Multi-level-of-detail three-dimensional model control system and device applied to digital twinning

Abstract

The invention provides a multi-level-of-detail three-dimensional model control system and device applied to digital twinning, and development work of a digital twinning system is carried out through standard and systematic data collection, model design and multi-level-of-detail three-dimensional model manufacturing. The system can evaluate the use environment and modeling objects of the digital twin system, and designs and optimizes the model used by the digital twin system; universal assets are flexibly used, and the three-dimensional model manufacturing speed is increased; by making three-dimensional model making standards and specifications, a project can effectively and efficiently carry out multi-person and multi-post cooperation, and finally by constructing a novel database, storage and integration of data, management and application of production materials and multi-angle and multi-detail-level display and evaluation of the model are achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of computer three-dimensional modeling, and in particular to a multi-level-of-detail three-dimensional model control system and device applied to digital twins. Background Art

[0002] Early digital twin systems had problems such as inconsistent 3D modeling accuracy standards, crude 3D animations, and uneven system image quality. With the development and iteration of computer hardware, new requirements have been put forward for digital twin systems.

[0003] With the improvement of computer hardware performance, computers have more powerful computing chips, which can render three-dimensional models with more complex structures and finer materials in real time without affecting the display rate, thereby obtaining better picture effects.

[0004] Therefore, in the digital twin system of the new era, higher quality 3D models should be used to create more realistic 3D simulation animations, render more realistic factory environments, and give full play to the performance of the new generation of computer hardware.

[0005] Higher-quality 3D models can give digital twin systems more functions. For example, the roaming function allows employees to get familiar with the factory environment in a simulated factory; the industrial training function allows vocational skills training without delaying production; and the equipment explosion and disassembly demonstration function allows employees to understand the internal details and operation of the equipment without delaying production.

[0006] However, simply increasing the details and precision of the model is not necessarily beneficial to the digital twin system:

[0007] On the one hand, the production of high-precision models is very time-consuming; on the other hand, the rendering of high-precision three-dimensional models takes up a lot of computer hardware performance.

[0008] Therefore, it is necessary to preferably use three-dimensional models with different levels of detail to build a digital twin system and to reasonably optimize the accuracy of the three-dimensional model and its display effect.

[0009] However, there has not been an effective multi-level-of-detail 3D model control system applied in the development of digital twin projects in this industry.

[0010] like Figure 1 As shown, before the application of the present invention, the three-dimensional model production process of the digital twin system has the problems of cumbersome and repetitive procedures, difficult model version control, overlapping work processes, and low efficiency of collaborative work among multiple trades and multiple personnel.

[0011] Therefore, how to design an effective multi-level of detail three-dimensional model control system to assist the development process of the digital twin system is a technical problem that needs to be solved urgently in the industry.

[0012] Patent document CN111061937A discloses a standardized design platform and a standardized design method. The standardized design platform system includes: a server side and at least one terminal. The terminal is provided with a drawing module, an upload module and a download module. The server side is provided with a design standard library; the terminal stores the standard parts constructed in the drawing module into the design standard library through the upload module and / or calls the standard parts in the design standard library to the drawing module through the download module.

[0013] However, the content of the patent system is relatively general and is a general solution for multiple people working together. It does not provide a specific solution for a specific industry or a specific instance. Summary of the invention

[0014] In view of the defects of the prior art, the object of the present invention is to provide a multi-level-of-detail three-dimensional model control system and device for digital twins.

[0015] The multi-level-of-detail three-dimensional model control system applied to digital twins provided by the present invention includes:

[0016] System control unit: provides account login function, manages accounts, and manages user permissions;

[0017] Evaluation unit: provides a variety of evaluation functions, evaluates the total number of faces and the total number of models of the 3D model based on the existing data, and evaluates the performance of the user's hardware system based on the existing data; if there is a lack of detailed surveying and mapping data or drawing data, the AI ​​algorithm is used to evaluate and calculate the equipment and plant dimensions based on the existing data;

[0018] 3D model control unit: standardize the 3D model production method, production process, naming management, and production accuracy, and ultimately produce a 3D model that can be included in the model library;

[0019] Export unit: provides the functions of exporting models and materials in multiple formats and converting between multiple formats, standardizes the model export process, and performs classification management;

[0020] Rapid design library unit: provides a general resource library consisting of 3D modeling, textures, and material resources that meet modeling standards, with classified storage and search functions, and unified model production standards;

[0021] Database unit: Classify and store the models and various data documents produced in this project, allow authorized personnel to search and edit various materials, and allow existing models to be displayed and exported.

[0022] Preferably, the system control unit comprises:

[0023] User login module: provides user login function, enabling different personnel to log in on the same or different terminals, and realize multi-personnel, multi-platform, and multi-account collaborative operations;

[0024] Permission management module: provides permission management functions to manage user permissions, so that multiple users can work together in an orderly manner, and users with different permissions do not interfere with each other;

[0025] Security module: provides functions such as automatic backup and multi-media storage to ensure the security of production data;

[0026] Account management module: provides functions such as storage, identification, and management of account information, and records the user's operation time and content on the system.

