Equipment digital simulation method based on PBR process

Abstract

The invention provides an equipment digital simulation method based on a PBR process, and belongs to the field of computer graphics, and the method comprises the following steps: S1, carrying out digital modeling by using industrial modeling software SOLIDWORKS according to the actual size of a product; s2, the curved surface model created in the SOLIDWORKS is exported to 3ds Max software; s3, the curved surface model is optimized in 3ds Max software; s4, importing the optimization model into Substance Painter software, and carrying out chartlet baking on the optimization model and the Substance Painter software; s5, carrying out detail enhancement on the chartlet on the basis of AI (Artificial Intelligence) in Substance Painter software; and S6, carrying out interactive rendering and digital twinning on the equipment model based on Unity. According to the method, full-process optimization of high-precision digital simulation of equipment is realized through multi-software collaboration and technical innovation.

Description

Technical Field

[0001] This application belongs to the field of computer graphics technology, specifically relating to a device digital simulation method based on the PBR process. Background Technology

[0002] In industrial fields such as mining machinery and heavy equipment, building high-precision digital simulation models of equipment is of great significance for product design verification, remote operation and maintenance monitoring, operator training, and fault prediction. With the development of digital twin technology, higher demands are placed on the realism, interactivity, and real-time performance of equipment digital models. Traditional 3D modeling methods usually render high-polygon models generated by CAD software (such as SOLIDWORKS). However, such models suffer from problems such as excessive polygon count, complex topology, and the inclusion of a large number of non-manifold geometries, making it difficult to meet the processing requirements of graphics software in areas such as UV unwrapping, texture baking, and real-time rendering. This results in low processing efficiency, texture distortion, or rendering lag.

[0003] In existing technologies, CAD models are typically preprocessed using manual simplification or general polygon reduction tools. However, this approach often lacks precision control over critical engineering dimensions, easily leading to model distortion. Furthermore, it fails to effectively address issues such as mesh distortion, overlapping surfaces, and normal errors caused by differences in modeling logic between CAD and CG software, severely impacting the baking quality of subsequent critical textures like Normal and Curvature. In addition, traditional material creation relies on manually drawing textures, a tedious and time-consuming process that struggles to realistically reproduce the complex surface conditions such as rust, wear, and mineral dust accumulation that occur during long-term operation. This results in a significant gap in appearance similarity between digital models and real equipment, typically falling short of 90%, thus limiting its application in high-fidelity simulation scenarios.

[0004] Meanwhile, most current equipment simulation systems remain at the stage of static display or simple animation playback, lacking the ability to dynamically link with real-world environmental data and equipment operating status. These systems typically cannot access multi-source environmental data such as light intensity, atmospheric particulate matter concentration, and humidity, nor can they integrate IoT data such as real-time equipment location, operating status, and fault warnings. This results in virtual models failing to accurately reflect the actual operating conditions of the equipment, making it difficult to achieve synchronized digital twin functionality. Furthermore, in large-scale equipment scenarios, resource loading efficiency is low, and there is a lack of intelligent resource scheduling mechanisms based on the user's perspective, making it difficult to achieve a smooth interactive experience with low latency and high frame rates. Summary of the Invention

[0005] To address at least one of the technical problems existing in the background art, this application provides a device digital simulation method based on the PBR process, which achieves full-process optimization of high-precision digital simulation of equipment through multi-software collaboration.

[0006] The technical solution adopted in this application is as follows: This application provides a device digital simulation method based on the PBR process, including: S1. Use industrial modeling software SOLIDWORKS to digitally model the product based on its actual dimensions: measure the actual dimensions of each component of the equipment in detail, and build a parametric surface model of the equipment based on the actual dimensions; S2. Export the surface model created in SOLIDWORKS to 3ds Max: Import the surface model created in SOLIDWORKS into 3ds Max in STEP file format; S3. Optimize the surface model in 3ds Max: Using the imported surface model as a reference, optimize the product model to obtain an optimized model, and then perform UV unwrapping on the optimized model; S4. Import the optimized model into Substance Painter software for texture baking: In Substance Painter software, bake seven textures for the reconstructed optimized model: Normal, Worldspace Normal, ID, AmbientOccusion, Curvature, Position, and Thickness. S5. AI-based detail enhancement of textures in Substance Painter software: Using GAN algorithm with real device images as training data, detailed textures of the device are generated in Substance Painter software and fused into the texture; S6. Interactive rendering and digital twin creation of device models based on Unity: Import the device model with completed texture design in Substance Painter into the Unity engine, perform physical rendering based on the device's environment to obtain a digital twin model, and connect the digital twin model to the device management platform through the network communication module.

[0007] According to one embodiment of this application, step S1 includes the following steps: S11. Define the key dimensional parameters of each component of the equipment using SOLIDWORKS software, determine the geometric constraint relationships, and obtain a parametric model that can be dynamically adjusted. S12. By establishing the assembly constraint relationships between the various components of the equipment, a parametric surface model is obtained.

[0008] According to one embodiment of this application, step S2 includes the following steps: S21. Using a topology optimization tool, analyze the imported surface model, identify and remove redundant faces and edges, and obtain a simplified model with a simplified geometric structure.

[0009] According to one embodiment of this application, step S3 includes the following steps: S31. The surface model is automatically split based on the UV splitting function of 3ds Max to generate the initial UV layout; S32. Manually adjust the UVs of a portion of the surface model.

[0010] According to one embodiment of this application, S3 further includes: S33. Using the sculpting tools in 3ds Max, a layered sculpting method is adopted, combined with the actual working conditions of the equipment, to sculpt micro-features on the surface of the curved model.

[0011] According to one embodiment of this application, step S4 includes the following steps: S41. The basic material parameters are obtained by generating the device material using the material generation technology of Substance Painter software; S42. Based on the basic material parameters and combined with the material properties, create a material effect that matches the actual appearance of the device; S43. Based on adaptive sampling technology, the sampling rate is adjusted according to the curvature and detail of the model surface.

[0012] According to one embodiment of this application, S42 further includes: For equipment made of different materials, different protective features are added to the basic material parameters: For metal parts, add an anti-rust coating; If the component is made of rubber, then set anti-aging parameters.

