Simulated production method and production system for outer coating layer of sofa

Abstract

The invention discloses a simulated production method for an outer coating layer of a sofa. The simulated production method comprises the following steps: S1, constructing and maintaining a decorative layer material and effect database; the database comprises a material basic parameter library for storing fabric physical and purchasing parameters, a visual effect mapping library for storing material digital twin chartlets and optical characteristic parameters, and a historical project library for associatively storing historical project data; s2, receiving a sofa three-dimensional model and decorative layer style design information input by a user, calling corresponding material data in the visual effect mapping library, and performing real-time visual simulation through a physics-based rendering engine to generate a decorative layer appearance effect picture; s3, after the effect is confirmed, performing 3D-to-2D processing on the decoration layer of the sofa three-dimensional model, automatically unfolding and calculating the shape and area of the two-dimensional cutting piece of each decoration layer component; and S4, automatically segmenting the two-dimensional cut piece obtained in the step S3.

Description

A simulated production method and system for the outer cover layer of a sofa Technical Field

[0001] This invention relates to the field of furniture processing systems, and in particular to a method and system for simulating the production of the outer cover layer of a sofa. Background Technology

[0002] Sofas generally consist of a main frame and an outer cover that surrounds it. Current outer covers include leather, fabric, and some specialty fabrics. The aesthetic appeal of sofas largely relies on the designer's subjective design, therefore there is no unified standard.

[0003] The standards include design standards and production standards. Therefore, ensuring the standardization of both processes has always been a development trend in the furniture industry.

[0004] For example, Chinese patent CN202210171097.7 mainly focuses on the specific production of the outer layer, but it cannot achieve the visual effect of the pattern or the pattern cutting standard, so the production and processing are still limited. Summary of the Invention

[0005] The main objective of this invention is to propose a simulated production method for the outer cover layer of sofas, aiming to achieve simulated production of the sofa's outer cover layer, thereby standardizing visual design (and thus improving the trend of sales market and production, that is, increasing market demand and supply according to customer needs); through the standardized production of the outer cover layer, production costs are reduced, while the existing design and board making time cycle of more than 3 days is reduced, thus improving production efficiency.

[0006] To achieve the above objectives, this invention proposes a method for simulating the production of sofa outer covers, comprising the following steps: S1: Constructing and maintaining a database of decorative layer materials and effects; the database includes a material basic parameter library storing fabric physical and procurement parameters, a visual effect mapping library storing material digital twin textures and optical property parameters, and a historical project library associated with historical project data (i.e., different light and shadow effects of different decorative layer materials, and visual effects of different stitching lines, achieving a dual database to achieve better simulation design effects and reduce waste or redesign steps after physical production; S2: Receiving the user-input sofa 3D model and decorative layer style design information, and calling the... The corresponding material data in the visual effect mapping library is used for real-time visualization simulation through a physically based rendering engine to generate a rendering of the decorative layer's appearance. The 3D model can be a designer's 3D drawing or a 3D scan of the product, which is then used to produce 3D drawings. Existing data can be used for the 3D scan. S3: After confirming the effect, the decorative layer of the sofa's 3D model is converted from 3D to 2D, automatically unfolded, and the shape and area of ​​the 2D cut pieces for each decorative layer component are calculated. This flattens the decorative layer, enabling cutting or pre-production of the decorative layer. S4: The 2D cut pieces obtained in step S3 are automatically segmented to generate individual cut components, and then applied to the 2D cut pieces. Add pre-defined seams adapted to the fabric type (i.e., in the actual 3D to 2D conversion process, the curved surface is pre-stretched through the planar U-axis to achieve the shaping and seam stitching at the seam); S5: Perform automatic cutting optimization. Based on the accumulated data in the historical project database, the system automatically selects one or a combination of rule-driven mode or data-driven prediction mode to generate an optimized cutting scheme; (i.e., determined according to the different needs of the designer, and further modified according to the designer's specific judgment or needs before being input into the database, thereby improving the database's judgment. Of course, this modification can be used to create a designer database to differentiate between the standard database and the non-standard design database. Of course, the design...) There are multiple designers, each with a different IP address and a defined database to improve the universality and distinctiveness of the design. Additionally, the database can be configured to grant designer permissions or system permissions based on the needs of different designers, thus avoiding system confusion. Furthermore, the designer's database can serve as a backup solution to enhance design flexibility. S6: Output the final production data package, which includes the final rendering, an optimized digital cutting diagram with seam annotations, a material list, and a cost analysis report, used to drive the CNC cutting machine for cutting. S7: Store the entire process data of this task as a complete data record in the historical project database for continuous system optimization.

[0007] The advantages of this application are: ① It integrates the traditional discrete sofa design, prototyping, and production processes into a closed-loop automated system. Compared with existing designs (such as relying on designer experience for 2D drawing and manual layout), this application achieves a fundamental transformation through "3D model-driven" and "automatic data-driven decision-making".

[0008] ② Integration and Foresight: In existing technologies, 3D modeling, material simulation, and cutting and layout are often independent processes. This method seamlessly integrates the three, especially in the design stage where high-precision rendering achieves "what you see is what you get," greatly reducing the number of physical prototypes and lowering development costs and time.