[0027] Preferably, the evaluation unit comprises:

[0028] Model evaluation module: Use AI algorithms and manual intervention to evaluate the time required for 3D modeling, the volume of 3D modeling, the number of faces required for models of different precisions, the total number of static models, and the total number of movable models;

[0029] Computer performance evaluation module: evaluates the user's computer performance based on software testing and computer hardware data;

[0030] Data evaluation module: Use AI algorithms or manual intervention to process the initial data and obtain abstract and digital modeling object data;

[0031] Comprehensive evaluation design module: Combining the model evaluation results and hardware performance evaluation results, a reference document is formed for the 3D model specifications used in different business scenarios, so that the balance among model accuracy, hardware performance, and display effect can be reasonably evaluated when formulating model production specifications.

[0032] Preferably, the three-dimensional model control unit comprises:

[0033] Model management module: specifies the naming and storage rules for 3D models, stores 3D models in a preset model library, enables 3D models to be reused multiple times in different scenes, and enables accurate search of 3D models in the model library according to the naming specifications;

[0034] Model production specification module: formulate unified and universal 3D model production specifications and applicable 3D model production processes, so that multiple model engineers can work together to produce 3D models with consistent standards;

[0035] Model feature module: It has the significant features of the modeling target and can quickly associate the modeling target in any perspective of the digital twin system, including the mapping, texture data, and material assets of the 3D model;

[0036] Multi-level detail model detail module: Based on the model feature module, model details with multiple levels of detail are produced according to the needs of the digital twin system, and classified and managed;

[0037] Animated model module: Use higher model production standards to produce three-dimensional models with animation production requirements, and specify the model's axial direction, axis position, motion trajectory, motion mode, and bound bone data.

[0038] Preferably, the export unit comprises:

[0039] Multi-format export module: provides the function of exporting 3D models and their materials and textures in multiple formats;

[0040] Format conversion module: provides the function of converting the formats of models and materials;

[0041] Export specification module: formulate corresponding export process specifications. After exporting the models and materials in the format required by the digital twin system, name and classify the exported files in accordance with the preset rules in a timely manner, and incorporate them into the database.

[0042] Preferably, the rapid design library unit comprises:

[0043] Storage module: stores existing 3D models and their materials, maps and textures;

[0044] Search module: Search for existing 3D models and their materials, maps, and textures;

[0045] Classification module: classify existing 3D models and their materials, maps, and textures;

[0046] Naming management module: Naming and managing existing 3D models and their materials, maps, and textures;

[0047] Version control module: perform version control on existing 3D models and their materials, maps, and textures.

[0048] Preferably, the database unit comprises:

[0049] Initial data storage module: obtain and store initial data and information, and only allow uploading and browsing;

[0050] Requirements document storage module: stores the digital twin system function design documents, environment model design documents, and equipment model design documents, so that model engineers can understand the requirements of the digital twin system for the three-dimensional model in a timely manner;

[0051] Data integration module: Analyze and extract initial data and information, use AI algorithms to analyze the image data of the modeling target, and obtain relevant data of the modeling target;

[0052] Model library module: stores and names newly created 3D models and their materials, maps, and textures, allows adding tags to related data, allows searching for related files, and performs version control on related files;

[0053] Export module: allows the export of 3D models and their materials, maps, and textures in multiple formats;

[0054] Format conversion module: allows the conversion of 3D models and their materials, maps, and textures between multiple formats;

[0055] Model display module: fully display the 3D model, switch between multi-level detail models of the same modeling object, and render the 3D model and its materials, maps, textures, lighting and shadows.

[0056] Preferably, the initial data collection process is as follows: according to the production process flow, image data of the modeling target is photographed; during the shooting process, modeling objects that meet the surveying and mapping conditions or objects near the modeling objects are surveyed and mapped, from small to large, to establish a dimensional reference system for the image data; according to the production line equipment model, data is collected on the equipment, and the data content includes relevant drawings of the production line equipment, equipment manuals, relevant production process diagrams, etc.; the collected data is classified and processed to form an initial database.

[0057] Preferably, the model making specifications include: the default unit setting of the 3D software; standardizing the model precision of each detail level of the 3D model; modeling by using four-sided face modeling; checking and repairing the defects on the model, checking whether there are redundant and useless isolated points, whether there are polygonal faces, whether there are open edges, and whether the normals are correct; setting the axis and axial direction of the 3D model of the movable parts; setting the smoothing group with reference to the surface structure of the modeling target, so that the arc of the 3D model is consistent with the arc of the actual device;

[0058] After the 3D model is completed, reset the 3D model; when splitting the UV of the 3D model, try to ensure that the edges of the same smooth group are not broken, and the UV fills the UV grid as much as possible. The UV of the general model needs to be split separately; when making the 3D model, it is advisable to make it from low-precision models to high-precision models in sequence, name the models of different detail levels according to the model naming rules, and classify and store them separately.

[0059] The multi-level-of-detail three-dimensional model control device applied to digital twins provided by the present invention is equipped with the multi-level-of-detail three-dimensional model control system applied to digital twins.