[0013] According to one embodiment of this application, step S5 includes the following steps: S51. A training dataset for the device is obtained by training high-definition operational images of the device using a generative adversarial network algorithm; S52. Based on the AI ​​image generation function, learn and analyze the device training dataset to generate detailed textures of the device and integrate them into the texture of the device model; S53. By setting parameters such as detail intensity, distribution range, and wear type, a personalized device model texture design is obtained, and different types of detail texture maps are adjusted and blended.

[0014] According to one embodiment of this application, step S6 includes the following steps: S61. Import the completed texture design model into the Unity engine and implement physically based rendering based on HDRP and Shader; S62. By accessing multi-source environmental data, device IoT data, and user interaction commands, a multimodal data fusion algorithm is written using C# scripts; S63. Based on the Timeline tool and animation system, generate animation sequences of device motion; perform scene rendering based on the Post-Processing Stack; S64. Implement progressive rendering by using the Addressable Assets system to manage resources; Multi-source environmental data must include at least one of the following: light intensity distribution, atmospheric particulate matter concentration, and humidity gradient change; Device IoT data includes at least one of the following: device real-time location, operating status, or fault warning signals.

[0015] According to one embodiment of this application, S64 further includes: When using the Addressable Assets system to manage resources and implement progressive rendering, model resources within the user's view are loaded first, with a loading response time of less than 0.5 seconds.

[0016] Beneficial effects: The technical solution involved in this application achieves end-to-end optimization of high-precision digital simulation of equipment through multi-software collaboration and technological innovation. First, parametric and precise modeling of the equipment is completed using SOLID WORKS. After determining the equipment's dimensional parameters and assembly relationships, the model is exported in a specific format. Next, structural optimization, UV unwrapping, and detailed sculpting are performed in 3ds Max. Subsequently, intelligent material generation and high-precision texture baking are implemented in Substance Painter, and AI is used to enhance texture details. The processed model is then imported into the Unity engine, a rendering environment is built, and multi-source data is integrated. Simultaneously, Unity's features are used to create animations and optimize interactions, and progressive rendering is achieved through dynamic resource management. Finally, the digital twin model is integrated into an intelligent management platform, supporting full lifecycle management of the equipment and providing technical support for equipment intelligence.

[0017] The technical solution involved in this application, from SOLIDWORKS parametric modeling to Substance Painter's AI detail enhancement, and then to Unity's interactive rendering and digital twin construction, improves the similarity between the surface model and the actual device's appearance and reduces critical dimensional errors. Furthermore, through the dynamic access of multi-source environmental data and device IoT data, it accurately simulates the surface features and material changes of the device under complex working conditions such as high humidity, high dust, and strong impact, significantly improving the model's realism and detail. Ultimately, this technical solution effectively reduces equipment operation and maintenance costs, improves overall work efficiency, and provides an efficient and accurate digital solution for the entire lifecycle management of equipment, realizing intelligent applications for equipment design optimization, remote monitoring, and maintenance management. Attached Figure Description

[0018] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1 A flowchart illustrating the device digital simulation method based on the PBR process provided in this application embodiment. Detailed Implementation

[0019] To more clearly illustrate the overall concept of this application, a detailed explanation is provided below with reference to the accompanying drawings.

[0020] Many specific details are set forth in the following description to provide a thorough understanding of this application. However, this application may also be implemented in other ways different from those described herein. Therefore, the scope of protection of this application is not limited to the specific embodiments disclosed below. It should be noted that, unless otherwise specified, the embodiments of this application and the features thereof can be combined with each other.

[0021] In this application, unless otherwise expressly specified and limited, the "above" or "below" of the second feature can mean that the first and second features are in direct contact, or that the first and second features are in indirect contact through an intermediate medium. In the description of this specification, references to terms such as "an embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described can be combined in any suitable manner in one or more embodiments or examples.

[0022] like Figure 1As shown, this application provides a device digital simulation method based on the PBR process, including: Step S1. Use industrial modeling software SOLIDWORKS to digitally model the product according to its actual dimensions: measure the actual dimensions of each component of the equipment in detail, and build a parametric surface model of the equipment based on the actual dimensions; Step S2. Export the surface model created in SOLIDWORKS to 3ds Max: Import the surface model created in SOLIDWORKS into 3ds Max in STEP file format; Step S3. Optimize the surface model in 3ds Max software: Using the imported surface model as a reference, optimize the product model to obtain an optimized model, and then perform UV unwrapping on the optimized model; Step S4. Import the optimized model into Substance Painter software for texture baking: In Substance Painter software, bake seven textures for the reconstructed optimized model: Normal, Worldspace Normal, ID, AmbientOccusion, Curvature, Position, and Thickness. Step S5. Enhance the details of the texture in Substance Painter software based on AI: Use the GAN algorithm with real device images as training data to generate detailed textures of the device in Substance Painter software and fuse them into the texture. Step S6. Perform interactive rendering and digital twin creation of the device model based on Unity: Import the device model with texture design completed in Substance Painter into the Unity engine, perform physical rendering based on multi-source environmental data to obtain a digital twin model, and connect the digital twin model to the device management platform through the network communication module.

[0023] In step S1, in order to achieve high-precision geometric modeling of the equipment, it is necessary to conduct comprehensive and detailed actual size measurements of each component of the equipment using professional measuring tools. This provides an accurate geometric basis for subsequent digital simulation, ensuring that the size of the modeled equipment is consistent with the actual equipment size and meets the accuracy requirements of the equipment engineering.

[0024] Specifically, using professional measuring tools, comprehensive and detailed actual dimensional measurements are taken of each component of the equipment (such as tunneling machines and electric shovels) to obtain accurate dimensional data. The measurement data is then input into SOLIDWORKS software, and its parametric modeling function is used to construct a parametric surface model based on the data, thus initially reconstructing the geometry of the equipment.

[0025] In step S2, since 3ds Max can perform structural optimization on the model, by exporting the surface model created in SOLIDWORKS to the 3ds Max software, the model efficiency is improved, providing an optimized model foundation for UV processing and texture baking.