[0009] ③ Adaptive optimization capability: An automatic mode switching (rule-driven → data-driven) based on historical data volume is introduced, enabling the system to learn and continuously optimize. Compared with the fixed and rigid material routing algorithms in existing technologies, this allows for the continuous accumulation of enterprise knowledge and the sustained improvement of material utilization.

[0010] ④ Knowledge Accumulation and Reuse: Data accumulation is clearly defined as a necessary step, transforming personal experience into reusable digital assets for enterprises, thus solving the problem of loss of technical experience caused by personnel turnover in traditional manufacturing. Attached Figure Description

[0011] Figure 1 is the overall flowchart; Figure 2 is a partial flowchart of S3; Figure 3 is a partial flowchart of S4. Detailed Implementation

[0013] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0014] It should be noted that if the embodiments of the present invention involve directional indicators (such as up, down, left, right, front, back, top, bottom, inside, outside, vertical, horizontal, longitudinal, counterclockwise, clockwise, circumferential, radial, axial, etc.), the directional indicators are only used to explain the relative positional relationship and movement of the components in a certain specific posture (as shown in the figure). If the specific posture changes, the directional indicators will also change accordingly.

[0015] Furthermore, if the embodiments of this invention involve descriptions such as "first" or "second," these descriptions are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, features defined with "first" or "second" may explicitly or implicitly include at least one of those features. Additionally, the technical solutions of the various embodiments can be combined with each other, but this must be based on the ability of those skilled in the art to implement them. If the combination of technical solutions is contradictory or impossible to implement, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed by this invention.

[0016] As shown in Figures 1 to 3, a method for simulating the production of an outer layer for a sofa includes the following steps: S1: Constructing and maintaining a database of decorative layer materials and effects; the database includes a material basic parameter library storing fabric physical and procurement parameters, a visual effect mapping library storing material digital twin textures and optical property parameters, and a historical project library associated with historical project data (i.e., different light and shadow effects of different decorative layer materials, and visual effects of different stitching, to achieve a dual database, thereby achieving better simulation design effects and reducing waste or redesign steps after physical production); S2: Receiving the user-input sofa 3D model and decorative layer style design information, and calling the visual effect mapping library... The corresponding material data in the visual effect mapping library is used for real-time visualization simulation through a physically based rendering engine to generate a rendering of the decorative layer's appearance. The 3D model can be a designer's 3D drawing or a 3D scan of the product, which is then used to produce 3D drawings. Existing data can be used for the 3D scan. S3: After confirming the effect, the decorative layer of the sofa's 3D model is converted from 3D to 2D, automatically unfolded, and the shape and area of ​​the 2D cut pieces for each decorative layer component are calculated. This flattens the decorative layer, enabling cutting or pre-production of the decorative layer. S4: The 2D cut pieces obtained in step S3 are automatically segmented to generate individual cut components, and these components are then applied to the 2D cut pieces. Add pre-defined seams that match the fabric type; that is, in the actual 3D to 2D conversion process, the curved surface is pre-stretched through the planar U-axis to achieve the shaping and seam stitching at the seam; S5: Perform automatic cutting optimization. The system automatically selects one or a combination of rule-driven mode or data-driven prediction mode based on the accumulated data in the historical project library to generate an optimized cutting scheme; that is, it is determined according to the different needs of the designer. Furthermore, it can be modified according to the designer's specific judgment or needs before being input into the database, thereby improving the database's judgment. Of course, this modification can be used to create a designer database to differentiate between the standard database and the non-standard design database. There are multiple databases, each defined with a different IP address, to improve the universality and distinctiveness of the design; additionally, they can be configured as either designer permissions or design system permissions according to the needs of different designers, thus avoiding confusion in the design system; another option is to use the designer's design database as a backup solution to improve design flexibility; S6: Output the final production data package, which includes the final rendering, the optimized digital cutting diagram with seam annotations, the bill of materials, and the cost analysis report, used to drive the CNC cutting machine for cutting; S7: Store the entire process data of this task as a complete data record in the historical project database for continuous system optimization.

[0017] The advantages of this application are: ① It integrates the traditional discrete sofa design, prototyping, and production processes into a closed-loop automated system. Compared with existing designs (such as relying on designer experience for 2D drawing and manual layout), this application achieves a fundamental transformation through "3D model-driven" and "automatic data-driven decision-making".

[0018] ② Integration and Foresight: In existing technologies, 3D modeling, material simulation, and cutting and layout are often independent processes. This method seamlessly integrates the three, especially in the design stage where high-precision rendering achieves "what you see is what you get," greatly reducing the number of physical prototypes and lowering development costs and time.

[0019] ③ Adaptive optimization capability: An automatic mode switching (rule-driven → data-driven) based on historical data volume is introduced, enabling the system to learn and continuously optimize. Compared with the fixed and rigid material routing algorithms in existing technologies, this allows for the continuous accumulation of enterprise knowledge and the sustained improvement of material utilization.

