, operation requirements, and structural descriptions. Based on the association between the request information and the design steps, the large model generates multiple design steps from the parsing results, and further matches API call instructions for each design step. For example, for the design step of "generating a support structure", the model will generate API instructions for calling support drawing and size setting; for the step of "checking structural interference", the model will match API call instructions related to interference detection.

[0097] For complex design steps, a combination of multiple API instructions may be required. The large model will automatically generate a multi-step instruction combination based on the serialized call relationship in the training data to ensure the integrity of the steps. For example, if the user requirement involves "generating a bracket with multiple holes and checking the gap", the large model may generate a combined instruction sequence for this, such as "draw the bracket" - "create holes" - "set gap detection parameters" - "perform gap detection".

[0098] Exemplarily, the obtaining, according to the request information, the design steps corresponding to the request information through the big model, and the API call instructions corresponding to each design step, include:

[0099] According to the request information and the preset step generation prompt template, a corresponding step generation prompt is obtained, wherein the step generation prompt template is used to prompt the large model to generate design steps for user interaction, and provide an API call instruction corresponding to each design step;

[0100] The step generation prompt is input into the large model to obtain the design steps output by the large model, as well as the single or sequence combination API call instructions corresponding to each design step.

[0101] In the above example, by using the preset step generation prompt template, the large model can be effectively guided to generate design steps that meet the requirements and match the corresponding API call instructions. In some embodiments, the request information input by the user is analyzed, and after extracting the key information, the most suitable one can be selected from the preset prompt templates based on the analysis results. The prompt template is a specific prompt structure that is designed to guide the large model to generate design steps that meet user requirements. Fill the request information into the selected prompt template to generate a step generation prompt. For example, if the request information is "Create a mechanical part with a bracket and holes", the step generation prompt may contain the following:

[0102] The example generates a step prompt: "You are a CAD design expert. Please create a mechanical part with a bracket and holes for the user. Please generate the design process step by step and match the corresponding API call instructions for each step."

[0103] In some embodiments, the step generation prompt template not only includes the user's design requirements, but also presets some fixed prompt contents to guide the large model to output complete and structured design steps. For example, the template may include the following contents:

[0104] “You are a CAD design expert.”

[0105] “Please generate design steps that meet user needs step by step.”

[0106] “Each step must be accompanied by specific API call instructions, and multiple API instructions can be combined if necessary.”

[0107] Because the template contains predefined structures and guidance content, the large model can better understand the flow of design steps and generate corresponding API instructions for each step, avoiding ambiguous output. For example, the user enters a complex requirement: "Add a bracket based on the existing model, perform interference detection and generate a design report." At this time, the large model can be guided by the step generation prompt template to gradually generate design steps to ensure that the execution of each step meets user requirements and is consistent and logical.

[0108] In other embodiments, if the generated design steps are also displayed on the display interface of the user end for the user to learn or as a basis for the user to input and adjust again, then the output of the design steps by the large model can be defined in the step generation prompt template to be more instructive and interactive. For example, while prompting the large model to generate the design steps, a guiding sentence prompting the user to give feedback or a sentence giving the user relevant suggestions can be added.

[0109] For example, the above content template may include the following contents:

[0110] You are a CAD model design expert. Follow the following requirements to gradually generate design steps that meet user needs. Each step must be accompanied by specific API call instructions. Multiple API instructions can be combined and called when necessary.

[0111] Step 1: Prompt the user to start the local CAD software or enable the cloud CAD service

[0112] Step 2: Draw basic graphics according to the text requirements entered by the user. After drawing, display the graphics and describe them.

[0113] Step 3: Graphics Add dimensions, parameters, and geometric constraints, which help ensure the accuracy and modifiability of the design.

[0114] Step 4: Begin refining the model, which may include adding fillets, chamfers, extrusions, or revolves.

[0115] Step 5: Perform a detailed design review to check that all dimensions, constraints, and shapes meet the design specifications.

[0116] Step 6: Generate a detailed design report. Please review the report and confirm that all information is accurate.

[0117] Finally, end with: "Congratulations on completing the design of the CAD model! If you need further modifications or have other design needs, I am always here to help you."