[0026] Specifically, the surface model created in SOLIDWORKS is imported into 3ds Max in STEP file format, and then analyzed using 3ds Max's topology optimization tool. The STEP format surface model effectively preserves the geometric and topological information of the model, reducing data exchange loss.

[0027] In step S3, the surface model is optimized using 3ds Max software. The optimized model structure is simpler, and UV unwrapping ensures a reasonable layout, enhances the detail of the model, and makes the model closer to the appearance of the actual device.

[0028] Specifically, 3ds Max uses its topology optimization tool to analyze the imported surface model, identify and remove redundant faces and edges, simplify the model's geometry, reduce model complexity, and improve subsequent processing efficiency without affecting model accuracy; based on 3ds Max's UV unwrapping function, the optimized surface model is automatically unwrapped to generate an initial UV layout.

[0029] By simplifying the surface model, the complexity of the device surface model can be reduced, improving the efficiency of subsequent processing. At the same time, it provides a more efficient model foundation for UV unwrapping and detail sculpting. The optimized model is more suitable for subsequent texture baking and rendering while ensuring appearance and accuracy. In addition, the simplification of the surface model reduces the computational burden of subsequent software processing and improves the overall work efficiency. UV unwrapping ensures the reasonable unfolding of texture maps and avoids texture stretching and deformation.

[0030] In step S4, seven texture maps—Normal, WorldspaceNormal, ID, AmbientOccusion, Curvature, Position, and Thickness—are baked on the optimized model in Substance Painter. This provides realistic surface details for the device's curved model and offers high-quality base textures for AI detail enhancement. Furthermore, these seven texture maps comprehensively capture the model's geometric details and material properties, providing a foundation for high-quality rendering. In step S5, in order to enhance the detail of the surface model, especially the surface features of the equipment under complex working conditions, and to make the surface model more realistic and closer to the actual state of the equipment in operation, the texture is enhanced with AI based on Substance Painter software.

[0031] Specifically, high-definition images of real equipment at the work site are collected as training data, covering the appearance of the equipment under different working conditions and at different stages of use. The GAN algorithm is used to learn and analyze the training data to generate highly realistic detailed textures. Combined with Substance Painter's layer blending modes and masking function, different types of detailed textures are finely adjusted and blended.

[0032] By using GAN algorithms with real device images as training data, highly realistic detailed textures can be generated, significantly improving the realism and detail of the model, making the model more similar to the appearance of the actual device; effectively solving the problem of simulating the dynamic changes of the device surface under real and complex working conditions.

[0033] In step S6, interactive rendering and digital twinning of the device model are performed using Unity, realizing interactive rendering of all elements of the device, constructing a digital twin model, ensuring the full lifecycle management of the device, and providing technical support for the intelligentization of the device.

[0034] Specifically, by importing the completed texture design model into the Unity engine, enabling HDRP, and combining it with a custom shader to achieve physically based rendering, the system integrates multi-source environmental data (light intensity distribution, atmospheric particulate matter concentration, humidity gradient changes) and equipment IoT data (real-time location, operating status, fault warning signals). A multimodal data fusion algorithm is written in C# to preprocess, extract features, and perform fusion calculations on various types of data. Animation sequences of equipment movement are generated using the Timeline tool and animation system. Resources are managed using the Addressable Assets system to achieve progressive rendering, prioritizing the loading of model resources within the user's viewpoint. The digital twin model is then connected to the intelligent management platform for mining equipment via a network communication module.

[0035] Furthermore, HDRP and custom shaders enable physically based rendering, accurately simulating the optical properties of mining equipment materials; multi-source data access allows rendering to perceive the environment and equipment status, achieving dynamic rendering and meeting the industry's demand for high-precision digital modeling of equipment; digital twin models are integrated into intelligent management platforms, supporting full lifecycle applications such as design optimization, remote monitoring, and maintenance management, effectively reducing equipment operation and maintenance costs by 40% and improving overall work efficiency by 80%, providing a complete and efficient solution for equipment intelligence.

[0036] The technical solution involved in this application achieves end-to-end optimization of high-precision digital simulation of equipment through multi-software collaboration and technological innovation. First, parametric and precise modeling of the equipment is completed using SOLID WORKS. After determining the equipment's dimensional parameters and assembly relationships, the model is exported in a specific format. Next, structural optimization, UV unwrapping, and detailed sculpting are performed in 3ds Max. Subsequently, intelligent material generation and high-precision texture baking are implemented in Substance Painter, and AI is used to enhance texture details. The processed model is then imported into the Unity engine, a rendering environment is built, and multi-source data is integrated. Simultaneously, Unity's features are used to create animations and optimize interactions, and progressive rendering is achieved through dynamic resource management. Finally, the digital twin model is integrated into an intelligent management platform, supporting full lifecycle management of the equipment and providing technical support for equipment intelligence.

[0037] The technical solution involved in this application, from SOLIDWORKS parametric modeling to Substance Painter's AI detail enhancement, and then to Unity's interactive rendering and digital twin construction, improves the similarity between the surface model and the actual device's appearance and reduces critical dimensional errors. Furthermore, through the dynamic access of multi-source environmental data and device IoT data, it accurately simulates the surface features and material changes of the device under complex working conditions such as high humidity, high dust, and strong impact, significantly improving the model's realism and detail. Ultimately, this technical solution effectively reduces equipment operation and maintenance costs, improves overall work efficiency, and provides an efficient and accurate digital solution for the entire lifecycle management of equipment, realizing intelligent applications for equipment design optimization, remote monitoring, and maintenance management.

[0038] In one embodiment of this application, step S1 includes the following steps: Step S11. Define the key dimensional parameters of each component of the equipment using SOLIDWORKS software, and determine the geometric constraint relationships to obtain a parametric model that can be dynamically adjusted. Step S12. By establishing the assembly constraint relationships between the various components of the equipment, a parametric surface model is obtained.

[0039] In step S11, by defining the key dimensional parameters of each component of the equipment and establishing the geometric constraint relationship between each component, it is ensured that all relevant dimensions can be automatically updated when a certain parameter of the equipment is modified, so as to achieve efficient iteration and precise control of the design and meet the strict requirements of the equipment for dimensional accuracy.