[0020] ④ Knowledge Accumulation and Reuse: Data accumulation is clearly defined as a necessary step, transforming personal experience into reusable digital assets for enterprises, thus solving the problem of loss of technical experience caused by personnel turnover in traditional manufacturing.

[0021] Specifically, the material basic parameter library includes: recording the type of leather, fabric and other materials (such as top-grain genuine leather, imitation leather, polyester material, cotton-polyester blend), density, thickness, abrasion resistance, color dry rubbing fastness (such as ≥4 grade), texture pattern, width, unit price, etc.

[0022] Visual Effects Mapping Library: Associates material parameters with the rendering engine. By collecting physical samples of materials and performing high-precision scanning, it generates digital twin textures, recording their visual characteristics such as gloss, softness, and wrinkle effects under different lighting conditions (e.g., natural light, warm light).

[0023] The "Design-Result-Cost" history records the design drawings, selected materials, actual output results (including client feedback), and final material consumption and cost data for each completed project.

[0024] Real-time rendering: Physically based rendering technology allows designers to "wrap" selected decorative layer materials onto 3D sofa models. The system can simulate the fit, drape, and seam effects of different materials on sofas of different styles (such as classical carvings and modern minimalist designs).

[0025] Parametric effect adjustment: Designers can adjust the stitch color, stitch length (e.g., 3-point thread for fabric, 4-point thread for leather), wrinkle style and other process details of the virtual fabric in real time, and see the effect change immediately.

[0026] Rule-based automatic nesting: The system automatically generates two-dimensional unfolded diagrams of each component of the decorative layer based on the design drawings. In the initial stage, it automatically nests the materials according to the material width and component shape, using classic algorithms (such as leftmost alignment and lowest horizontal line algorithms) to generate a preliminary cutting plan.

[0027] Data-driven predictive nesting: This function is activated once the system has accumulated enough historical data ("when a certain number of data records are reached"). Based on new design drawings and material selections, the system prioritizes matching the most similar successful cutting schemes from the historical database. Simultaneously, it analyzes the implicit relationship between material parameters (such as texture directionality and elasticity) and cutting utilization through machine learning models to predict the optimal nesting method, thereby maximizing material utilization and reducing waste.

[0028] Cost simulation and solution comparison: For the same design, the system can simulate cutting schemes and total costs using materials of different grades or prices (such as branded fabrics and ordinary fabrics), and generate comparison reports to assist in procurement and decision-making.

[0029] Seamlessly integrate the simulation system with the actual production process.

[0030] Design and production data are integrated: the optimized cutting scheme can directly generate digital cutting instructions, which are transmitted to the CNC cutting machine to realize "design as production" and reduce intermediate errors.

[0031] The key points are: 1. Time cost optimization, reducing the time cycle from design to production; 2. Design cost optimization, enabling diverse comparative designs through light and shadow judgment and a material basic parameter library; 3. Reduced material waste, effectively reducing material waste during plate making. One aspect is that after separating the individual cut parts, multiple cut parts can be integrated onto the same board before cutting; however, this design is mainly applied to designs that can be cut in a non-integral manner, such as separate armrest covering, separate seat covering, and separate backrest covering; of course, even for integrated cutting designs, the cut areas can be simulated for covering products such as cushions and armrests, further reducing material waste and achieving environmentally friendly utilization; 4. Through multi-dimensional light and shadow rendering, the judgment of market demand can be improved, thereby reducing the probability of unsold products.

[0032] Specifically, the 3D to 2D conversion process in step S3 adopts an equal area unfolding step (i.e., unfolding first and then compensating, mainly surface compensation). The difference between automated plate making and manual plate making is that manual plate making can achieve maximum utilization, but involves multiple designs (such as a higher defect rate); while automated plate making can achieve standard consistency (i.e., a larger material allowance, but can be adjusted according to different production or designer plans), that is, its production has linear variation.

[0033] Combining the energy optimization equation to minimize the global stretching error, a piecewise unfolding and splicing compensation strategy is adopted for non-developable surfaces to ensure that the error between the unfolded 2D contour area and the 3D decorative layer surface area does not exceed 3%. Specifically, S3 includes: S31: Decorative layer surface extraction; Separating the independent surfaces of the decorative layer from the overall 3D model of the sofa using a boundary segmentation algorithm to form a decorative layer mesh model: S = {M_i | i=1,2,...,n}; where S is the overall area; where (M_i) is a single triangular mesh face of the decorative layer; S32: Surface topology analysis; Establishing the adjacency relationship matrix (A_{n×n}) of each triangular face in the mesh model, (A_{ij}=1) indicates that the i-th triangular face is adjacent to the j-th triangular face, (A_{ij}=0) indicates that they are not adjacent; S33: Parametric mapping establishment, selecting the non-closed edge of the decorative layer as the unfolding baseline and mapping it to the 2D plane u-axis, defining the 2D parametric coordinates (u_k, v_k) of each vertex, satisfying the 3D coordinates of adjacent vertices. The stretching error between the distance and the 2D mapping distance is within the range of 1% to 3%. The 2D coordinates of the vertices are solved using the decoration layer optimization equation: E = sum_{(v_k,v_l)∈E} (frac{L_{kl_{2D}}}{L_{kl}} - 1)^2, where E is the set of edges in the mesh model, (L_{kl}) is the 3D distance between adjacent vertices, and (L_{kl_{2D}}) is the 2D mapping distance between adjacent vertices. S36: Error correction for non-developable surfaces: The non-developable region is divided into several sub-surfaces according to the curvature threshold. After each sub-surface is individually parameterized and unfolded, the splicing error between adjacent sub-surfaces is calculated and compensation is added. S35: 2D contour generation: A 2D triangular mesh is constructed, and an edge extraction algorithm is used to simplify the boundaries, generating a smooth 2D closed contour. The error between the contour area and the 3D decoration layer surface area does not exceed 3%.