[0118] Based on the prompt template of the above example, the steps involved in generating the interaction with the user are as follows:

[0119] As an expert in CAD model design, I will guide you through the entire design process. Let's get started!

[0120] Step 1: Launch the CAD software

[0121] First, please launch your local CAD software, or if you are using a cloud service, open your CAD service web page. Make sure all the tools are ready so that we can start designing.

[0122] Please let me know when you are ready and we will proceed to the next step.

[0123] Step 2: Draw the basic shape

[0124] Now, draw the basic graphics in the CAD software according to the text requirements you provided.

[0125] After drawing, show your drawing (show CAD model)

[0126] Step 3: Add Dimensional and Geometric Constraints

[0127] Next, add the necessary dimensions and geometric constraints to your shape. This will help ensure the accuracy of your design and allow you to easily modify the design in the future.

[0128] Step 4: Refine the model

[0129] Start refining your model, which may include adding fillets, chamfers, extrusions, or revolves to make the model more complete. (After refining, display the CAD model)

[0130] Step 5: Design Review

[0131] Now, conduct a detailed design review to check that all dimensions, constraints, and shapes meet the design specifications.

[0132] Please provide your review results to ensure the design is correct.

[0133] Step 6: Generate Design Report

[0134] Based on your design process and review results, I will help you generate a detailed design report.

[0135] Please review the report and confirm that all information is accurate.

[0136] Ending:

[0137] Congratulations on completing your CAD model! If you require further modifications or have other design needs, I am always here to help. Your design is now ready to move on to the next stage, or be adjusted as needed.

[0138] S203: Generate target CAD model data according to the design steps and corresponding API call instructions.

[0139] In this step, according to the design steps and API call instructions generated in the previous step, the API is called step by step to complete various design requirements and generate target CAD model data that meets user needs. That is, by calling the API sequentially, the system superimposes the data generated in each step on the model data and gradually forms a complete CAD model. The generated target CAD model data may include but is not limited to complete geometric information, structural characteristics, and analysis results (such as interference analysis).

[0140] In the specific process, API instructions are called step by step in the order of design steps to build the geometric features of the CAD model. For example, if the user's requirement is to "add a bracket and perform interference detection", the system will first execute the instructions for drawing basic geometric shapes, adding brackets, creating hole structures, and finally performing interference detection. After each API call instruction is executed, the generated data of the current step can be saved to the CAD model data structure, so that the model remains complete and meets the user's design requirements at each step. For example, after the bracket and hole structure are generated, they are added to the model data so that the complete model information can be accessed when performing interference detection in the next step.

[0141] S204: Display the design steps on the display interface of the user terminal, and render the target CAD model based on the target CAD model data.

[0142] In this step, the design steps and generated model data can be intuitively displayed to the user, helping the user to view the effect of each design step in real time and understand the final form of the model. Using the target CAD model data for 3D rendering enables the user to view the details of the model and verify whether the design effect meets the requirements.

[0143] Exemplarily, rendering the target CAD model based on the target CAD model data includes:

[0144] The target CAD model data is rendered into an HTML file by a WebGL-based 3D library;

[0145] According to the HTML file, when a page displaying the target CAD model in the user-end interface is loaded, a three-dimensional view of the target CAD model is generated and displayed.

[0146] In the above example, by embedding the target CAD model data into the HTML file, it has the ability to render a three-dimensional model in the browser, which is convenient for users to view and interact. Specifically, first load the target CAD model data generated by the API instruction. The data may include information such as geometry, size, material and texture. By calling the WebGL 3D library (such as Three.js), the CAD model data is converted into a visual three-dimensional graphic. The WebGL library (Web Graphics Library) embeds the rendered three-dimensional model into the HTML file, allowing the browser to directly load and display the model view. In the loaded HTML file, the user can see a three-dimensional model (such as a cube or a support structure), and use the mouse to rotate and zoom to view different perspectives.