[0040] Specifically, in SOLIDWORKS, editable parameter variables are set for the key dimensions of each component of the equipment (such as the telescopic length of the cutting arm, the track track pitch, the diameter of the drive wheel, etc.). By labeling key dimensions and naming variables, mathematical relationships between dimensions are established (such as relating gear module and number of teeth, pitch circle diameter, etc. through equations). Geometric constraints (including horizontal, vertical, tangent, concentric, etc.) are applied to ensure the precise positional relationships of sketch elements. The sketch is constructed according to the principle of "positioning → shaping" to ensure that the sketch is fully defined (without blue unconstrained lines), achieving a fully constrained state of the sketch. Basic features (such as extruded bosses / bases, revolved features) are generated and detailed features (such as fillets, chamfers, arrays) are added to form a complete parametric model. It should be noted that in this application, parametric modeling ensures that the error of key dimensions is controlled within 0.5mm.

[0041] The embodiments of this application, through parametric design, ensure that critical dimensional errors are controlled within 0.5mm, meeting the stringent precision requirements of mining equipment. By establishing relationships between dimensions, a highly efficient design mode of "modifying one parameter and automatically updating the entire model" is achieved, shortening design change time. Geometric constraints ensure the accuracy and consistency of the sketches, avoiding subsequent feature failures due to insufficient constraints, and providing a precise geometric basis for subsequent equipment assembly and simulation, ensuring that the model's dimensions match those of the actual equipment. Parametric models significantly improve design efficiency, reduce repetitive work, and allow designers to focus more on innovation and optimization rather than mechanical dimensional modifications.

[0042] In step S12, by establishing the assembly constraint relationship between the various components of the equipment, the mechanical assembly relationship between the various components of the equipment can be realistically reproduced, ensuring that the components maintain the correct relative position and motion relationship when the equipment changes dynamically, thus providing an accurate assembly basis for the digital simulation of the equipment.

[0043] Specifically, in the SOLIDWORKS assembly environment, the "Fixed" constraint is used to fix the position of the reference component in the assembly; the "Coincident" constraint is applied to align the axes of cylindrical (or hole) shapes to achieve axial alignment between components; the "Contact" constraint is used to establish the contact relationship between two planes or between a cylinder (or sphere) and a plane; the "Offset" constraint is used to set a specific distance between two components to restrict their relative positional relationship; advanced constraints such as "Contour Center Constraint" are used to ensure the precise alignment of the component contour centers; and the "Path Fit" constraint is used to realize the motion relationship of components along a specified path to simulate the actual motion state of the equipment.

[0044] By employing precise assembly constraints, the mechanical structure of the equipment can be realistically reproduced, ensuring that each component maintains the correct relative position during dynamic changes, thus providing an accurate foundation for subsequent equipment motion simulation. Furthermore, the parameterization of assembly constraints enables automatic adjustment of assembly relationships when a key dimension is modified, significantly reducing the workload of assembly modifications. This provides a clear and accurate assembly model for subsequent 3ds Max model optimization, avoiding model processing problems caused by incorrect assembly relationships. It effectively solves the challenge of digitally modeling complex assembly structures in mining equipment, ensuring consistency between the model and the actual equipment's assembly relationships, and laying a reliable foundation for equipment digital twins.

[0045] In one embodiment of this application, step S2 includes the following steps: Step S21. Analyze the imported surface model using a topology optimization tool, identify and remove redundant faces and edges, and obtain a simplified model with a simplified geometric structure.

[0046] To address the issue of excessive model complexity caused by differences in modeling logic during the conversion from industrial design software (SOLIDWORKS) to 3D graphics processing software (3ds Max), topological optimization of the model was performed. This significantly reduced the number of facets and data redundancy while preserving key geometric features and surface accuracy, providing a lightweight and structurally sound digital model foundation for subsequent UV unwrapping, texture baking, and real-time rendering.

[0047] Specifically, the high-precision surface model exported from SOLIDWORKS in STEP format is imported into the 3ds Max environment; the model is automatically analyzed using 3ds Max's built-in "ProOptimizer", "Quad Chamfer" or third-party plugins such as "Graphite ModelingTools"; by scanning the model for densely distributed triangular faces, overlapping vertices, tiny edges and other unnecessary geometric elements, redundant faces and edges that do not contribute significantly to visual representation but increase computational burden are identified.

[0048] Using an intelligent polygon reduction algorithm, while maintaining the accuracy of the model outline, curvature variation areas, and key structures (such as the cutting head tooth seat, track plate connection holes, etc.), high-density polygon areas are reconstructed into a simpler and more uniform quadrilateral topology. The optimized model is then checked for edge streamlines to ensure that it is suitable for subsequent sculpting, animation rigging, and UV unwrapping operations, avoiding the appearance of extreme points or narrow triangular faces. The final output is a "simplified model" that can easily reproduce the original device geometry and has an efficient topology, preparing it for material processing in Substance Painter.

[0049] This solution intelligently simplifies high-precision surface models exported from SOLIDWORKS using topology optimization tools. While maintaining the model's appearance and accuracy, it reduces the number of faces by 30%-50%, significantly improving the overall efficiency of UV unwrapping, texture baking, and rendering. Simultaneously, it effectively eliminates problems such as mesh distortion and overlapping faces caused by differences in modeling logic between CAD and CG software, ensuring the baking quality of seven textures, including Normal and Curvature. The optimized, regularized mesh structure not only reduces the GPU burden on real-time engines like Unity, creating conditions for smooth 60fps interaction and progressive rendering, but also enhances compatibility with AI-driven detail generation in Substance Painter, achieving synergistic optimization of industrial precision and graphics efficiency.

[0050] In one embodiment of this application, step S3 includes the following steps: Step S31. Automatically split the surface model based on the UV splitting function of 3ds Max to generate the initial UV layout; Step S32. Manually adjust the UVs of a portion of the surface model; Step S33. Using the sculpting tools in 3ds Max, employ a layered sculpting method, and combine this with the actual working conditions of the equipment, sculpt microscopic features on the surface of the curved model.