[0034] This method solves the problem of accurately unfolding complex curved surfaces of sofas; existing simple unfolding methods lead to large deformations, affecting the fit of the finished product. This method, through isothermal parameterization and energy optimization, strictly controls the error to within 3%, ensuring high fidelity from digital model to physical pattern piece, and avoiding material waste and seam wrinkles caused by unfolding errors from the source.

[0035] Capability to handle complex curved surfaces: For non-developable curved surfaces such as sofa armrests and backrests, the innovative segmented unfolding + splicing compensation strategy provides an effective engineering solution, breaking through the limitations of existing CAD software in unfolding highly complex curved surfaces.

[0036] The pre-defined seams mentioned in step S4 include splicing seams, decorative seams, and quilting seams; when adding seams, the seam parameters are automatically matched according to the fabric type, and the depth compensation parameters of the seams are synchronized to the three-dimensional visualization simulation engine to simulate the three-dimensional light and shadow effect at the seams.

[0037] The specific process of automatic segmentation of the 2D cut pieces in step S4 includes: S41: Feature extraction: Identify structural mapping points, curvature change points, and texture alignment points from the 2D contour; S42: Segmentation line generation: Construct a candidate set of segmentation lines based on feature points, and use a dynamic programming algorithm to select the optimal combination of segmentation lines. The objective function is (F = α·N + β·sum_{i=1}^q frac{A_{part i}}{A_{2D}} + γ·frac{L_{total seam}}{A_{2D}}), where N is the number of parts after segmentation, (q=N), (α=-0.5), (β=0.3), (γ=-0.2); S43: Segmentation verification and adjustment: Verify whether the size and shape of each part after segmentation meet the cutting requirements. If not, adjust the position of the segmentation lines or add segmentation lines.

[0038] The process of adding the pre-defined seam in step S42 includes: S421: Seam position positioning: The splicing seam is aligned with the edge of the 2D component, and the distance from the edge is 5 times the fabric thickness; the decorative seam is generated by projection based on the three-dimensional texture features of the 3D sofa model; the quilting seam is generated as a closed seam 5mm inside the outline of the 2D component; S422: Seam parameter adaptation: The seam parameters are automatically matched according to the fabric type, or a user-defined seam style is used and the rationality of the parameters is verified; S423: Three-dimensional rendering association: The depth compensation parameters of the seam are synchronized to the three-dimensional visualization simulation engine to simulate the changes in light and shadow at the seam, and the seam position information is synchronized to the final production data package.

[0039] The "stitch" was elevated from a mere manufacturing instruction to an important design element, thus achieving unification between the design and manufacturing ends.

[0040] In existing technologies, stitching is represented only as lines on 2D drawings, making it impossible to preview the actual 3D effect. This method, by associating depth compensation parameters and a physically based rendering engine, can realistically simulate the texture of stitching during the design phase, greatly enriching design expressiveness and making visual evaluation more accurate.

[0041] Parametric and automated: The sewing parameters are automatically matched according to the fabric type, avoiding human error, ensuring the consistency between design intent and production process, and improving the efficiency and reliability of customized design.

[0042] 3D to 2D unfolding algorithm and calculation formula; for example, for the common "developable surface + locally non-developable surface" mixed features of sofa decorative layers, the unfolding method is adopted. The core is to map the 3D mesh to the 2D plane while keeping the local angle of the surface unchanged, and at the same time minimize the area error.

[0043] Select a non-closed edge of the decorative layer as the unfolding baseline L0 and map it onto the u-axis of the 2D plane.

[0044] The data-driven prediction mode described in step S5 includes: a. Feature extraction: extracting the feature vector of the current design task, including the shape complexity of each component, material texture directionality requirements, material elastic coefficient, total component area and aspect ratio; b. Scheme matching and prediction: matching the feature vector of the current task with cases in the historical project library for similarity, and / or using machine learning models to predict the material layout guide line; c. Multi-scheme optimization and comparison: generating at least one prediction-based optimization scheme and comparing it with the rule-driven benchmark scheme in terms of material utilization, estimated cost and time.

[0045] This upgrades the nesting scheme (i.e., the pattern-making step) from "experience-dependent" and "fixed rule" to "data-driven" and "predictively automated." By extracting multi-dimensional feature vectors, the system can gain a deeper understanding of the essence of the design task, rather than simply arranging geometric shapes. Machine learning is used for prediction (e.g., decision tree models, convolutional neural networks) to generate predetermined nesting schemes.

