Light-weight assembly type mobile equipment industrial design generation method and system based on artificial intelligence

Abstract

The invention relates to the technical field of industrial design, and discloses a lightweight assembly type mobile equipment industrial design generation method based on artificial intelligence, and the method comprises the following steps: S1, constructing and training a dual-stage AI model which comprises a concept generation module and an industrial conversion module; s2, receiving multi-modal design demand information input by a user; s3, generating at least one visual concept design scheme through the concept generation module; and S4, in response to a selection instruction of a user for the visual concept design scheme, the selected concept scheme is converted into an industrial-grade design scheme through the industrial conversion module, and the industrial-grade design scheme comprises at least one of a bill of material, a technical drawing and a manufacturing instruction. Through the two-stage AI model, full-automatic generation from fuzzy text originality to an accurate industrial-grade manufacturing scheme is realized for the first time, the design period is shortened from several weeks to several hours, and the design cost is reduced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of industrial design, in particular to a lightweight assembly type mobile equipment industrial design generation method and system based on artificial intelligence. BACKGROUND

[0002] Lightweight assembly type mobile equipment is increasingly widely used in retail, catering, tourism and other fields due to its flexibility, mobility and versatility. However, the current design process of this industry mainly relies on manual work, which has the following outstanding problems:

[0003] 1. Long design cycle and high cost: The traditional design process involves concept communication, sketching, 3D modeling, structural design, material selection and other links, which requires the collaboration of designers, structural engineers and other parties, resulting in high communication costs and a cycle of several weeks or even months.

[0004] 2. Low innovation efficiency: Designers' creativity is limited by their personal experience and knowledge range, making it difficult to produce a large number of innovative solutions with aesthetics, functionality and cost efficiency in a short period of time.

[0005] 3. Design and manufacturing are disconnected: Conceptual design solutions often lack in-depth consideration of material costs, processing technology and structural feasibility, making it difficult to implement the solution in subsequent production and manufacturing stages or causing cost overruns.

[0006] 4. High threshold for personalized customization: For small and medium-sized enterprises or individual users, the cost of designing personalized equipment is high, limiting the further development of the market. SUMMARY

[0007] The present application aims to provide a lightweight assembly type mobile equipment industrial design generation method and system based on artificial intelligence to solve the problems raised in the background.

[0008] To achieve the above-mentioned purpose, the present application provides the following technical solution: a lightweight assembly type mobile equipment industrial design generation method based on artificial intelligence, comprising the following steps:

[0009] S1, a two-stage AI model is constructed and trained, the two-stage AI model comprising a concept generation module and an industrial transformation module;

[0010] S2, receiving user inputted multi-modal design requirement information;

[0011] S3, generating at least one visual concept design solution through the concept generation module;

[0012] S4, in response to a user's selection instruction of the visual concept design scheme, converting the selected concept scheme into an industrial-level design scheme through the industrial conversion module, the industrial-level design scheme including at least one of a bill of materials, technical drawings and manufacturing instructions.

[0013] As a preferred technical solution of the present application, the concept generation module adopts a multi-modal neural network model, receives at least one of text, image, style option and function parameter as input, and outputs corresponding three-dimensional rendering or three-dimensional model.

[0014] As a preferred technical solution of the present application, the industrial conversion module adopts a graph neural network or a knowledge graph model, integrates structural mechanics rules, material attribute parameters and manufacturing cost constraints, and automatically analyzes the structure and layout of the concept scheme and generates industrial-level data.

[0015] As a preferred technical solution of the present application, the industrial conversion module performs at least one of the following operations:

[0016] Identify the main structure, load-bearing part and functional partition in the concept model;

[0017] Generate a bill of materials according to a material knowledge base;

[0018] Automatically generate two-dimensional engineering drawings and three-dimensional assembly drawings;

[0019] Generate process instructions or numerical control machining codes for guiding production.

[0020] As a preferred technical solution of the present application, the step S1 further includes the following steps before the step S1:

[0021] S01, a professional database for lightweight assembly type mobile equipment is constructed, the professional database integrating image data, three-dimensional models, technical drawings, bill of materials, material attribute parameters, cost data and functional layout information of multiple equipment categories;

[0022] S02, the data in the database is cleaned, classified and structuredly labeled, and the labeling dimensions include equipment category, design style, structural component, functional area, material type and cost interval;

[0023] S03, according to the labeled data, the concept generation module and the industrial conversion module are respectively supervised trained and knowledge fused.