[0060] Compared with the prior art, the present invention has the following beneficial effects:

[0061] (1) The present invention provides a multi-level-of-detail three-dimensional model control system applied to digital twins, comprehensively evaluates the application environment of the digital twin system, and optimally designs and produces multi-level-of-detail three-dimensional models, thereby achieving the purpose of giving full play to the software and hardware performance of the digital twin system;

[0062] (2) Comprehensively integrate the results of data collection and data processing, formulate standardized modeling specifications and processes, improve the universality of models between multiple projects, effectively reduce the quality differences of 3D models produced by different 3D modeling engineers, and improve the collaboration efficiency between multiple personnel and multiple trades;

[0063] (3) Improving resource management and control capabilities, effectively reducing repetitive work, and lowering production costs can significantly improve the efficiency of the early stages of digital twin system development. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Other features, objects and advantages of the present invention will become more apparent from the detailed description of non-limiting embodiments made with reference to the following drawings:

[0065] Figure 1 This is an embodiment before the application of the present invention;

[0066] Figure 2 It is a schematic diagram of the system structure of the present invention;

[0067] Figure 3 It is a schematic diagram of the association between the example of the present invention and the industrial digital twin visualization system;

[0068] Figure 4 The figure is a schematic diagram of an example model tree after the application of the present invention. DETAILED DESCRIPTION

[0069] The present invention is described in detail below in conjunction with specific embodiments. The following embodiments will help those skilled in the art to further understand the present invention, but are not intended to limit the present invention in any form. It should be noted that, for those of ordinary skill in the art, several changes and improvements can also be made without departing from the concept of the present invention. These all belong to the protection scope of the present invention.

[0070] Example

[0071] The present invention formulates standardized modeling specifications and processes to effectively reduce the quality differences of 3D models produced by different 3D modeling engineers, improve the model versatility between multiple projects, form an effective and practical database unit, and preferably apply multi-level detailed 3D models to digital twins; improve resource management and control capabilities, reduce repetitive work, reduce production costs, greatly improve the efficiency of the early work of digital twin system development, and give full play to the software and hardware performance of the digital twin system.

[0072] like Figure 2 According to the present invention, a multi-level-of-detail three-dimensional model control system provided by the present invention includes:

[0073] System control unit: provides account login and account management functions, manages the rights of users, and ensures the safety of production materials

[0074] Evaluation unit: Evaluate the total number of faces and the total number of models based on the existing data, evaluate the performance of the user's hardware system based on the existing data, and formulate 3D model production specifications based on the comprehensive evaluation results to achieve a balance between model accuracy, hardware performance and display effect. If there is no detailed surveying and mapping data or drawing data, it is necessary to use AI algorithms to evaluate and calculate the equipment and plant dimensions based on the existing data;

[0075] 3D model control unit: standardize model making methods, standardize model making processes, standardize model naming methods, unify model making accuracy, form a model library, and save integration time;

[0076] 3D model export unit: provides model and material export and format conversion functions, standardizes model export process, and does a good job of classification management;

[0077] Rapid Design Library Unit: A general resource library consisting of 3D modeling, textures, and material resources that meet modeling standards and have the same classification storage and search functions as the database unit, which can save time in 3D model production.

[0078] Database unit: Classify and store the models and various data documents produced in this project, allow authorized personnel to search and edit various materials, and allow existing models to be displayed and exported.

[0079] Specifically, the system control unit includes the following functional modules:

[0080] User login module: provides user login function, enabling different personnel to log in on the same or different terminals, and realize multi-personnel, multi-platform, and multi-account collaborative operations;

[0081] Permission management module: manages user permissions so that multiple users can work together in an orderly manner and users with different permissions will not interfere with each other;

[0082] Production data security module: adopts automatic backup, multi-media storage and other methods to ensure the security of production data to the greatest extent possible, avoiding risks such as production data being deleted by mistake and data leakage;

[0083] Account management module: stores, identifies and manages user account information, and records user operation time and content.

[0084] Specifically, the evaluation unit includes the following functional modules:

[0085] Model evaluation module: After obtaining the original data of the 3D modeling target, professional modeling engineers or AI algorithms that have completed large-scale model training will interpret the original data, evaluate the size of the 3D modeling target, the number of faces required for multi-level-of-detail models, the total number of static models, and the total number of dynamic models, and evaluate the working hours required for 3D modeling based on the evaluation results.

[0086] The model evaluation algorithm includes the following steps:

[0087] (1) Data Collection

[0088] Collect data related to 3D modeling, such as the time taken by different model engineers to model different types and levels of detail.

[0089] (2) Data preprocessing

[0090] The raw data is cleaned, converted, and processed to obtain usable data.

[0091] (3) Feature extraction

[0092] Extract meaningful features from data for model building.

[0093] (4) Model construction

[0094] Build a prediction model based on the training data. There are usually the following types of prediction models:

[0095] Linear regression model, the mathematical model formula involved in the linear regression model is as follows:

[0096] y=β0+β1x1+β2x2+β3x3+…+β n x n +∈

[0097] Among them, y is the output variable, x1,x2,x3,…,x n are input variables, β0,β1,β2,β3+,…,+β n are model parameters, ∈ is the error term;

[0098] Support vector machine model, the mathematical model formula involved in the support vector machine model is as follows:

[0099] f(x)=sgn(ω T x+b)

[0100] Among them, f(x) is the output variable, ω is the model parameter, b is the bias term, and x is the input variable;

[0101] Decision tree model, the mathematical model formula involved in the decision tree model is as follows:

[0102]

[0103] Where D(x) is the output variable, d1, d2, …, d n are the leaf nodes of the decision tree, R1, R2, …, R n is the branch of the decision tree;

[0104] Neural network model, the mathematical model formula involved in the neural network model is as follows:

[0105] y=f(ω T x+b)

[0106] Among them, y is the output variable, ω is the model parameter, b is the bias term, x is the input variable, and f is the activation function.

[0107] (5) Model evaluation

[0108] Evaluate the performance of your model and make adjustments and optimizations.