[0046] It also provides a multi-solution comparison function, which provides designers with quantitative basis, enabling them to make the best trade-off between multiple objectives such as cost, time, and material utilization, which is of great value in actual production.

[0047] An automated simulation production system for sofa upholstery layers includes: a database module for storing a database of upholstery layer materials and effects; a 3D simulation conversion module for real-time visualization simulation; a 3D-to-2D conversion and seam-separation module for 3D-to-2D processing, automatic segmentation of 2D cut pieces, and addition of predetermined seams; a cutting optimization module for automatic cutting optimization; a data output module for outputting production data packages; and a management and update module for managing project processes and controlling data accumulation and model update mechanisms.

[0048] This claim is a system claim corresponding to the method, seeking to protect the hardware and software architecture for implementing the innovative method.

[0049] System-level innovation: The value of this system lies in integrating disparate tools into a collaborative, organic whole. Each module performs its own function while working closely together (e.g., the database module provides data support for rendering and optimization, and the learning module ensures system evolution), forming a complete "design-simulation-production-optimization" ecosystem, whose overall efficiency is greater than the sum of its individual functions.

[0050] Enhancing industrial competitiveness: The existence of this system enables enterprises to respond quickly to market changes, conduct flexible production of small batches and multiple varieties, improve the advanced manufacturing capabilities of the traditional furniture industry, and thus effectively reduce production costs, thereby improving market competitiveness while ensuring high-quality products; the key points are: 1. Time cost; 2. Design cost; 3. Material waste; 4. Through multi-dimensional light and shadow rendering, the judgment of market demand can be improved, thereby reducing the probability of unsold products.

[0051] The database module also includes a suture parameter library, which stores parameter information for different types of sutures and is associated with the visual effect mapping library to support visual preview of suture effects.

[0052] By librarying the suture parameters and linking them with other databases, the design elements (sutures) have been standardized and made manageable. This not only ensures design consistency but also makes sutures a callable and previewable asset, much like materials, greatly improving design efficiency and professionalism.

[0053] The cropping optimization module also includes a manual intervention optimization module. This module compares the physical production and virtual generation of the data, determines whether the data meets the requirements based on the differences, and stores the optimized data in the system improvement module or directly stores the data that meets the requirements. This enables data updating and judgment. The continuous optimization can also use existing learning models, such as decision tree models, to continuously select the best solution; for example, convolutional neural networks (CNNs) are good at processing data with spatial or temporal structure, such as images and speech. They extract local features through convolutional layers and reduce dimensionality through pooling layers, thereby improving the conversion efficiency of 3D to 2D and optimizing edge data, thus improving the finished product effect.

[0054] Capability to handle complex curved surfaces: For non-developable curved surfaces such as sofa armrests and backrests, the innovative segmented unfolding + splicing compensation strategy provides an effective engineering solution, breaking through the limitations of existing CAD software in unfolding highly complex curved surfaces.

[0055] Specifically, one design implementation example is as follows: (I) 3D Decorative Layer Preprocessing and Surface Extraction: From the overall 3D model of the sofa, the independent surfaces of the decorative layer (leather layer) are separated using a boundary segmentation algorithm. The triangular mesh faces of the sofa 3D model are traversed, and mesh units belonging to the decorative layer are selected (determined by material ID, layer attributes, or geometric relationships) to form an independent decorative layer mesh model S={Mi|i=1,2,...,n}, where Mi is a single triangular mesh face of the decorative layer.

[0056] Topological relationship analysis of surfaces: An adjacency matrix An×n is established for each triangular face in the mesh model, where Aij=1 indicates that the i-th triangular face is adjacent to the j-th triangular face, and Aij=0 indicates that they are not adjacent. This adjacency matrix ensures the continuity of surfaces during unfolding, avoiding tearing or overlapping.

[0057] Key parameter definition: Mesh vertex coordinates: Let the vertex set of the decorative layer mesh model be V={vk(xk,yk,zk)|k=1,2,...,m}, where (xk, yk, zk) are 3D coordinates. Spatial coordinates; Triangular face side lengths: For a triangular face Mi, the three side lengths are: Lab=(xa−xb)2+(ya−yb)2+(za−zb)2; Lbc=(xb−xc)2+(yb−yc)2+(zb−zc)2; Lca=(xc−xa)2+(yc−ya)2+(zc−za)2; Surface curvature: Calculate the Gaussian curvature K and mean curvature H of each vertex to determine the surface type (developable / non-developable surface): Gaussian curvature K=R1R21 (where R1 is the radius of curvature of the two principal curvatures at the vertex); mean curvature H=21(R11+R21); When K=0, it is a developable surface (such as a cylinder or cone), which can be directly developed; when K=0, it is a non-developable surface (such as a sphere or parabola), which requires an approximate development method.