[0051] In step S31, the imported simplified model undergoes initial UV unwrapping using 3ds Max's built-in "Unwrap UVW" modifier and its automatic splitting functions such as "Quick Planar Mapping" and "Unwrap Mapping". Specifically, the system automatically identifies potential UV seam locations based on the model's geometric curvature, edge orientation, and facet distribution, dividing the complex surface into multiple flattenable UV islands. After generating the initial UV layout, the UV distribution density is viewed through the "UV Editor" to ensure that each part of the UV is distributed as evenly as possible within the 0-1 texture space; and overlapping or over-compressed areas generated by automatic splitting are initially marked, providing a basis for subsequent manual adjustments.

[0052] Automatic UV splitting provides an efficient initial texture layout scheme for complex mining equipment models, significantly shortening UV processing time; and the generated initial UV layout lays the foundation for subsequent manual fine-tuning, avoiding the high workload and low consistency problems caused by completely manual splitting.

[0053] In step S32, based on automatic UV splitting, the surface model at key parts of the device is manually finely adjusted to optimize UV ​​space utilization, ensure that high detail areas obtain sufficient texture resolution, and enhance the realism of the final material representation.

[0054] Specifically, for key visual areas of the equipment (such as the cutting head blade array, track plate connection structure, and cab observation window), manually adjust the size and arrangement of UV islands in the UV editor; use tools such as "relax," "flatten," and "scale" to optimize the shape of UV islands, reduce stretching deformation, and ensure uniform texture mapping; prioritize allocating more UV space to areas with high detail requirements, such as enlarging the UV islands in the cutting head area to give them more pixel resources at the same texture resolution; adjust the position of UV seams to hide them in areas obscured by the equipment structure or in non-visually important areas to avoid exposing seam defects during rendering; ensure that all UV islands are strictly within the 0-1 texture coordinate range to avoid texture overflow or duplication.

[0055] Manual adjustments ensure higher texture density in key areas, making subsequent AI-generated details such as wear and cracks clearer and more realistic. Optimized UV space utilization avoids wasting texture resources, improves overall mapping efficiency and quality, reduces texture stretching and deformation, ensures natural material appearance in curved transition areas, and increases the overall similarity of material appearance to over 95%. This provides precise spatial basis for subsequent layer mask control in Substance Painter, allowing effects such as rust and mineral dust to be accurately distributed in vulnerable areas of the device.

[0056] In step S33, while maintaining the macroscopic geometric accuracy of the model, the microscopic damage and surface features formed by the mining equipment in the real working environment are added using the sculpting tools of 3ds Max to enhance the visual realism and working condition reproduction of the curved surface model.

[0057] Specifically, use the "Sculpt" modifier in 3ds Max or combine it with ZBrush-style brush tools to sculpt microscopic geometric features on the model's surface. The process employs a "layered sculpting" approach: the first layer sculpts large-scale features (such as pits formed by rock impacts and weld undulations); the second layer sculpts medium-scale features (such as machining marks and bolt indentations); and the third layer sculpts fine features (such as scratches and micro-cracks). This sculpting process references real-world images of mining equipment in operation, simulating the physical damage to the equipment surface caused by conditions such as tunneling machines cutting rock and electric shovels excavating ore. Different sculpting intensities are applied to different components; for example, high-wear areas such as the cutting head and track plates have a high sculpting density, while areas such as support structures and protective covers remain relatively flat. After sculpting, the model's normal information is updated to ensure that subsequent texture baking accurately captures the newly added micro-geometric changes. By employing a layered sculpting method, the surface damage characteristics of mining equipment under complex working conditions can be systematically reproduced, filling the gap in the microscopic detail representation of traditional modeling. Furthermore, the microscopic geometric features provide a realistic geometric basis for the Normal and Curvature textures of PBR materials, significantly enhancing the material's light and shadow performance. This provides a reasonable adhesion basis for the AI ​​textures (such as rust spots and mineral dust accumulation) generated by the GAN algorithm in Substance Painter, making the AI ​​generation effect more in line with physical laws. It also improves the similarity between the digital model and the appearance of the real equipment, making the simulation results more credible and meeting the needs of high-precision digital simulation of mining equipment. By completing the sculpting in the 3ds Max stage, the limitations of relying solely on normal maps to forge details in Substance Painter are avoided, achieving collaborative optimization of geometry and materials.

[0058] In one embodiment of this application, step S4 includes the following steps: Step S41. Generate basic material parameters for the device material using the material generation technology of Substance Painter software; Step S42. Based on the basic material parameters and combined with the material properties, create a material effect that matches the actual appearance of the device; Step S43. Based on adaptive sampling technology, adjust the sampling rate according to the curvature and detail of the model surface.

[0059] In step S41, based on the actual material type used in the device, the software's built-in intelligent algorithm automatically generates basic material parameters that conform to physical properties, providing a scientific basis for constructing a realistic and credible digital model.

[0060] Specifically, based on the actual materials of each component of the equipment (such as high-strength alloy steel, wear-resistant rubber, anti-corrosion coating, etc.), the corresponding "smart material" or "material generator" function is called in Substance Painter. The software automatically matches and generates basic parameters that conform to the material type based on a pre-set physical material database, including core PBR attributes such as base color, metallicity, roughness, and normal intensity. The generated basic material parameters serve as the starting point for subsequent manual adjustments and AI enhancements, ensuring that the initial state of the material has physical accuracy.

[0061] In step S42, based on the intelligently generated basic materials, personalized adjustments are made in combination with the user's observation and understanding of the actual appearance of the equipment, so that the material representation of the digital model is highly consistent with the appearance of the real equipment under complex working conditions in the mine.

[0062] Specifically, based on the basic material parameters, users can manually adjust key attributes such as base color, roughness, and metallicity according to the actual characteristics of the device, such as color, gloss, and surface treatment process. Using Substance Painter's layer system, users can construct multi-layered material structures to simulate the composite state of the device surface, such as a metal base + oil stain layer + mineral dust cover. The spatial distribution of different material attributes can be controlled through masks, such as increasing roughness in weld areas and adding rust effects to bolt connections. The overall aging degree and dirt status of the materials can be adjusted according to the device's usage environment (high humidity, high dust, strong impact) to enhance scene reproduction.