[0024] As a preferred technical solution of the present application, the concept generation module adopts a multi-modal neural network integrating a Transformer structure and a generative adversarial network or a diffusion model, extracts multi-modal features of user input through a text encoder and an image encoder, and performs semantic alignment and feature enhancement in a fusion layer to generate high-quality visual concept schemes meeting aesthetic and functional requirements.

[0025] As a preferred technical solution of the present application, the industrial transformation module performs manufacturability transformation on the selected concept scheme through an integrated engineering knowledge base, specifically including:

[0026] S11, structurally decompose the concept scheme according to a predefined equipment structure unit library and connection rules, and identify a bearing structure, an enclosure structure and a functional module;

[0027] S12, automatically match and optimize the selection of materials and specifications according to the decomposed structure units, in combination with a material performance database and a cost database;

[0028] S13, automatically generate a structured bill of materials containing component details, specifications, quantities and assembly relationships according to the matching results and assembly logic;

[0029] S14, parametrically generate a technical drawing set including at least general assembly drawings, subassembly drawings and part drawings in accordance with engineering drawing standards according to the structured bill of materials and the spatial relationship of the equipment structure units.

[0030] The lightweight assembly type mobile equipment industrial design generation system based on artificial intelligence comprises:

[0031] A two-stage AI model including a concept generation module and an industrial transformation module;

[0032] A user interaction interface for receiving user input multi-modal design requirements;

[0033] A processing unit for calling the concept generation module to generate visual concept schemes and calling the industrial transformation module to transform the user-selected concept schemes into industrial-level design schemes;

[0034] An output unit for outputting an industrial design package containing at least one of a bill of materials, technical drawings and manufacturing instructions.

[0035] As a preferred technical solution of the present application, the system further comprises:

[0036] An integrated professional database storing image data, three-dimensional models, technical drawings, bill of materials, material attribute parameters, cost data and functional layout information of multiple equipment categories, and connected with the two-stage AI model;

[0037] A data preprocessing module for cleaning, classifying and multi-dimensionally structuring the data in the professional database.

[0038] A model training and optimization module for supervised training and continuous optimization of the concept generation module and the industrial transformation module based on the labeled data.

[0039] A user feedback collection module for collecting user rating data, modification behavior and adoption results of the generated scheme.

[0040] A cloud deployment and scheduling module for encapsulating the trained two-stage AI model as an API service and deploying it on a SaaS platform.

[0041] A knowledge base updating engine for dynamically adjusting the rules and constraints in the industrial transformation module according to new materials, new processes or new standards.

[0042] As a preferred technical solution of the present application, the system further comprises:

[0043] A scheme manufacturability verification module for virtual assembly simulation, structural strength analysis, material utilization rate evaluation and manufacturing cost review of the industrial design scheme.

[0044] A supply chain coordination interface for data interfacing of the generated bill of materials with external supplier systems.

[0045] A design version management unit for version recording and backtracking management of the concept scheme and the industrial scheme generated by the user at different stages.

[0046] A cross-platform output adapter for converting the industrial design package into various mainstream CAD formats and manufacturing execution system compatible data structures.

[0047] Compared with the prior art, the present application has the following advantages:

[0048] 1. Through the two-stage AI model, the full automatic generation from the fuzzy text idea to the accurate industrial manufacturing scheme is realized for the first time, the design cycle is shortened from several weeks to several hours, and the design cost is reduced.

[0049] 2. The industrial transformation module deeply integrates structural mechanics, material properties and manufacturing cost constraints to ensure that the generated scheme has high manufacturability, effectively avoids the disconnection between design and production, and realizes what you see is what you get.

[0050] 3. Through the simple multi-modal interaction interface, users without professional design background can also quickly generate mobile equipment schemes that meet professional standards, greatly releasing the potential of market individual customization.