[0109] (6) Model application

[0110] Apply the optimized model to the actual 3D modeling process, optimize task allocation, reasonably evaluate working hours, improve work efficiency, and rationally utilize the production capacity of 3D modeling engineers.

[0111] Data evaluation module: Use AI algorithms or manual intervention to process the initial data, obtain abstract and digital modeling object data, and use the preferred data identification method to identify the size of the modeling object;

[0112] In the process of data evaluation, it is necessary to pre-process the initial data in different ways and methods according to the format of the initial data.

[0113] For example, for CAD / Creo drawings, you can directly extract the model's dimensional information through the API, or export the dimensional information to a standard format (such as CSV format) and read the data in the corresponding application.

[0114] If there is no surveying and mapping data or CAD / Creo drawings with dimensional information, and only image data of the relevant modeling object is obtained, the image data needs to be processed to obtain the modeling object dimensional information required for 3D modeling.

[0115] Among them, the data algorithm evaluation processing stage includes the following steps:

[0116] (1) Preprocessing: Preprocessing methods include denoising, contrast enhancement, etc., to improve image quality and facilitate subsequent processing.

[0117] Common algorithms for denoising are:

[0118] Mean filtering: Sum the values ​​of the current pixel and its neighboring pixels, and divide the result by the number of neighboring pixels.

[0119] The algorithm formula involved is:

[0120]

[0121] Where F(x,y) is the filtered pixel value, f(x,y) is the original pixel value, and N is the number of neighboring pixels;

[0122] Median filtering: Sort the values ​​of the current pixel and its neighboring pixels, and then take the middle value.

[0123] The algorithm formula involved is:

[0124] F(x,y)=f(x,y)

[0125] Among them, F(x,y) is the pixel value after filtering, and f(x,y) is the original pixel value;

[0126] Gaussian filtering: Use Gaussian function as the filtering function to smooth and blur the image.

[0127] The algorithm formula involved is:

[0128]

[0129] Among them, F(x,y) is the pixel value after filtering, f(x,y) is the original pixel value, and σ is the standard deviation of the Gaussian filter.

[0130] (2) Edge detection

[0131] Edge detection algorithms are used to identify edges in an image. Common edge detection algorithms include:

[0132] Gradient method, the algorithm formula involved is:

[0133]

[0134] Among them, G(x,y) is the gradient value, Gx(x,y) and Gy(x,y) are the gradients in the x direction and y direction respectively;

[0135] Laplace method, the algorithm formula involved is:

[0136] L(x,y)=f(x,y)*(-Δ)

[0137] Where L(x,y) is the pixel value after Laplace filtering, f(x,y) is the original pixel value, and Δ is the Laplace operator;

[0138] Difstein method, the algorithm formula involved is:

[0139] D(x,y)=f(x,y)*(-Δ 2 )

[0140] Where D(x,y) is the pixel value after Difstein filtering, f(x,y) is the original pixel value, Δ 2 is the square of the Difstein operator.

[0141] (3) Contour recognition and tracking: Use FindContours in OpenCV to identify the shape of the modeled object.

[0142] (4) Appearance correction

[0143] During shooting, the image captured by the lens will distort the picture to a certain extent. At this time, the checkerboard correction method needs to be used to correct the image.

[0144] Use the chessboard for calibration, calculate the distortion coefficient of the lens, use the Fisheye module in OpenCV, use cv2.fisheye.calibrate to calculate the distortion coefficient according to the chessboard calibration result, use the cv2.fisheye.initUndistortRectifyMap function to calculate the mapping matrix, and use cv2.remap for correction according to the mapping matrix.

[0145] (5) Dimension measurement: Calculate the actual size of the part based on the scale mark in the image or the size of the known reference object. The calculation formula is:

[0146]

[0147] (6) Result output: The measured target size is output and marked on the image in digital form.

[0148] Hardware performance evaluation module: To evaluate the operating environment of the digital twin system, develop an optimal test software, and perform a performance evaluation on the computer hardware performance of the digital twin system based on the software test results and the acquired computer hardware data of the digital twin system.

[0149] Comprehensive evaluation design module: Combine the model evaluation results and hardware performance evaluation results to form a reference document for the 3D model specifications used in different business scenarios. Preferably, formulate model production specifications based on the reference document so that the digital twin system can achieve better demonstration effects under limited hardware conditions.

[0150] Specifically, the three-dimensional model control unit includes the following functional modules:

[0151] Model production specification module: In 3D model production, each model engineer has different 3D model production habits. Therefore, in order to ensure the standardization of the final 3D model product, a model specification needs to be formulated before the model is officially produced.

[0152] In the specific implementation, Figure 3 As shown in the figure, the industrial digital twin production visualization system includes a variety of functions, such as scene roaming function, equipment display function, explosion and splitting function, production animation demonstration function, material tracking function, fault alarm function, etc.

[0153] Each function of the industrial digital twin production visualization system has different requirements for the level of detail of the 3D model;

[0154] For example, in the scene roaming function, the focus is usually on the overall orientation layout of the models; the camera angle, running speed, running angle; the rendering effect and smoothness of the demonstration screen, etc. The accuracy requirements of the 3D model are lower than the details of the single model;

[0155] In this function, the number of single 3D models is usually higher than that of single 3D models in other functions. If you blindly pursue the details of the model, the number of triangles in the scene will be too high due to too many model details, which will occupy too many hardware resources, resulting in poor picture material and shadow rendering effects and reduced picture smoothness.