[0147] In some embodiments, the large model can be used to generate an HTML file for rendering. In this scenario, the large model not only outputs design steps and API instructions, but can also directly generate an HTML file containing WebGL rendering code according to design requirements. The call to the large model can be spontaneous by the system. For example, after obtaining the CAD model data, the system can propose a display or rendering requirement, such as "generate a three-dimensional view with a bracket and display it." After receiving the requirement, the large model generates an HTML file template based on the model data and embeds the corresponding WebGL rendering code.

[0148] For example, if the user's request is to "generate and display a 3D model of a bracket with holes", the large model will generate an HTML file containing the model structure, with Three.js code embedded in it, to convert the user's needs into a 3D view display.

[0149] Exemplarily, before obtaining the request information from the user terminal, the method further includes:

[0150] In response to a click operation of a start control by an operator on a development platform, a corresponding component list is obtained, wherein the development platform is used by the operator to arrange a workflow generated by a CAD model, and the component list includes a conditional branch and a plurality of functional components;

[0151] In response to the operator's operation of adding components based on the component list, the classification of the user's original request information is defined, and the text, document, or CAD model file to be processed and the user's requirement description based on the CAD model file to be processed are identified and processed to obtain the request information, and the design steps corresponding to the request information are obtained through the large model, as well as the API call instructions corresponding to each design step, to obtain the workflow of the target CAD model.

[0152] As in the above example, the steps in this embodiment can be implemented by operators choreographing on the development platform in advance. The development platform can be an integrated environment of an existing computer-aided design (CAD) software, which can provide an intuitive graphical user interface (GUI) so that operators can easily choreograph CAD workflows without writing complex codes. Each component in the component library is designed to be configurable and extensible to accommodate different design requirements and API calls. Integration with various CAD software and external services enables the platform to call the corresponding API to generate and modify models.

[0153] Figure 3 FIG. 1 is a schematic diagram of the development platform interface according to this embodiment. Figure 3 As shown, when the operator clicks the "Start" control, a component list will appear, which contains multifunctional "nodes" and "tools" for workflow arrangement. When a component in the selected component list is not clicked to add, the corresponding introduction content will pop up to explain the function of the selected component. As shown in the schematic diagram, when the "problem classifier" in the component list is selected, the content box displayed on the right explains the role of the problem classifier (the explanation of the problem classifier is only an example). When you continue to click Add, the editing window of the problem classifier will pop up, which can be used to increase the type of problem classification and define the classification criteria for each type. In this embodiment, text, documents and models can be distinguished by arranging the problem classifier. Exemplarily, when arranging other parts, for example, according to the user needs after classification processing, the design steps are further split, and the API call instructions are matched. The functions of these parts can be realized by adding components. The above process can be implemented in some existing open source development platforms. When there are obvious differences in specific application scenarios, the plug-in mechanism of the open platform can also allow operators to develop specific functional components so that these functional components can be integrated into the workflow.

[0154] It should be noted that the above development process is achieved through the prompting technology of Chain of Thought (CoT), which breaks down complex problems into simple steps to form a clear execution path, and then uses the few-shot learning method to allow the agent to quickly learn task execution through a small number of examples. This method imitates the human thinking process and finds solutions to problems through step-by-step reasoning, making the large language model have stronger reasoning ability and flexibility in solving problems.

[0155] The CAD model generation method based on the industrial large model provided in this embodiment determines the design steps corresponding to the request information and the API call instructions corresponding to each design step based on the large model according to the request information of the user end, that is, the demand information of the user to generate or process the CAD model. The API call instruction can be a single one or a sequence combination. Among them, the large model is obtained through fine-tuning training with a large corpus and related background. According to the design steps and the corresponding API call instructions, the target CAD model data is generated. Finally, the design steps and the rendered target CAD model are displayed on the display interface of the user end. This method realizes automatic modeling through the generalization ability of the large model, reduces the manual burden, automates complex design tasks, and improves the accuracy and efficiency of the workflow. It simplifies the user interface and natural language interaction, and improves the user experience and processing efficiency.