[0051] 4、AI model can comprehensively optimize aesthetics, function, cost and other multi-dimensional information, generate balanced solutions that traditional design methods cannot achieve, and improve overall design and production efficiency through supply chain collaboration interface and version management.

[0052] 5、System integration user feedback mechanism and knowledge base update engine, can continuously optimize the model according to new materials, new processes or user behavior data, and maintain technical advancement. BRIEF DESCRIPTION OF DRAWINGS

[0053] Fig. 1 The overall flowchart of the lightweight assembly type mobile equipment industrial design generation method based on artificial intelligence of the present application is shown in the figure.

[0054] Fig. 2 The structure and data processing flowchart of the industrial conversion module in the present application is shown in the figure. DETAILED DESCRIPTION

[0055] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0056] Embodiment 1

[0057] Please refer to Figs. 1-2 The present application provides a lightweight assembly type mobile equipment industrial design generation method based on artificial intelligence, which includes the following steps:

[0058] S1, a two-stage AI model is constructed and trained, and the two-stage AI model includes a concept generation module and an industrial conversion module;

[0059] S2, receiving user input multi-modal design requirement information;

[0060] S3, generating at least one visual concept design scheme through the concept generation module;

[0061] S4, in response to the user's selection instruction of the visual concept design scheme, converting the selected concept scheme into an industrial grade design scheme through the industrial conversion module, and the industrial grade design scheme includes at least one of a bill of materials, technical drawings and manufacturing instructions.

[0062] Further, the concept generation module adopts a multi-modal neural network model, receives at least one of text, image, style option and function parameter as input, and outputs corresponding three-dimensional effect picture or three-dimensional model.

[0063] Further, the industrial transformation module adopts a graph neural network or a knowledge graph model, integrates structural mechanics rules, material attribute parameters, and manufacturing cost constraints, and automatically analyzes the structure and layout of the conceptual scheme and generates industrial-level data.

[0064] Further, the industrial transformation module performs at least one of the following operations:

[0065] Identify the main structure, load-bearing part, and functional partition in the conceptual model;

[0066] Generate a bill of materials according to the material knowledge base;

[0067] Automatically generate two-dimensional engineering drawings and three-dimensional assembly drawings;

[0068] Generate process instructions or numerical control machining codes for production guidance.

[0069] Further, the step S1 further includes the following steps:

[0070] S01, a professional database for lightweight assembly mobile equipment is constructed, and the professional database integrates image data, three-dimensional models, technical drawings, bill of materials, material attribute parameters, cost data, and functional layout information of multiple equipment categories;

[0071] S02, the data in the database is cleaned, classified, and structurally labeled, and the labeling dimensions include equipment categories, design styles, structural components, functional areas, material types, and cost intervals;

[0072] S03, according to the labeled data, the concept generation module and the industrial transformation module are respectively supervised and trained and knowledge is fused.

[0073] Further, the concept generation module adopts a multi-modal neural network that integrates a Transformer structure and a generative adversarial network or a diffusion model, extracts multi-modal features of user input through a text encoder and an image encoder, and performs semantic alignment and feature enhancement in a fusion layer to generate high-quality visual conceptual schemes that meet aesthetic and functional requirements.

[0074] Further, the industrial transformation module performs manufacturability transformation on the selected conceptual scheme by integrating an engineering knowledge base, specifically including:

[0075] S11, according to the pre-defined equipment structure unit library and connection rules, the conceptual scheme is structurally decomposed to identify the load-bearing structure, the enclosure structure, and the functional module;

[0076] S12, according to the decomposed structure unit, combined with the material performance database and the cost database, the material and specification are automatically matched and optimized;

[0077] S13, automatically generating a structured bill of material containing component details, specifications, quantities, and assembly relationships according to the matching results and assembly logic;

[0078] S14, parametrically generating a technical drawing set including at least general assembly drawings, subassembly drawings, and part drawings in accordance with engineering drawing standards based on the structured bill of material and spatial relationships of equipment structural units.

[0079] The lightweight assembly type mobile equipment industrial design generation system based on artificial intelligence comprises:

[0080] The two-stage AI model comprises a concept generation module and an industrial transformation module.

[0081] The user interaction interface is used to receive user input multi-modal design requirements.