[0063] In step S43, the allocation of computational resources during the texture baking process is optimized to improve baking efficiency while ensuring texture quality, ensuring sufficient sampling accuracy in high-detail areas, and avoiding waste of computational resources in flat areas.

[0064] Specifically, in Substance Painter's texture baking settings, enabling the "Adaptive Sampling" function allows the system to automatically analyze the geometric features of the model's surface, identifying high-curvature areas (such as pits, bumps, and sharp edges) and densely detailed areas (such as sculpted textures and threaded structures). For high-curvature and high-detail areas, the sampling rate is automatically increased, increasing the number of ray tracing iterations to ensure that textures such as Normal and Curvature accurately capture subtle geometric changes. For flat or low-detail areas, the sampling rate is reduced, decreasing computational load while maintaining basic quality. This dynamically balances rendering quality and baking time, generating high-quality Normal, Worldspace Normal, ID, Ambient Occlusion, Curvature, Position, and Thickness textures.

[0065] This application utilizes the material generation technology of Substance Painter software to automatically construct basic material parameters that conform to physical properties. Combined with material attributes, it achieves high-fidelity reproduction of the material appearance of mining equipment under complex working conditions. Furthermore, adaptive sampling technology dynamically adjusts the baking sampling rate based on the surface curvature and detail density of the model, ensuring sufficient accuracy in high-detail areas while avoiding wasted computational resources in flat areas. This technical solution significantly improves the efficiency and scientific rigor of material generation, guaranteeing the realistic response of PBR materials under illumination and enhancing the engineering rationality of degradation effects such as rust, wear, and aging. It also improves the overall similarity of the material appearance of the digital model, providing a high-quality texture data foundation for subsequent AI detail enhancement and real-time rendering, effectively supporting the realization of high-precision digital simulation of mining equipment.

[0066] In one embodiment of this application, step S42 further includes: For equipment made of different materials, different protective features are added to the basic material parameters: For metal parts, add an anti-rust coating; If the component is made of rubber, then set anti-aging parameters.

[0067] To address the performance degradation characteristics of different material components in mining equipment under actual working conditions, protective attributes are added in a differentiated manner to realistically reproduce the physical properties and appearance evolution of metal and rubber materials in complex environments, thereby improving the engineering accuracy and visual realism of digital models in terms of material representation.

[0068] Specifically, in Substance Painter, based on the basic material parameters generated in step S41, the components of the equipment are classified according to their actual material types. For metal components (such as cutting arms, rotary tables, track frames, etc.), a "rust-proof coating" simulation layer is added to the material layer, setting up a multi-layer structure to reflect the anti-corrosion system of primer, intermediate coating, and topcoat, and configuring corresponding adhesion decay and oxidation diffusion parameters. For rubber components (such as sealing rings, shock-absorbing pads, track plate inserts, etc.), "anti-aging parameters" are set, and the aging process under long-term exposure to ultraviolet rays, oil stains, and mechanical stress is simulated by adjusting the material's elastic decay, surface crack generation rate, discoloration coefficient, and other properties.

[0069] The strength and distribution of protective features are differentiated based on the actual location of the components on the equipment and the usage environment. For example, the coating loss rate of exposed metal surfaces is higher than that of internal structures, and the aging parameters of rubber components in high-temperature areas are higher. These protective features are integrated into the material system in a non-destructive layer manner, supporting subsequent dynamic adjustments and visual previews. This enables accurate simulation of the service behavior of components made of different materials in harsh mining environments, overcoming the shortcomings of the "one-size-fits-all" approach in traditional material modeling. Simultaneously, adding anti-rust coatings allows metal components to realistically exhibit typical damage morphologies such as coating peeling, edge corrosion, and pitting in the rendering, significantly improving the material's operational condition accuracy. Setting anti-aging parameters allows rubber components to accurately represent aging characteristics such as hardening, cracking, and discoloration, enhancing the model's credibility in long-term simulations.

[0070] In one embodiment of this application, step S5 includes the following steps: Step S51. Train the high-definition operation images of the device using a generative adversarial network algorithm to obtain the device training dataset; Step S52. Based on the AI ​​image generation function, learn and analyze the device training dataset to generate device detail textures and integrate them into the texture map of the device model; Step S53. By setting parameters such as detail intensity, distribution range, and wear type, a personalized device model texture design is obtained, and different types of detail texture maps are adjusted and blended.

[0071] In step S51, a high-quality image dataset with the appearance features of real mining equipment is constructed to provide learning samples that conform to actual working conditions for subsequent AI-driven detailed texture generation, ensuring that the generated results have physical authenticity and scene relevance.

[0072] Specifically, high-resolution images of mining equipment such as tunneling machines, electric shovels, and mining trucks at different operational stages (newly manufactured, mid-term use, and heavily worn) are collected from the mining site, covering surface conditions under multiple angles and lighting conditions. The images are preprocessed, including noise reduction, color correction, and size normalization, and key areas (such as cutting heads, tracks, and articulated parts) and damage types (rust, scratches, and dust accumulation) are labeled. Unsupervised learning is performed on the preprocessed image set using WGAN or StyleGAN architectures in generative adversarial networks, enabling the network to extract and learn the distribution patterns of microscopic features on the equipment surface. During training, a discriminator continuously provides feedback on the differences between generated and real images, optimizing the generator parameters until the generator can stably output texture samples that conform to the statistical characteristics of the equipment's appearance. Finally, a structured and scalable "equipment training dataset" is output, containing the original image features and the latent feature space learned by the GAN.

[0073] In step S52: High-fidelity micro-detail textures (such as rust spots, scratches, oil stains, mineral dust, etc.) are automatically extracted and generated from the training dataset using a deep learning model, and seamlessly integrated into the PBR mapping system of the equipment model, significantly improving the visual realism and working condition reproduction of the digital model.

[0074] Specifically, the equipment training dataset generated in step S51 is input into an AI image generation model (such as CycleGAN or Pix2Pix) for end-to-end learning and inference. Based on the semantic information of the input equipment region (such as "cutting head holder" and "track plate connection hole"), the model automatically generates detailed texture maps for the corresponding locations, including channels such as Normal, Roughness, Metallic, and Ambient Occlusion. The generated detailed textures are then multi-channel fused with the base map baked in step S43, using layer overlay, mask control, and intensity adjustment to achieve a natural transition. During the fusion process, the physical consistency of the PBR material is maintained to ensure that the roughness changes and normal bumps generated by the AI ​​exhibit realistic physical responses under illumination. The generated results can be visualized and manually intervened to ensure that the texture distribution conforms to engineering common sense and actual observation.