[0156] This embodiment provides a CAD model generation method based on an industrial large model. Figure 4 The process of the CAD model generation method based on the large model provided in the embodiment of the present application Figure 2 .like Figure 4 As shown, in this embodiment Figure 2 Based on the embodiment, the pre-training process of the large model is described in detail. The method includes:

[0157] S401. Obtain sample data, wherein the sample data includes a requirement example input by a user, a design step related to each design requirement, a single or sequence combination of API call instructions corresponding to each of the design steps, and a functional description of the API call instructions.

[0158] In this step, in order to enable the large model to generate design steps and match appropriate API call instructions in the CAD model design plan, the pre-training process needs to be combined with specific domain knowledge and task requirements. Among them, the functional description details the purpose and function of each API call to ensure that the model can understand how to use these APIs. It provides background information for the model to help it make accurate choices when generating steps and instructions. When generating design steps, the model can use the functional description to determine which steps require which APIs. The functional description provides context so that the model can take into account the specific APIs that need to be used in actual operations when outputting design steps.

[0159] Exemplarily, the obtaining of sample data includes:

[0160] Obtain user input requirement examples, design steps related to each design requirement, and corresponding single or sequence combination API call instructions;

[0161] Obtaining a function description corresponding to each API in the API call instruction, wherein the function description includes a function description, parameter information, and usage scenario of the corresponding API;

[0162] Annotate and associate the requirements input by the user with the corresponding design steps, and annotate and associate each design step with the corresponding API call instruction;

[0163] The sample data is obtained based on the associated data and the functional description.

[0164] It should be noted that the relationship between annotated user input and generated design steps. This is the basis of model learning, helping it understand how different requests are converted into specific steps. Each design step is associated with its corresponding API call instruction annotation to ensure that the model understands which API calls can be used to implement specific steps. Exemplarily, the user input requirement examples can be obtained by collecting user historical inputs, and the user's historical data best represents the general needs of users in actual scenarios. The design steps and corresponding API call instructions related to each design requirement can be implemented by preset, and when the data volume demand is large, the design steps and API calls can be captured and recorded by using automated tools. For example, software monitoring tools can be used to automatically record the designer's operation steps in CAD software; or design-related tasks can be distributed to a large number of designers to quickly collect a large amount of sample data, etc., without limitation. The API call instructions matched by the design steps can be implemented from a preset API instruction library, which contains detailed information on the available APIs (including functional descriptions) and sample codes or instructions on how to call these APIs. As important information in the sample data, the functional description can ensure that the model can effectively learn and generate design steps and API call instructions through the role of auxiliary information.

[0165] For example, in the API instruction library, some more complex and commonly used design steps can be specially preset corresponding to API call instructions. These design steps often require calling a series of APIs to be implemented. For example, if you want to determine whether two specified CAD models can be assembled together, one of the design steps will involve determining the interference of the two model components. The usual process for checking interference is to read the component information of the two compared CAD models, and then perform interference checking and analyze the inspection results. Performing interference checking requires calling a series of APIs to implement, which can be preset according to the following process:

[0166] Use the UF_MODL_check_interference function to check whether two bodies interfere.

[0167] If a more detailed gap analysis is required, first create a dataset using UF_CLEAR_create_dataset.

[0168] Use UF_CLEAR_set_obj_list to set the list of objects involved in the analysis.

[0169] Call UF_CLEAR_set_analysis_mode to set the analysis mode.

[0170] To perform a clearance analysis, use UF_CLEAR_do_clearance_analysis.

[0171] Use UF_CLEAR_count_interferences to count the number of interferences.

[0172] If there is interference, use UF_CLEAR_ask_next_interf and UF_CLEAR_ask_interf_data to obtain specific interference data.

[0173] After similar presets are completed in the API instruction library, when the API call instructions corresponding to the design steps are matched, the above API call combination can be matched and determined through the large model.

[0174] S402: Based on the sample data, fine-tune the initial large model so that the initial large model updates the model parameters through back propagation to minimize the loss parameters and obtain a trained large model. The initial large model is used to indicate a model that is pre-trained using a CAD design-related corpus.