[0156] Therefore, in this function, the detail level of the 3D model can be reduced while retaining the features of the 3D model;

[0157] In the device display function, a single device is displayed independently in the system screen. At this time, higher requirements are placed on the external details of the model. Under the premise that the system performance allows, the refinement of the appearance of the device 3D model should be improved to enhance the demonstration effect of the model.

[0158] The explosion and splitting function places higher demands on the precision of the 3D model. It requires not only precise modeling of the device appearance, but also detailed modeling of the interior of the model, so as to realize the function of displaying the appearance and internal structure of a single model.

[0159] In the production animation demonstration function, the focus shifts from the model itself to the model action. At this time, the number of single models displayed in the screen is usually more than that in the single device display function, and less than that in the scene roaming function. Therefore, in this function, the level of detail of the model can be appropriately lower than that of the single model display function, but higher than that of the scene roaming function, so as to reasonably transfer more computing resources for better real-time rendering of production animation;

[0160] In the material tracking function, the object of attention is not the model itself, but the position of the model relative to the overall scene. At this time, the model itself can even be simplified and abstracted into a three-dimensional icon or even a picture or a text.

[0161] In the fault alarm function, the object of concern is not the model itself, but the position of the model relative to the overall scene and the position of the specific model relative to the single device. At this time, the model is required to be simplified, but the position of the model assembled on the single device model must be accurate and have geometric modeling features and material features that can quickly correspond to the physical equipment components.

[0162] In actual operation, the model making specifications should at least stipulate the following contents:

[0163] 1. Default unit settings for 3D software

[0164] 3D software generally supports multiple size units. Changing the default unit of the software after the 3D model is completed may cause confusion in the scale of the 3D model. Therefore, the unit should be specified at the beginning of the 3D modeling work.

[0165] In this way, when multiple model engineers work together, the final 3D model will have consistent size units.

[0166] 2. Model refinement at each level of detail of the 3D model

[0167] To realize the digital twin visualization system, the 3D model needs to be imported into the engine in advance. 3D models of different fineness have different requirements on computer performance, so the number of faces of the model needs to be controlled when making the 3D model;

[0168] Therefore, it is necessary to standardize the model refinement at each level by combining the design results of the multi-level-of-detail model;

[0169] For example, in the 3D model required for the explosion splitting function, each screw and each gear needs to be completely produced, and depending on the size of the parts, the 3D model of the parts needs to be chamfered to further enhance the appearance of details. In the scene overview function, only a feature model that can show the characteristics of the equipment needs to be placed in the corresponding position.

[0170] For example, in Figure 4 In the example of a digital twin production process simulation system in the steel industry shown in the figure, the following multi-level detail modeling regulations are made:

[0171] LOD0, 3D icons, simple patches, 2D images;

[0172] LOD1, a model built with simple geometric bodies, can basically reflect the appearance characteristics of the modeled object, and can express the area, volume, height, position, and orientation information of the modeled object. It is suitable for scenes with a viewing distance of more than 100 meters and can display large-scale factory areas, cities, etc.

[0173] LOD2, based on the LOD1 model, adds a feature detail model with an actual size greater than 1 meter, which can more accurately express the appearance of the modeled target. It is suitable for scenes with a viewing distance of 20-100 meters to show a single factory building or large equipment;

[0174] LOD3, based on the LOD2 model, adds chamfers, grooves, convexities and other details to models with actual sizes greater than 1 meter, and adds detailed models with actual sizes greater than 0.2 meters. It is suitable for viewing distances of 2-30 meters and presents a single device or production line;

[0175] LOD4, based on the LOD3 model, adds detailed models with a size greater than 0.02 meters. It is suitable for a viewing distance of 5 meters and provides detailed display of a single device or production line.

[0176] LOD5, based on the LOD4 model, adds chamfers, grooves, convexities and other details to models with actual sizes greater than 0.02 meters, suitable for rendering model running animations;

[0177] LOD6, based on LOD5, makes the internal structure of the model, and the internal structure model is made according to the LOD4 standard;

[0178] LOD7, based on LOD5, makes the internal structure of the model, and the internal structure model is made according to the LOD5 standard;

[0179] LOD8, a model for 3D printing, is made based on LOD7 or LOD5. The 3D model needs to be modified according to the requirements of the 3D printing model, because the model is scaled at a ratio of 1:100 during printing, and the accuracy of 3D printing is limited, so there is basically no need to add details to the model; and because the strength of 3D printing materials is limited, some small parts need to be thickened to avoid breakage during printing or transportation;

[0180] LOD9, a model for video rendering, is made based on LOD7 or LOD5, with no limit on the number of faces. It prioritizes the best rendering effect and models every detail;

[0181] LOD·TEST, a test model, several spherical monomer models, used in computer performance testing programs, with the number of faces ranging from more than 100,000 to hundreds of millions.

[0182] 3. Flatness of the model

[0183] In the process of making three-dimensional models, quadrilateral modeling is often used. Each quadrilateral is composed of two triangular faces. If the quadrilateral has a large bend, the two vertices of the quadrilateral need to be connected to make the model smoother.

[0184] 4. Model flaws

[0185] After completing the production of the 3D model, the defects on the model should be checked and repaired, including checking whether there are redundant and useless isolated points, whether there are polygons, whether there are open edges, whether the normals are correct, etc. The above situations may cause problems in UV unfolding, mapping, light and shadow rendering, etc.