[0058] (II) 3D to 2D unfolding algorithm and calculation formula For the common mixed features of "developable surface + local non-developable surface" in sofa decorative layer, the "isothermal parametric unfolding method" is adopted. The core is to map the 3D mesh to the 2D plane while keeping the local angle of the surface unchanged, and at the same time minimize the area error.

[0059] Parametric mapping is established by selecting a non-closed edge of the decorative layer as the unfolding baseline L0 and mapping it onto the u-axis of the 2D plane. The 3D length of the baseline L03D = ∑seg∈L0Lseg; (Lseg is the segment length of the baseline). After mapping, the 2D length L02D = L03D (ensuring that the baseline is not stretched).

[0060] For each vertex vk, define its parametric coordinates (uk, vk) in the 2D plane, satisfying: for adjacent vertices vk and vl, the 3D distance Lkl = (xk−xl)2+(yk−yl)2+(zk−zl)2; the 2D mapping distance Lkl2D = (uk−ul)2+(vk−vl)2; the constraint condition is: LklLkl2D = 1±ε (ε is the allowable stretching error, with a value range of 0.01~0.03, i.e. 1%~3%, which is in line with the elasticity tolerance range of sofa fabric).

[0061] Establish an energy optimization equation to minimize the global stretching error: E=∑(vk,vl)∈E(LklLkl2D−1)2 where E is the set of edges of the mesh model. Solve for (uk,vk) using the gradient descent method to obtain the 2D coordinates of all vertices.

[0062] Error correction for non-developable surfaces: For local non-developable regions with Gaussian curvature K=0 (such as the curved surface of sofa armrests and the curvature of backrests), a "segmented unfolding + splicing compensation" strategy is adopted: the non-developable region is divided into several sub-surfaces S1, S2, ..., Sp according to the curvature threshold Kth (empirical value Kth=0.001mm−2), and the absolute value of Gaussian curvature of each sub-surface |K|≤Kth, which is approximately a developable surface; the parametric unfolding of step 1 is performed separately for each sub-surface to obtain the 2D contour of the sub-surface; the splicing error of adjacent sub-surfaces is calculated, and a compensation amount Δ / 2 is added to the splicing edge of the 2D contour to ensure that the spliced ​​surface fits the 3D shape.

[0063] 2D contour generation: Traverse the 2D parameterized vertex set and construct a 2D triangular mesh according to the triangular face connection relationship of the original 3D mesh; use an edge extraction algorithm (such as the Douglas-Puk algorithm) to simplify the boundary of the 2D mesh, remove redundant vertices, and generate a smooth 2D closed contour.

[0064] The area of ​​the outline, A2D = 21 |∑i=1n(uivi+1−ui+1vi) |(un+1,vn+1) = (u1,v1), must satisfy |A3DA2D−A3D| ≤ ε (ε = 0.03) with the surface area of ​​the 3D decorative layer, A3D = ∑i=1nAMi (AMi is the area of ​​the 3D triangular face).

[0065] II. Implementation Scheme for Automatic Segmentation of 2D Decorative Layer; The core of 2D decorative layer segmentation is to automatically divide the complete 2D outline into multiple independent cutting components (such as backrest pieces, armrest pieces, seat cushion pieces, etc.) based on factors such as sofa structural characteristics, fabric texture direction, and cutting efficiency. The segmentation process follows the principles of "structure priority, texture adaptation, material saving and high efficiency".

[0066] (a) Segmentation rule definition: Structural feature driven rule: Extract the structural boundary information of the 3D sofa model (such as the connection between the backrest and the armrest, and the boundary line between the seat cushion and the backrest), and map it onto the 2D outline to form the initial segmentation line; Each segmented component must correspond to an independent functional area of ​​the 3D model, and the edge of the component must coincide with the fold and seam position of the sofa structure to ensure that the splicing conforms to the 3D form.

[0067] Fabric matching rules: For fabrics with a texture direction (such as leather texture, striped fabric), the dividing line must be parallel or perpendicular to the texture direction to avoid texture confusion after splicing; the component size must match the material width (call the width data W from the material basic parameter library), and the maximum side length of each component ≤ W−2Δ (Δ is the cutting allowance, default 20mm).

[0068] Material-saving optimization rules: The number of components after division should be minimized to reduce the number of seams and material waste; the shape of the components should be as close as possible to rectangles or regular polygons to facilitate subsequent material layout optimization (improving material utilization).

[0069] (ii) Automatic segmentation process feature extraction: Identify key feature points from 2D contours, including: structural mapping points: projection points of 3D structural boundary lines onto 2D contours; curvature change points: vertices of curvature change on 2D contours (such as corners, arc changes); texture alignment points: alignment reference points determined according to the fabric texture direction (called from the visual effects mapping library).

[0070] Segmentation line generation: A candidate set of segmentation lines is constructed based on feature points. Each candidate segmentation line must meet the following requirements: it does not cross the functional area of ​​the component, it is adapted to the texture direction, and the size of the segmented component meets the width requirement. A dynamic programming algorithm is used to select the optimal combination of segmentation lines. The objective function is: total seam length of the component, where: N is the number of components after segmentation, q=N, α=−0.5 (minimize the number of components), β=0.3 (maximize the area ratio of a single component), γ=−0.2 (minimize the total length of the seam). Optimal segmentation is achieved through weight balancing.