[0082] The processing unit is used to call the concept generation module to generate a visual concept scheme, and call the industrial transformation module to transform the user-selected concept scheme into an industrial-level design scheme.

[0083] The output unit is used to output an industrial design package containing at least one of a bill of materials, technical drawings, and manufacturing instructions.

[0084] Further, the system further comprises:

[0085] The integrated professional database stores image data, three-dimensional models, technical drawings, bills of materials, material attribute parameters, cost data, and functional layout information of multiple equipment categories, and is connected with the two-stage AI model.

[0086] The data preprocessing module is used to clean, classify, and multi-dimensionally structure the data in the professional database.

[0087] The model training and optimization module is used to supervise the training and continuous optimization of the concept generation module and the industrial transformation module according to the labeled data.

[0088] The user feedback collection module is used to collect user rating data, modification behavior, and adoption results of the generated scheme.

[0089] The cloud deployment and scheduling module is used to encapsulate the trained two-stage AI model as an API service and deploy it on a SaaS platform.

[0090] The knowledge base updating engine is used to dynamically adjust the rules and constraints in the industrial transformation module according to new materials, new processes, or new standards.

[0091] Further, the system further comprises:

[0092] A scheme manufacturability verification module is configured to perform virtual assembly simulation, structural strength analysis, material utilization assessment, and manufacturing cost review on the industrial design scheme.

[0093] A supply chain coordination interface is configured to data interface the generated bill of materials with external supplier systems.

[0094] A design version management unit is configured to record and manage versions of the conceptual scheme and the industrial scheme generated by the user at different stages.

[0095] A cross-platform output adapter is configured to convert the industrial design package into various mainstream CAD formats and manufacturing execution system compatible data structures.

[0096] Embodiment 2

[0097] In this embodiment, a lightweight mobile shop for urban flash retail is designed as an example to further illustrate the specific implementation process of the present application:

[0098] 1. User input: The user selects a mobile shop in the category through the SaaS platform interface, selects a modern minimalist style, and inputs the text requirement: a large glass display window is needed, which contains a product display area and a small cash desk, the overall weight needs to be controlled below 1.5 tons, and a modern building facade picture is uploaded as a style reference.

[0099] 2. Concept generation: The concept generation module (a multi-modal neural network combining Transformer and diffusion model) performs semantic analysis and feature fusion on the user input, the text encoder extracts key functional requirements such as glass display window and product display area, the image encoder extracts line, color and material features from the reference picture, and the network outputs four different layout and facade design modern minimalist style mobile shop three-dimensional effect pictures within 3 minutes for the user to choose.

[0100] 3. Industrial transformation: After the user selects one of the schemes with aluminum profiles and tempered glass as the main facade, the industrial transformation module is triggered, which performs the following automatic processing based on graph neural networks:

[0101] Structural decomposition: identifies the main load-bearing frame (aluminum profile), the enclosure structure (glass panel, composite panel), and the functional module (display rack, cash desk).

[0102] Material matching and BOM generation: According to the lightweight (density) and strength (yield strength) requirements, 6061 aluminum alloy profiles are matched from the material library for the load-bearing frame, and 12mm tempered glass is matched for the facade. A structured bill of materials (BOM) containing all profile specifications, glass sizes, connectors (stainless steel screws, angle codes) and quantities is automatically generated, totaling 78 entries, with real-time cost estimates.

[0103] Drawing generation: A complete set of technical drawings is generated parametrically, including general assembly drawings (showing overall size and layout), sub-assembly drawings (showing frame connection methods), and glass panel and display rack part drawings.

[0104] Manufacturing instruction output: Synchronous generation of aluminum profile cutting code for CNC machine tools and assembly process guidebook.

[0105] 4. Scheme delivery and verification: The manufacturability verification module of the scheme performs virtual assembly simulation and finite element structural strength analysis on the generated industrial scheme, confirming that the maximum stress is below the allowable stress of the material and the stability is qualified. Finally, the user sends the core materials in the BOM table to the cooperating aluminum profile and glass suppliers for inquiry through the supply chain collaboration interface.