[0175] In this step, fine-tuning is to conduct targeted training on a model that has been pre-trained with a large-scale data set to adapt to a specific task or field. Exemplarily, a model that supports multi-task learning and large-scale natural language processing (such as the Transformer architecture, the GPT series model, etc.) can be used as the initial large model. In the specific fine-tuning process, the data of a specific task can be input into the model, and the model can be fine-tuned to optimize its output. For CAD design scenarios, it includes generating design steps and API call instructions based on the input. The model parameters are updated by back propagation to minimize the loss function, and iterative optimization is performed using mini-batch gradient descent. In some cases, in order to maintain the generalization ability of the model, it is possible to choose to freeze some layers of the model and only fine-tune the last few layers. When the fine-tuning of the initial large model is completed, the model is fine-tuned to be able to recognize complex user needs, decompose them into multi-step tasks, and generate a suitable combination of API instructions.

[0176] The large-model-based CAD model generation method provided in this embodiment realizes the automation and intelligence of CAD model generation by utilizing large-scale pre-trained models and fine-tuning, effectively improving design efficiency and accuracy, lowering the design threshold, optimizing the design process, and enhancing user experience.

[0177] This embodiment also provides a CAD model generation device based on industrial large models. Figure 5 A CAD model generation device diagram based on an industrial large model is provided in an embodiment of the present invention, such as Figure 5As shown, the CAD model generating device 50 includes:

[0178] The acquisition module 501 is used to acquire request information from the user, where the request information is used to indicate the user's requirement information for generating or processing a CAD model;

[0179] In a possible implementation, the acquisition module 501 is specifically used to:

[0180] Acquire original request information input by the user end, where the original request information is a text, a document, or a CAD model file to be processed and a requirement description of the user end based on the CAD model file to be processed;

[0181] If the original request information is text, then the text is the request information;

[0182] If the original request information is a document, extracting the target text in the document and using the target text as the request information;

[0183] If the original request information is a CAD model file to be processed and the user terminal has a requirement description based on the CAD model file to be processed, the API instruction for parsing the CAD model file to be processed in the preset API call instruction is called to obtain the CAD model information to be processed, and the CAD model information to be processed and the requirement description are used as the request information.

[0184] In a possible implementation, the acquisition module 501 is specifically used to:

[0185] In response to a click operation of a start control by an operator on a development platform, a corresponding component list is obtained, wherein the development platform is used by the operator to arrange a workflow generated by a CAD model, and the component list includes a conditional branch and a plurality of functional components;

[0186] In response to the operator's operation of adding components based on the component list, the classification of the user's original request information is defined, and the text, document, or CAD model file to be processed and the user's requirement description based on the CAD model file to be processed are identified and processed to obtain the request information, and the design steps corresponding to the request information are obtained through the large model, as well as the API call instructions corresponding to each design step, to obtain the workflow of the target CAD model.

[0187] Processing module 502, used to determine, according to the request information, the design steps corresponding to the request information and the API call instructions corresponding to each design step through the big model; wherein there is an association relationship between the request information and the design steps, and the big model is obtained by training based on the association relationship, the preset API call instructions and the corresponding function description;

[0188] In a possible implementation, the processing module 502 is specifically configured to:

[0189] According to the request information and the preset step generation prompt template, a corresponding step generation prompt is obtained, wherein the step generation prompt template is used to prompt the large model to generate design steps for user interaction, and provide an API call instruction corresponding to each design step;

[0190] The step generation prompt is input into the large model to obtain the design steps output by the large model, as well as the single or sequence combination API call instructions corresponding to each design step.

[0191] In a possible implementation, the processing module 502 is specifically configured to:

[0192] Acquire sample data, wherein the sample data includes a requirement example input by a user, a design step related to each design requirement, a single or sequence-combined API call instruction corresponding to each of the design steps, and a functional description of the API call instruction;

[0193] Based on the sample data, the initial large model is fine-tuned so that the initial large model updates the model parameters through back propagation to minimize the loss parameters, thereby obtaining a trained large model; wherein the initial large model is used to indicate a model that is pre-trained using a CAD design-related corpus.