[0071] Segmentation Verification and Adjustment: Verify whether the size and shape of each segmented component meet the cutting requirements (e.g., the minimum component area is ≥0.1㎡ to avoid wasting materials on overly small components); if there are components that do not meet the requirements, automatically adjust the position of the segmentation lines or add segmentation lines until all components meet the rules.

[0072] III. Pre-selected Seam Addition and 3D Enhancement Solution Seams are not only the physical connection between decorative layers, but also a key element in enhancing the three-dimensionality of the design. By precisely adding pre-selected seams to the 2D decorative layer, the assembled 3D decorative layer can exhibit three-dimensional effects such as raised textures and layered structures after cutting and production.

[0073] (I) Sewing Type and Parameter Definition The system database calls the preset sewing parameter library, which includes the following core parameters: Sewing Type | Function | Thread Width (d) | Spacing (s) | Depth Compensation (h) | Applicable Area | Joint | Component Connection | 3~5mm | -1~2mm | Component Edge | Decorative Sewing | Three-Dimensional Shaping | 2~3mm | 10~20mm | 0.5~1mm | Backrest and Seat Cushion Surface | Seating Sewing | Reinforcement Contour | 2mm | 5~8mm | 0.3~0.5mm | Armrest Edge and Backrest Contour Note: Depth compensation h refers to the amount of fabric compression at the seam, used to simulate a three-dimensional concave effect in 3D rendering, which is then reflected by the physical rendering engine.

[0074] (II) Automatic Seam Addition Process and Seam Positioning: Splicing Seam: Automatically fits the edge contour of the 2D component, with a distance of t = 10-15mm from the edge (adjusted according to fabric thickness, calling the fabric thickness δ from the material base parameter library, t = 5δ); Decorative Seam: Based on the three-dimensional texture features of the 3D sofa model (such as the concave-convex shape of the backrest), the 3D texture is projected onto the 2D component to form the seam path. The projection formula is: Let point P(x,y,z) on the 3D texture line have a projection point P′(u,v) on the 2D component that satisfies the parameterized mapping relationship (consistent with the mapping matrix during 3D to 2D conversion); Seam Stitching: Generates a closed seam 5mm inside the contour line of the 2D component, reinforcing the three-dimensional contour of the component edge.

[0075] Sewing parameter adaptation: Automatically match sewing parameters according to fabric type (e.g., use 3mm thread width for leather and 2mm thread width for fabric); if the user customizes the sewing style (input through style design information), the user parameters will be used first, while verifying the rationality of the parameters (e.g., the thread width should not exceed twice the fabric thickness to avoid sewing breakage).

[0076] 3D rendering linkage: While adding stitches to the 2D decorative layer, the depth compensation parameter h of the stitches is recorded; the stitch information is synchronized to the 3D visualization simulation engine, and during real-time rendering, the light and shadow changes at the stitches are simulated through physical rendering algorithms (shadows are formed in recessed areas, and highlights are formed in raised areas), so that the 3D effect brought by the stitches can be previewed during the design stage; the stitch position information is synchronized to the final production data package, clearly marking the coordinates, length, and parameters of the stitches, guiding the CNC cutting machine to reserve stitch holes or indentations, and ensuring that the 3D effect after production meets the design expectations.

[0077] IV. Integration and Data Flow with Existing Systems: The 3D to 2D conversion, segmentation, and seam addition modules serve as preprocessing units before the "Automatic Cutting Optimization Engine Module." The specific integration process is as follows: After the 3D visualization simulation engine confirms the effect, it outputs the 3D model of the decorative layer to this module; after completing the 3D to 2D conversion, automatic segmentation, and seam addition, this module outputs a set of 2D cut pieces with seam annotations, which is then fed into the automatic cutting optimization engine for nesting optimization; finally, a "seam parameter list" is added to the final production data package, output synchronously with the cutting diagram and material list, driving CNC equipment production.

[0078] Database association: During the 3D to 2D conversion process, parameters such as fabric thickness and elasticity coefficient from the material basic parameter library are called for error correction and segmentation rule adaptation; the seam parameter library is associated with the visual effect mapping library, and the seam styles, colors, and other effects corresponding to different fabrics can be previewed through the visualization engine; the 2D conversion error data, segmentation scheme, seam parameters, etc. of this task are stored in the historical project library as part of the data accumulation, and are used for subsequent optimization of the error threshold and segmentation rules of the 3D to 2D algorithm.

[0079] In practical implementations, fully automated computation results in high system maintenance costs. Therefore, a partial approach with some designer decisions is generally adopted to reduce the system failure rate. Meanwhile, 3D to 2D conversion is a relatively existing design. The key point is the compensation scheme for the seam position, which mainly focuses on (material stretching and invisible surface compensation).

[0080] The above description is merely a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural transformations made using the contents of the present invention's specification and drawings under the inventive concept of the present invention, or direct / indirect applications in other related technical fields, are included within the patent protection scope of the present invention.