[0106] Comparative example

[0107] To demonstrate the superiority of the present application, a comparative example is set up to show the process of completing and implementing the same design task (modern minimalist style mobile shop) using the traditional manual design process:

[0108] 1. Demand communication and concept design: The designer and the user communicate through multiple meetings, which takes about 2 days, and draws 3 versions of concept sketches.

[0109] 2. Three-dimensional modeling: The designer uses SketchUp or Rhino software to create a three-dimensional model based on the selected sketches, which takes about 3 days.

[0110] 3. Structural design and material selection: The structural engineer intervenes, analyzes the structural rationality of the three-dimensional model, and uses CAD software to redraw the structural diagram, while manually querying the material manual for selection and cost estimation. This process takes about 5 days.

[0111] 4. Engineering drawing: The engineer draws detailed general assembly drawings, sub-assembly drawings and part drawings, which takes about 3 days.

[0112] 5. Scheme review: Multiple rounds of review and modification are carried out within the team, which takes about 2 days.

[0113] The traditional process takes about 15 working days from concept to producible scheme, and is heavily dependent on the professional experience and collaboration efficiency of designers and engineers.

[0114] Experimental data and interpretation

[0115] To quantitatively evaluate the effect of the present application, we selected 10 different types of mobile equipment design tasks (including dining cars, retail cars, display cabins, etc.), and used the AI system of the present application (experimental group) and the traditional manual process (control group) respectively to implement them. The key data comparison is as follows:

[0116]

[0117] 1. Efficiency and cost revolution: Experimental data shows that the present application shortens the design cycle from the "week / day" level to the "hour" level, and reduces the cost from the "ten thousand yuan" level to the "hundred yuan" level, which is mainly due to the realization of full-link automation by the two-stage AI model, and the elimination of communication and time cost of multi-link cooperation of manual work.

[0118] 2. Innovation and diversity: The AI system can generate more conceptual schemes in a short time, which is due to the strong combination innovation and style transfer ability of the concept generation module under the training of massive data, which breaks through the experience limitations of individual designers.

[0119] 3. Manufacturing feasibility and quality: The high first-time manufacturability rate proves that the embedded engineering knowledge base (structural mechanics rules, material constraints, cost data) in the industrial transformation module plays a key "design for manufacturing" guarantee role in the scheme generation stage, avoiding design defects from the source.

[0120] 4. User satisfaction: The improvement of user satisfaction is on the one hand due to the extremely fast response speed and extremely low cost, and on the other hand because the AI generated scheme performs well in professionalism, feasibility and fit, and realizes continuous optimization through the user feedback collection module.

[0121] In summary, the experimental data fully verify the significant technical progress and commercial value of the present application in improving design efficiency, greatly reducing cost, enhancing innovation ability and ensuring manufacturing feasibility.

[0122] Although embodiments of the present application have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the present application, and the scope of the present application is defined by the appended claims and their equivalents.

Claims

1. A method for generating industrial design for lightweight prefabricated mobile equipment based on artificial intelligence, characterized in that, Includes the following steps: S1. Construct and train a two-stage AI model, which includes a concept generation module and an industrial transformation module; S2. Receive multimodal design requirement information input by the user; S3. Generate at least one visual concept design scheme through the concept generation module; S4. In response to the user's selection instruction for the visual concept design scheme, the selected concept scheme is transformed into an industrial-grade design scheme through the industrial conversion module. The industrial-grade design scheme includes at least one of a bill of materials, technical drawings, and manufacturing instructions.

2. The method for generating industrial design for lightweight prefabricated mobile equipment based on artificial intelligence according to claim 1, characterized in that, The concept generation module uses a multimodal neural network model, which receives at least one of text, images, style options and function parameters as input, and outputs the corresponding 3D rendering or 3D model.

3. The method for generating industrial design for lightweight prefabricated mobile equipment based on artificial intelligence according to claim 1, characterized in that, The industrial transformation module adopts a graph neural network or knowledge graph model, which integrates structural mechanics rules, material property parameters and manufacturing cost constraints, and automatically analyzes the structure and layout of the conceptual scheme to generate industrial-grade data.