[0075] In step S53, by providing a user-controllable parameterized adjustment mechanism, fine-grained control of AI-generated detailed textures is achieved, meeting the personalized needs of different device states (new machines, old machines, faulty machines) and different application scenarios (simulation, training, demonstration).

[0076] Specifically, Substance Painter or its customized interface provides a visual parameter adjustment panel, allowing users to set key parameters such as "detail intensity," "distribution density," and "wear type" (e.g., uniform wear, localized peeling, edge corrosion). Users can adjust the coverage area and depth of the rust layer, or control the thickness and distribution area of ​​mineral dust accumulation, based on the actual service life or maintenance status of the equipment. It supports independent control and layered fusion of different types of detail textures (e.g., mechanical damage, chemical corrosion, environmental deposition), achieving multi-factor coupling effects. A non-destructive editing workflow is adopted, with all adjustments saved as layers and masks, supporting modification at any time and version rollback. Furthermore, the parameter adjustment results are fed back to the preview window in real time, allowing users to intuitively evaluate the final effect.

[0077] In one embodiment of this application, step S6 includes the following steps: Step S61. Import the model with completed texture design into the Unity engine and implement physically based rendering based on HDRP and Shader; Step S62. By accessing multi-source environmental data, device IoT data, and user interaction commands, a multimodal data fusion algorithm is written using C# script; Step S63. Generate an animation sequence of device motion based on the Timeline tool and animation system; perform scene rendering based on the Post-Processing Stack; Step S64. Use the Addressable Assets system to manage resources and implement progressive rendering; Multi-source environmental data must include at least one of the following: light intensity distribution, atmospheric particulate matter concentration, and humidity gradient change; Device IoT data includes at least one of the following: device real-time location, operating status, or fault warning signals.

[0078] Step S61: By utilizing the Unity engine's High Dynamic Range Rendering Pipeline (HDRP) and advanced shader technology, high-quality, realistic rendering of the mining equipment model with complex texture details is achieved to achieve visual realism.

[0079] Specifically, the 3D model with texture design completed in the previous steps will be exported to a Unity-supported format and imported into a Unity project; using the functions provided by Unity HDRP, the lighting settings in the scene will be configured, including light source type, intensity, color, etc., to simulate the lighting conditions in a real mining environment; custom or preset shaders will be applied, which can accurately simulate the performance of materials such as metal, rust, and mineral dust under different lighting conditions to ensure that the generated image conforms to physical laws; material parameters, such as roughness and metallicity, will be adjusted so that the model surface exhibits accurate optical properties during rendering.

[0080] In step S62, by integrating data streams from different sources, including environmental data, device operating status information, and user input, and by writing C# scripts, real-time data processing and fusion are achieved, thereby enhancing the interactivity and dynamic response capabilities of the virtual scene.

[0081] Specifically, the system receives and parses multi-source environmental data from sensor networks, weather stations, and other data sources based on development interfaces or existing APIs, and establishes a data collection mechanism for IoT devices. For example, it obtains real-time location, operating status, and fault warning signals from edge devices via the MQTT protocol. C# scripts are written to adjust elements in the virtual scene based on the received data, such as changing the light intensity to reflect the sun's angle at the current time, or updating the device status display. Combined with user interaction commands, such as mouse clicks and keyboard input, it enables the manipulation and control of objects within the scene.

[0082] In step S63, the system creates a dynamic animation sequence to demonstrate the device's operation process and uses a post-processing stack to optimize the final output image quality, thereby improving the overall visual experience.

[0083] Specifically, Unity's Timeline tool is used to orchestrate the sequence of actions for device components, including specific actions during startup, operation, and shutdown phases. Complex mechanical motion animations are created using the Animator or Animation window to ensure smooth and natural animation. The Post-Processing Stack is used to add effects such as Bloom, Color Grading, and Depth of Field to enhance realism and immersion. Unity's Addressable Assets system is employed for resource management, combining cloud servers and edge computing nodes to achieve dynamic loading and unloading of model resources. When users interact with VR / AR devices, high-detail model resources for critical areas are prioritized based on actual device operation data, enabling progressive rendering and improving the smoothness of the interactive experience.

[0084] In step S64, progressive rendering is achieved by using the Addressable Assets system to manage resources, thereby optimizing resource management and loading performance for large projects and ensuring that high-quality visual effects are maintained while reducing initial loading time and memory usage.

[0085] Specifically, Unity's Addressable Asset System is used to tag and organize various resources in a project, including models, textures, and audio files. It supports full lifecycle applications such as design optimization (e.g., virtual prototype testing, structural strength simulation), remote monitoring (e.g., real-time status visualization, fault prediction and early warning), and maintenance management (e.g., maintenance path planning, spare parts inventory optimization), providing strong support for intelligent management of mining equipment.

[0086] This application imports the completed texture design model into the Unity engine and uses the HDRP high-fidelity rendering pipeline and customized shaders to achieve physically based rendering, ensuring that the material representation of mining equipment under complex lighting conditions is realistic and believable. By accessing multi-source environmental data (such as light intensity distribution, atmospheric particulate matter concentration, and humidity gradient changes) and equipment IoT data (such as real-time location, operating status, and fault warning signals), combined with user interaction commands, a multi-modal data fusion algorithm is built using C# scripts to achieve dynamic synchronization and intelligent response between the virtual model and the real working conditions. Based on the Timeline and animation system, a precise sequence of equipment motion animations is generated, and post-processing such as depth of field, color correction, and floodlighting are performed using the Post-Processing Stack, significantly improving visual expressiveness and immersion. At the same time, the Addressable Assets system is used to manage resources in a modular manner, prioritizing the loading of model resources within the user's viewpoint to ensure that the loading response time is less than 0.5 seconds, effectively supporting progressive rendering of large-scale scenes. This technical solution not only achieves a smooth interactive experience at high frame rates (60fps), but also constructs a high-precision, highly interactive, and low-latency digital twin system for mining equipment through the collaborative optimization of physically realistic rendering, multi-source data-driven approaches, and efficient resource scheduling, providing solid technical support for equipment simulation, remote monitoring, and intelligent training.