[0186] 5. Axis and axial settings of the model

[0187] The three-dimensional models used in digital twins often have parts that need to move. At this time, the three-dimensional models of the movable parts must be set with targeted axis and axial directions so that their movement patterns in the animation and actual equipment movements are consistent.

[0188] If there is no special movement requirement, the model axis position should be set on the plane where the model contacts the ground.

[0189] 6. Smooth group settings of the model

[0190] The three-dimensional model should set up reasonable smoothing groups in reference to the surface structure of the modeling target so that the arc of the three-dimensional model is consistent with the arc of the actual equipment.

[0191] 7. Reset transformation of the model

[0192] After the 3D model is completed, the model itself will carry a lot of software operation history records. After multiple calculations by the 3D software, abnormalities may occur. Therefore, after the 3D model is completed, the 3D model should be reset in time to delete the historical records carried by the model.

[0193] 8. Model UV splitting and allocation

[0194] When splitting the UV of a 3D model, you should generally follow the principle that the edges of the same smooth group are not separated and the UV grid is filled as much as possible.

[0195] The UVs of common models should be split separately to optimize UV ​​resource allocation and make the models, their textures, and materials universally usable.

[0196] 9. Texture size control

[0197] In order to save hardware performance and meet the needs of model versatility, the material and texture size of the model should be controlled;

[0198] For example, a general bolt can use a 128*32 texture, an asphalt road can use a 1024*1024 loop texture, an excavator can use a 4096*4096 texture or four 2048*2048 textures, etc.

[0199] 10. 3D model production process

[0200] When making a three-dimensional model, the order should be from making a medium-precision model to a low-precision model and a high-precision model. When making it, it should be named according to the naming rules of models with different levels of detail.

[0201] In model making, models should be classified and stored separately in a proactive manner. For example, the three-dimensional models of parts such as rivets and gears will appear repeatedly in the subsequent production process and can be reused in other projects. They should be stored in the rapid design library as general models after production is completed, and reused in the next appropriate application scenario, thereby reducing the production time and cost of the three-dimensional model.

[0202] Model management specification module: standardize the naming, classification and format of models and materials;

[0203] Specifically, at least the following contents should be stipulated:

[0204] 1. Naming of 3D models and materials

[0205] In specific implementation, the names of models and materials should be unique and consist of any combination of numbers, letters or special characters. Chinese characters can be used for naming if the system and software allow it.

[0206] If there is no special need, the naming of models should be carried out according to obvious rules, such as the name and model of the three-dimensional model in reality; the number on the specific equipment and other well-documented rules.

[0207] 2. Naming of 3D model materials, textures and other assets

[0208] In specific implementations, the materials, textures and other assets of the 3D model should correspond one-to-one with the names of the 3D model. However, if the materials, textures and other assets are general assets, they can also follow the naming rules corresponding to the real names or the naming rules of related general assets.

[0209] 3. Classification of 3D models and materials

[0210] In specific implementation, three-dimensional models should be classified and stored according to the actual purpose of the model, storage format, asset naming rules, etc., to facilitate the search, modification and call of models and materials.

[0211] Model feature module: It has the significant features of the modeling target and can quickly associate the modeling target from any perspective of the digital twin system; the model feature module also includes the 3D model's mapping, texture data, material assets and other content.

[0212] In the specific implementation, modeling engineers are required to use the simplest model possible to express the modeling characteristics of the modeling object, laying the foundation for the subsequent multi-level detailed model production. The model modeling, number of triangles, and model wiring in this module are all optimized, and there is no room for further optimization;

[0213] Model multi-level of detail module: Based on the 3D model requirements of each function of the digital twin system, on the basis of the model feature module, a multi-level of detail model is produced in accordance with the 3D model production specifications, and classified and managed according to the level of detail, so that the multi-level of detail model can be optimally applied in the digital twin system;

[0214] Movable model module: Three-dimensional models with animation production requirements should be produced using higher model production standards, and the model's axial direction, axis position, motion trajectory, motion mode, bound bones and other data must be specified to ensure that the three-dimensional model can operate normally in motion animation.

[0215] A multi-detail model application management system rapid design library unit provided by the present invention has the following functions:

[0216] The rapid design library unit is essentially a partial reference to an existing optimized database, which classifies, stores and manages existing 3D models and material textures, and can search, update and reference the above data.

[0217] A multi-level-of-detail three-dimensional model control system database unit provided by the present invention has the following functional modules:

[0218] Original data module: Preferably, the original data of the modeling object is obtained using multiple ways and methods. Among them, the method of obtaining the original data includes:

[0219] Use laser radar to scan the modeling target and obtain the three-dimensional point cloud data of the modeling target; obtain the CAD data of the modeling target, obtain the remote sensing data of the modeling range; obtain the surveying and mapping data of the modeling target, etc.;

[0220] In specific implementation, there are usually four ways to model digital twins using three-dimensional models: GIS (geographic information system) spatial modeling, oblique photography modeling, BIM (building information modeling) modeling, and manual modeling.

[0221] Specifically, GIS spatial modeling is a method of automatically generating three-dimensional models based on GIS data. The models generated by this method are usually simple and non-textured models.

[0222] Specifically, oblique photography modeling is a method of automatically generating 3D models based on aerial photography and close-range measurement technology. This method can fully perceive complex scenes in a large-scale, high-precision, and high-definition manner. However, this automatic generation method has many problems: for example, the number of 3D models is huge, the bottom of the model is greatly deformed due to the large number of occlusions in densely built areas, and the feature points of objects with small cross-sections such as wires are rarely matched, resulting in missing models.