[0194] In a possible implementation, the processing module 502 is specifically configured to:

[0195] Obtain user input requirement examples, design steps related to each design requirement, and corresponding single or sequence combination API call instructions;

[0196] Obtaining a function description corresponding to each API in the API call instruction, wherein the function description includes a function description, parameter information, and usage scenario of the corresponding API;

[0197] Annotate and associate the requirements input by the user with the corresponding design steps, and annotate and associate each design step with the corresponding API call instruction;

[0198] The sample data is obtained based on the associated data and the functional description.

[0199] A generating module 503, used to generate target CAD model data according to the design steps and corresponding API call instructions;

[0200] The display module 504 is used to display the design steps on the display interface of the user terminal and render the target CAD model based on the target CAD model data.

[0201] In a possible implementation, the display module 504 is specifically used for:

[0202] The target CAD model data is rendered into an HTML file by a WebGL-based 3D library;

[0203] According to the HTML file, when a page displaying the target CAD model in the user-end interface is loaded, a three-dimensional view of the target CAD model is generated and displayed.

[0204] The present embodiment provides a CAD model generation device based on industrial large models, which can execute the CAD model generation method based on large models provided in the above method embodiment. Its implementation principle and technical effects are similar, and will not be described in detail in this embodiment.

[0205] Figure 6 Schematic diagram of the hardware of the electronic device provided by the embodiment of the present invention. Figure 6 As shown, the electronic device 60 provided in this embodiment includes: at least one processor 601 and a memory 602. The device 60 also includes a communication component 603. The processor 601, the memory 602 and the communication component 603 are connected via a bus 604.

[0206] In a specific implementation process, at least one processor 601 executes the computer-executable instructions stored in the memory 602, so that at least one processor 601 executes the above method.

[0207] The specific implementation process of the processor 601 can be found in the above method embodiment, and its implementation principle and technical effect are similar, so this embodiment will not be repeated here.

[0208] In the above Figure 6In the illustrated embodiment, it should be understood that the processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), etc. A general-purpose processor may be a microprocessor or any conventional processor, etc. The steps of the method disclosed in the invention may be directly implemented as being executed by a hardware processor, or may be executed by a combination of hardware and software modules in the processor.

[0209] The memory may include a high-speed memory (Random Access Memory, RAM), and may also include a non-volatile memory (NVM), such as at least one disk storage.

[0210] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, the bus in the drawings of this application is not limited to only one bus or one type of bus.

[0211] The present application also provides a computer-readable storage medium, in which computer-executable instructions are stored. When a processor executes the computer-executable instructions, the method described above is implemented.

[0212] The above-mentioned readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk. The readable storage medium can be any available medium that can be accessed by a general or special-purpose computer.

[0213] An exemplary readable storage medium is coupled to a processor so that the processor can read information from the readable storage medium and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can be located in an application specific integrated circuit (Application Specific Integrated Circuits, referred to as: ASIC). Of course, the processor and the readable storage medium can also exist in the device as discrete components.

[0214] The division of the units is only a logical function division, and there may be other divisions in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection through some interface, device or unit, which may be electrical, mechanical or other forms.

[0215] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0216] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0217] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc., which can store program codes.

[0218] Those skilled in the art can understand that all or part of the steps of implementing the above-mentioned method embodiments can be completed by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, the steps of the above-mentioned method embodiments are executed; and the aforementioned storage medium includes: ROM, RAM, disk or optical disk and other media that can store program codes.

[0219] So far, the technical solution of the present application has been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present application is obviously not limited to these specific embodiments, and the above embodiments are only used to illustrate the technical solution of the present application rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein by equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A CAD model generation method based on industrial large models, characterized in that: The method comprises: Obtaining request information from a user, where the request information is used to indicate the user's requirement information for generating or processing a CAD model; According to the request information, the design steps corresponding to the request information and the API call instructions corresponding to each design step are determined by the big model; wherein there is an association relationship between the request information and the design steps, and the big model is pre-trained based on the association relationship, the preset API call instructions and the corresponding function description; Generate target CAD model data according to the design steps and corresponding API call instructions; The design steps are displayed on a display interface of the user terminal, and the target CAD model is rendered based on the target CAD model data.