[0087] In one embodiment of this application, step S64 further includes: When using the Addressable Assets system to manage resources and implement progressive rendering, model resources within the user's view are loaded first, with a loading response time of less than 0.5 seconds.

[0088] By using the Addressable Assets system for resource management, prioritizing the loading of model resources within the user's current field of view ensures the immediate presentation of key visual content, achieving low-latency resource response under an immersive interactive experience and meeting the needs of high-real-time industrial simulation applications.

[0089] Specifically, the Unity Addressables system is used to group and manage mining equipment models by spatial region or functional module, dividing model resources into multiple independently loadable Addressable units. At runtime, the camera's frustum culling detects the model area covered by the current viewpoint, dynamically identifying "resources within the view area" that require priority loading. Combined with resource preloading and caching strategies, areas that may enter the viewpoint are predicted before the user's viewpoint changes, and low-priority resources are loaded in advance to ensure the main process loading efficiency. Performance monitoring tools optimize the loading process, ensuring that the end-to-end response time from issuing a loading request to resource visibility is controlled within 0.5 seconds.

[0090] This solution effectively reduces redundant resource loading, lowers memory usage and bandwidth consumption, and improves overall system stability through a space-aware loading strategy.

[0091] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

[0092] The above description is merely an embodiment of this application and is not intended to limit this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A device digital simulation method based on PBR process, characterized in that, include: S1. Use industrial modeling software SOLIDWORKS to digitally model the product based on its actual dimensions: measure the actual dimensions of each component of the equipment in detail, and build a parametric surface model of the equipment based on the actual dimensions; S2. Export the surface model created in SOLIDWORKS to 3ds Max: Import the surface model created in SOLIDWORKS into 3ds Max in STEP file format; S3. Optimize the surface model in 3ds Max: Using the imported surface model as a reference, optimize the product model to obtain an optimized model, and then perform UV unwrapping on the optimized model; S4. Import the optimized model into Substance Painter software for texture baking: In Substance Painter software, bake seven textures for the reconstructed optimized model: Normal, Worldspace Normal, ID, AmbientOccusion, Curvature, Position, and Thickness. S5. AI-based detail enhancement of textures in Substance Painter software: Using GAN algorithm with real device images as training data, detailed textures of the device are generated in Substance Painter software and fused into the texture; S6. Interactive rendering and digital twin creation of device models based on Unity: Import the device model with completed texture design in Substance Painter into the Unity engine, perform physical rendering based on the device's environment to obtain a digital twin model, and connect the digital twin model to the device management platform through the network communication module.

2. The equipment digital simulation method based on PBR process according to claim 1, characterized in that, S1 includes the following steps: S11. Define the key dimensional parameters of each component of the equipment using SOLIDWORKS software, determine the geometric constraint relationships, and obtain a parametric model that can be dynamically adjusted. S12. By establishing the assembly constraint relationships between the various components of the equipment, a parametric surface model is obtained.

3. The equipment digital simulation method based on PBR process according to claim 1, characterized in that, S2 includes the following steps: S21. Using a topology optimization tool, analyze the imported surface model, identify and remove redundant faces and edges, and obtain a simplified model with a simplified geometric structure.

4. The equipment digital simulation method based on the PBR process according to claim 1, characterized in that, S3 includes the following steps: S31. The surface model is automatically split based on the UV splitting function of 3ds Max to generate the initial UV layout; S32. Manually adjust the UVs of a portion of the surface model.

5. The equipment digital simulation method based on the PBR process according to claim 4, characterized in that, S3 further includes: S33. Using the sculpting tools in 3ds Max, a layered sculpting method is adopted, combined with the actual working conditions of the equipment, to sculpt micro-features on the surface of the curved model.

6. The equipment digital simulation method based on the PBR process according to claim 1, characterized in that, S4 includes the following steps: S41. The basic material parameters are obtained by generating the device material using the material generation technology of Substance Painter software; S42. Based on the basic material parameters and combined with the material properties, create a material effect that matches the actual appearance of the device; S43. Based on adaptive sampling technology, the sampling rate is adjusted according to the curvature and detail of the model surface.

7. The equipment digital simulation method based on the PBR process according to claim 6, characterized in that, S42 further includes: For equipment made of different materials, different protective features are added to the basic material parameters: For metal parts, add an anti-rust coating; If the component is made of rubber, then set anti-aging parameters.

8. The equipment digital simulation method based on PBR process according to claim 1, characterized in that, S5 includes the following steps: S51. A training dataset for the device is obtained by training high-definition operational images of the device using a generative adversarial network algorithm; S52. Based on the AI ​​image generation function, learn and analyze the device training dataset to generate detailed textures of the device and integrate them into the texture of the device model; S53. By setting parameters such as detail intensity, distribution range, and wear type, a personalized device model texture design is obtained, and different types of detail texture maps are adjusted and blended.

9. The equipment digital simulation method based on PBR process according to claim 1, characterized in that, S6 includes the following steps: S61. Import the completed texture design model into the Unity engine and implement physically based rendering based on HDRP and Shader; S62. By accessing multi-source environmental data, device IoT data, and user interaction commands, a multimodal data fusion algorithm is written using C# scripts; S63. Based on the Timeline tool and animation system, generate animation sequences of device motion; perform scene rendering based on the Post-Processing Stack; S64. Implement progressive rendering by using the Addressable Assets system to manage resources; Multi-source environmental data must include at least one of the following: light intensity distribution, atmospheric particulate matter concentration, and humidity gradient change; Device IoT data includes at least one of the following: device real-time location, operating status, or fault warning signals.

10. The equipment digital simulation method based on the PBR process according to claim 9, characterized in that, S64 further includes: When using the Addressable Assets system to manage resources and implement progressive rendering, model resources within the user's view are loaded first, with a loading response time of less than 0.5 seconds.

Images