[0223] Specifically, BIM is a new tool in architecture, engineering and civil engineering. It is defined as a building or construction engineering information model that consists of complete and sufficient information to support new product development management and can be directly interpreted by computer applications. BIM usually contains a huge amount of information and a large number of model faces. The BIM model needs to be lightweight before it can be used in the digital twin system.

[0224] Specifically, manual modeling is a method in which a professional 3D modeling engineer manually enters the data of the modeling object into the 3D software to produce a 3D model. This method is the most widely used method in the production process of digital twin 3D models. The defects of the above three modeling methods can all be compensated by manual modeling.

[0225] In the specific implementation, a necessary modification review system should be set up, and the original data can only be overwritten or deleted after being reviewed by high-authority personnel to avoid data loss due to misoperation;

[0226] by Figure 4 Taking the steel production simulation system shown in the figure as an example, when collecting the original data of the production line, the user who customized this system only provided CAD drawings and a small amount of image data when the production line was refurbished more than ten years ago.

[0227] In order to ensure that the 3D modeling work can be carried out normally, preferably, this system provides a solution for collecting original data:

[0228] Step 1: Take images of the modeling target according to the production process. During the shooting process, you need to use the optimal shooting angle to shoot the modeling target. For example, the image data is required to clearly show the volume relationship and relative position of the production line equipment, or to clearly show the details and structure of the production line equipment;

[0229] Step 2: During the shooting process, the modeling object or the object near the modeling object that has the conditions for surveying and mapping is surveyed and mapped, from small to large, to establish a size reference system for the image data, so as to facilitate the size assessment of the modeling object in the subsequent size assessment work;

[0230] Step 3: Collect relevant information about the equipment according to the production line equipment model, including relevant drawings of the production line equipment, equipment manuals, relevant production process diagrams, etc.

[0231] Step 4: Classify and process the collected data to form an original database.

[0232] Model library module: In the specific implementation, the original data of the modeling object, the processed data, the completed 3D models and texture materials, and the 3D models and texture materials of various historical versions are classified and stored, integrated into an asset database, and labels are added to the above data files so that the data files can be quickly and accurately consulted, called, and modified in the database;

[0233] Demand document storage module: stores preliminary design documents such as digital twin system function design documents, environment model design documents, and equipment model design documents, so that model engineers can timely understand the changes in the digital twin system's requirements for the 3D model through document updates;

[0234] Data integration module: Analyze and extract initial data and information. If necessary, use AI algorithms to analyze the image data of the modeling target to obtain data for making a three-dimensional model of the modeling target.

[0235] Export module: Different engines need to import 3D models, materials, and texture files in different formats to develop the digital twin system in the engine. Therefore, it is necessary to export the 3D model and its textures and materials in different formats.

[0236] Format conversion module: convert the format of models and materials;

[0237] Export specification module: When exporting models and materials in the format required by the digital twin system, the exported files should be named and classified in accordance with relevant regulations in a timely manner. And included in the database, it is convenient to call it again in the future and control the version of the digital twin system, avoiding time waste and version confusion caused by re-export;

[0238] Display module: The 3D model can be displayed in all directions before entering the engine. The multi-level detailed models of the same modeling object can be switched. The material, texture, lighting and shadow effects of the 3D model can be adjusted at multiple levels to show the display effects under different rendering accuracies.

[0239] Those skilled in the art know that, in addition to implementing the system, device and its various modules provided by the present invention in a purely computer-readable program code, it is entirely possible to implement the same program in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers and embedded microcontrollers by logically programming the method steps. Therefore, the system, device and its various modules provided by the present invention can be considered as a hardware component, and the modules included therein for implementing various programs can also be considered as structures within the hardware component; the modules for implementing various functions can also be considered as both software programs for implementing the method and structures within the hardware component.

[0240] The above describes the specific embodiments of the present invention. It should be understood that the present invention is not limited to the above specific embodiments, and those skilled in the art can make various changes or modifications within the scope of the claims, which does not affect the essence of the present invention. In the absence of conflict, the embodiments of the present application and the features in the embodiments can be combined with each other arbitrarily.

Claims

1. A multi-level-of-detail three-dimensional model control system applied to digital twins, characterized in that: include: System control unit: provides account login function, performs account management, and performs permission management for users; Evaluation unit: Evaluate the total number of faces and total number of models of the 3D model based on the existing data, and evaluate the performance of the user's hardware system based on the existing data; if there is a lack of detailed surveying and mapping data or drawing data, use AI algorithms to evaluate and calculate the equipment and plant dimensions based on the existing data; 3D model control unit: standardize model making methods and processes, manage model naming, unify model making accuracy, and form a model library; Export unit: provides multi-format export and multi-format conversion functions for models and materials, standardizes the model export process, and classifies and manages the exported files; Rapid design library unit: a general resource library consisting of 3D modeling, textures, and material resources that meet modeling standards, with classified storage and search functions, and unified model production standards; Database unit: Classify and store the models and various data documents produced in this project, allow authorized personnel to search and edit various materials, and allow existing models to be displayed and exported.