2. The method according to claim 1, characterized in that The step of obtaining the design steps corresponding to the request information and the API call instructions corresponding to each design step through the large model according to the request information includes: According to the request information and the preset step generation prompt template, a corresponding step generation prompt is obtained, wherein the step generation prompt template is used to prompt the large model to generate design steps for user interaction, and provide an API call instruction corresponding to each design step; The step generation prompt is input into the large model to obtain the design steps output by the large model, as well as the single or sequence combination API call instructions corresponding to each design step.

3. The method according to claim 1, characterized in that The pre-training of the large model includes: Acquire sample data, wherein the sample data includes a requirement example input by a user, a design step related to each design requirement, a single or sequence-combined API call instruction corresponding to each of the design steps, and a functional description of the API call instruction; Based on the sample data, the initial large model is fine-tuned so that the initial large model updates the model parameters through back propagation to minimize the loss parameters, thereby obtaining a trained large model; wherein the initial large model is used to indicate a model that is pre-trained using a CAD design-related corpus.

4. The method according to claim 3, characterized in that The obtaining of sample data comprises: Obtain user input requirement examples, design steps related to each design requirement, and corresponding single or sequence combination API call instructions; Obtaining a function description corresponding to each API in the API call instruction, wherein the function description includes a function description, parameter information, and usage scenario of the corresponding API; Annotate and associate the requirements input by the user with the corresponding design steps, and annotate and associate each design step with the corresponding API call instruction; The sample data is obtained based on the associated data and the functional description.

5. The method according to claim 1, characterized in that The obtaining of the request information of the user terminal includes: Acquire original request information input by the user end, where the original request information is a text, a document, or a CAD model file to be processed and a requirement description of the user end based on the CAD model file to be processed; If the original request information is text, then the text is the request information; If the original request information is a document, extracting the target text in the document and using the target text as the request information; If the original request information is a CAD model file to be processed and the user terminal has a requirement description based on the CAD model file to be processed, the API instruction for parsing the CAD model file to be processed in the preset API call instruction is called to obtain the CAD model information to be processed, and the CAD model information to be processed and the requirement description are used as the request information.

6. The method according to claim 5, characterized in that Before obtaining the request information of the user terminal, the method further includes: In response to a click operation of a start control by an operator on a development platform, a corresponding component list is obtained, wherein the development platform is used by the operator to arrange a workflow generated by a CAD model, and the component list includes a conditional branch and a plurality of functional components; In response to the operator's operation of adding components based on the component list, the classification of the user's original request information is defined, and the text, document, or CAD model file to be processed and the user's requirement description based on the CAD model file to be processed are identified and processed to obtain the request information, and the design steps corresponding to the request information are obtained through the large model, as well as the API call instructions corresponding to each design step, to obtain the workflow of the target CAD model.

7. The method according to claim 1, characterized in that The step of rendering the target CAD model based on the target CAD model data comprises: The target CAD model data is rendered into an HTML file by a WebGL-based 3D library; According to the HTML file, when a page displaying the target CAD model in the user-end interface is loaded, a three-dimensional view of the target CAD model is generated and displayed.

8. A CAD model generation device based on industrial large models, characterized in that: The device comprises: An acquisition module is used to acquire request information from a user, where the request information is used to indicate the user's requirement information for generating or processing a CAD model; A processing module, configured to determine, according to the request information, the design steps corresponding to the request information and the API call instructions corresponding to each design step through a large model; wherein there is an association relationship between the request information and the design steps, and the large model is obtained by training based on the association relationship, preset API call instructions and corresponding function descriptions; A generation module, used to generate target CAD model data according to the design steps and corresponding API call instructions; A display module is used to display the design steps on the display interface of the user terminal and render the target CAD model based on the target CAD model data.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, and when the computer-executable instructions are executed by a processor, they are used to implement the method according to any one of claims 1 to 7.

10. An electronic device, characterized in that: include: at least one processor and memory; wherein, The memory stores computer-executable instructions; The at least one processor executes the computer-executable instructions stored in the memory, so that the at least one processor performs the method according to any one of claims 1 to 7.

Images