4. The method for generating industrial design for lightweight prefabricated mobile equipment based on artificial intelligence according to claim 1, characterized in that, The industrial conversion module performs at least one of the following operations: Identify the main structure, load-bearing components, and functional zones in the conceptual model; Generate a bill of materials based on the materials knowledge base; Automatically generate 2D engineering drawings and 3D assembly drawings; Generate process specifications or CNC machining codes to guide production.

5. The method for generating industrial design for lightweight prefabricated mobile equipment based on artificial intelligence according to claim 1, characterized in that, The following steps are included before step S1: S01. Construct a professional database for lightweight prefabricated mobile equipment. The professional database integrates image data, 3D models, technical drawings, bills of materials, material property parameters, cost data and functional layout information for multiple equipment categories. S02. Clean, classify and structure the data in the database. The labeling dimensions include equipment category, design style, structural components, functional area, material type and cost range. S03. Based on the labeled data, supervised training and knowledge fusion are performed on the concept generation module and the industrial transformation module, respectively.

6. The method for generating industrial design for lightweight prefabricated mobile equipment based on artificial intelligence according to claim 1, characterized in that, The concept generation module employs a multimodal neural network that integrates a Transformer structure with a generative adversarial network or a diffusion model. It extracts multimodal features from user input through a text encoder and an image encoder, and performs semantic alignment and feature enhancement in the fusion layer to generate a high-quality visual concept solution that meets aesthetic and functional requirements.

7. The method for generating industrial design for lightweight prefabricated mobile equipment based on artificial intelligence according to claim 1, characterized in that, The industrial transformation module performs manufacturability transformation on selected conceptual solutions through an integrated engineering knowledge base, specifically including: S11. Based on the predefined equipment structure unit library and connection rules, the conceptual scheme is structurally decomposed to identify the load-bearing structure, enclosure structure and functional modules. S12. Based on the decomposed structural units, and in conjunction with the material performance database and cost database, perform automatic matching and optimization selection of materials and specifications. S13. Based on the matching results and assembly logic, automatically generate a structured bill of materials containing component details, specifications, quantities and assembly relationships. S14. Based on the structured bill of materials and the spatial relationships of the equipment structural units, generate a set of technical drawings that conforms to engineering drawing standards, including at least general assembly drawings, sub-assembly drawings, and part drawings.

8. An industrial design generation system for lightweight prefabricated mobile equipment based on artificial intelligence, characterized in that, include: A two-stage AI model, comprising a concept generation module and an industrial transformation module; User interaction interface, used to receive user input for multimodal design requirements; The processing unit is used to call the concept generation module to generate a visual concept scheme, and to call the industrial transformation module to transform the user-selected concept scheme into an industrial-grade design scheme. The output unit is used to output an industrial design package containing at least one of a bill of materials, technical drawings, and manufacturing instructions.

9. The lightweight prefabricated mobile equipment industrial design generation system based on artificial intelligence according to claim 8, characterized in that, The system also includes: An integrated professional database stores image data, 3D models, technical drawings, bills of materials, material property parameters, cost data, and functional layout information for multiple equipment categories, and is connected to the dual-stage AI model. The data preprocessing module is used to clean, classify, and perform multi-dimensional structured annotation on the data in the professional database. The model training and optimization module is used to perform supervised training and continuous optimization of the concept generation module and the industrial transformation module based on the labeled data. The user feedback collection module is used to collect user ratings, modification actions, and adoption results for the generated solutions; The cloud deployment and scheduling module is used to encapsulate the trained two-stage AI model into an API service and deploy it on the SaaS platform; The knowledge base update engine is used to dynamically adjust the rules and constraints in the industrial transformation module based on new materials, new processes, or new standards.

10. The lightweight prefabricated mobile equipment industrial design generation system based on artificial intelligence according to claim 8, characterized in that, The system also includes: The manufacturability verification module is used to perform virtual assembly simulation, structural strength analysis, material utilization assessment and manufacturing cost verification of industrial design schemes. The supply chain collaboration interface is used to connect the generated bill of materials with external supplier systems. The version management unit is designed to record and retrospectively manage the version of conceptual and industrial solutions generated by users at different stages. A cross-platform output adapter for converting industrial design packages into data structures compatible with various mainstream CAD formats and manufacturing execution systems.

Images