2. The multi-level-of-detail three-dimensional model control system applied to digital twins according to claim 1, characterized in that: The system control unit comprises: User login module: Use the user login function to enable different personnel to log in on the same or different terminals, and realize multi-personnel, multi-platform, and multi-account collaborative operations; Permission management module: manages user permissions, enables multiple users to work together in an orderly manner, and prevents users with different permissions from interfering with each other; Security module: automatic backup and multi-media storage are used to ensure the security of production data; Account management module: stores, identifies and manages user account information, and records the user's operation time and content on the system.

3. The multi-level-of-detail three-dimensional model control system applied to digital twins according to claim 1, characterized in that: The evaluation unit comprises: Model evaluation module: Use AI algorithms and manual intervention to evaluate the time required for 3D modeling, the volume of 3D modeling, the number of faces required for models of different precisions, the total number of static models, and the total number of movable models; Computer performance evaluation module: evaluates the user's computer performance based on software testing and computer hardware data; Data evaluation module: Use AI algorithms or manual intervention to process the initial data and obtain abstract and digital modeling object data; Comprehensive evaluation design module: Combining the model evaluation results and hardware performance evaluation results, a reference document is formed for the 3D model specifications used in different business scenarios, so as to achieve a balance among model accuracy, hardware performance, and display effect in the model production specifications.

4. The multi-level-of-detail three-dimensional model control system applied to digital twins according to claim 1, characterized in that: The three-dimensional model control unit comprises: Model management module: stipulates the naming rules and storage rules of 3D models, standardizes the naming and storage methods of 3D models, stores 3D models in the model library, enables 3D models to be reused multiple times in different scenes, and enables accurate search and finding of 3D models in the model library according to the naming specifications; Model production specification module: formulate unified and universal 3D model production specifications, formulate applicable 3D model production processes, and enable multiple model engineers to work together to produce 3D models with consistent standards; Model feature module: It has the significant features of the modeling target and can quickly associate the modeling target in any perspective of the digital twin system; the model feature module also contains the mapping, texture data, and material asset content of the 3D model; Multi-level-of-detail model module: Based on the model feature module, model details with multiple levels of detail are produced according to the needs of the digital twin system, and classified and managed according to the level of detail; Animated model module: Use higher model production standards to produce three-dimensional models with animation production requirements, and specify the model's axial direction, axis position, motion trajectory, motion mode, and bound bone data.

5. The multi-level-of-detail three-dimensional model control system applied to digital twins according to claim 1, characterized in that: The export unit comprises: Multi-format export module: export 3D models and their materials and textures in multiple formats; Format conversion module: convert the formats of models and materials; Export specification module: After exporting the models and materials in the format required by the digital twin system, the exported files are named and classified in a timely manner according to the preset rules and incorporated into the database.

6. The multi-level-of-detail three-dimensional model control system applied to digital twins according to claim 1, characterized in that: The rapid design library unit comprises: Categorize, store and manage existing 3D models and their materials, maps and textures, and search, update and reference the above data; Newly created general models, general textures and materials are collected into the library and classified, stored and managed.

7. The multi-level-of-detail three-dimensional model control system applied to digital twins according to claim 1, characterized in that: The database unit comprises: Initial data storage module: obtain and store initial data and information, and only allow uploading and browsing; Requirements document storage module: stores the digital twin system function design documents, environment model design documents, and equipment model design documents, so that modeling engineers can promptly understand the requirements of the digital twin system for the 3D model; Data integration module: Analyze and extract initial data and information, use AI algorithms to analyze the image data of the modeling target, and obtain data for making a three-dimensional model of the modeling target; Model library module: stores and names newly created 3D models and their materials, maps, and textures, allows adding tags to related data, allows searching for related files, and performs version control on related files; Export module: allows the export of 3D models and their materials, maps, and textures in multiple formats; Format conversion module: allows the conversion of 3D models and their materials, maps, and textures between multiple formats; Model display module: fully display the 3D model, switch between multi-level detail models of the same modeling object, and render the 3D model and its materials, maps, textures, lighting and shadows.

8. The multi-level-of-detail three-dimensional model control system applied to digital twins according to claim 1, characterized in that: The initial data collection process is as follows: according to the production process, the image data of the modeling target is photographed; during the shooting process, the modeling objects that meet the surveying and mapping conditions and the objects near the modeling objects are surveyed and mapped, from small to large, to establish a dimensional reference system for the image data; according to the production line equipment model, the relevant equipment data is collected, and the data content includes relevant drawings of the production line equipment, equipment manuals, and relevant production process diagrams; the collected data is classified and processed to form an initial database.

9. The multi-level-of-detail three-dimensional model control system applied to digital twins according to claim 4, characterized in that: Model making specifications include: default unit settings for 3D software; specification of model precision at each level of detail of the 3D model; modeling using four-sided face modeling; checking and repairing defects on the model, checking whether there are redundant and useless isolated points, whether there are polygons, whether there are open edges, and whether the normals are correct; setting the axis and axial direction of the 3D model of movable parts; setting the smoothing group with reference to the surface structure of the modeling target, so that the arc of the 3D model is consistent with the arc of the actual device; After the 3D model is completed, reset the 3D model; when splitting the UV of the 3D model, ensure that the edges of the same smooth group are not broken, the UV occupies the UV grid, and the UV of the general model is split separately; when making the 3D model, start from low-precision models to high-precision models, name the models of different detail levels according to the naming rules of the model, and classify and store them separately.

10. A multi-level-of-detail three-dimensional model control device applied to digital twins, characterized in that: A multi-level-of-detail three-dimensional model control system for digital twins according to any one of claims 1 to 9.

Images