Claims

1. A method for simulating the production of an outer cover layer for a sofa, characterized in that: Includes the following steps: S1: Construct and maintain a database of decorative layer materials and effects; the database includes a material basic parameter library storing fabric physical and procurement parameters, a visual effect mapping library storing material digital twin textures and optical property parameters, and a historical project library that stores historical project data; S2: Receive the user-input sofa 3D model and decorative layer style design information, call the corresponding material data in the visual effect mapping library, and perform real-time visualization simulation through a physically based rendering engine to generate a decorative layer appearance effect image; S3: After confirming the effect, the decorative layer of the sofa 3D model is converted from 3D to 2D, automatically unfolded and the shape and area of ​​the 2D cut pieces of each decorative layer component are calculated. S4: Automatically segment the two-dimensional cutting piece obtained in step S3 to generate individual cutting components, and add predetermined seams adapted to the fabric type to the two-dimensional cutting piece; S5: Perform automatic cutting optimization. Based on the data accumulation in the historical project library, the system automatically selects one or a combination of rule-driven mode or data-driven prediction mode to generate an optimized cutting scheme; S6: Output the final production data package, which includes the final effect diagram, the optimized digital cutting diagram with seam annotations, a bill of materials, and a cost analysis report, used to drive the CNC cutting bed for cutting; S7: Store the entire process data of this task as a complete data record in the historical project library for continuous system optimization.

2. The method for simulating the production of the outer cover layer of a sofa as described in claim 1, characterized in that, The 3D to 2D conversion process described in step S3 adopts an equal area unfolding step, combines energy optimization equations to minimize global stretching error, and adopts a segmented unfolding and splicing compensation strategy for non-developable surfaces to ensure that the error between the unfolded 2D contour area and the 3D decorative layer surface area does not exceed 3%.

3. The method for simulating the production of the outer cover layer of a sofa as described in claim 2, characterized in that, Specifically, S3 includes: S31: Surface extraction of the decorative layer; Independent surfaces of the decorative layer are separated from the overall 3D model of the sofa using a boundary segmentation algorithm, forming a decorative layer mesh model: S = {M_i | i=1,2,...,n}; where S is the overall area; and (M_i) is a single triangular mesh face of the decorative layer; S32: Surface topology analysis; An adjacency matrix (A_{n×n}) is established for each triangular face in the mesh model, where (A_{ij}=1) indicates that the i-th triangular face is adjacent to the j-th triangular face, and (A_{ij}=0) indicates that they are not adjacent; S33: Parametric mapping establishment; Non-closed edges of the decorative layer are selected as the expansion baseline and mapped to the 2D plane u-axis. The 2D parametric coordinates (u_k, v_k) of each vertex are defined, satisfying that the stretching error between the 3D distance and the 2D mapping distance between adjacent vertices is within the range of 1% to 3%. This is achieved through the decorative layer optimization equation: E = sum_{(v_k,v_l)∈E} (frac{L_{kl_{2D}}}{L_{kl}} - 1)^2; Solve for the 2D coordinates of the vertices, where E is the set of edges of the mesh model, (L_{kl}) is the 3D distance between adjacent vertices, and (L_{kl_{2D}}) is the 2D mapping distance between adjacent vertices; S36: Error correction for non-developable surfaces: Divide the non-developable region into several sub-surfaces according to the curvature threshold. After each sub-surface is parametrically unfolded, calculate the splicing error between adjacent sub-surfaces and add compensation; S35: 2D contour generation: Construct a 2D triangular mesh, use an edge extraction algorithm to simplify the boundary, and generate a smooth 2D closed contour. The error between the contour area and the surface area of ​​the 3D decorative layer does not exceed 3%.

4. The method for simulating the production of the outer cover layer of a sofa as described in claim 1, characterized in that, The pre-defined seams mentioned in step S4 include splicing seams, decorative seams, and quilting seams; when adding seams, the seam parameters are automatically matched according to the fabric type, and the depth compensation parameters of the seams are synchronized to the three-dimensional visualization simulation engine to simulate the three-dimensional light and shadow effect at the seams.

5. The method for simulating the production of the outer cover layer of a sofa as described in claim 1, characterized in that, The data-driven prediction mode described in step S5 includes: a. Feature extraction: extracting the feature vector of the current design task, including the shape complexity of each component, material texture directionality requirements, material elastic coefficient, total component area and aspect ratio; b. Scheme matching and prediction: matching the feature vector of the current task with cases in the historical project library for similarity, and / or using machine learning models to predict the material layout guide line; c. Multi-scheme optimization and comparison: generating at least one prediction-based optimization scheme and comparing it with the rule-driven benchmark scheme in terms of material utilization, estimated cost and time.

6. An automated simulation production system for sofa decorative layers, characterized in that, The system includes a production system database module employing any one of claims 1-5 for simulating the production of the outer cover layer of a sofa, the database module being used to implement the function of the decorative layer material and effect database; a three-dimensional simulation conversion module, the three-dimensional simulation conversion module being used to implement real-time visualization simulation function; a 3D to 2D conversion and seam segmentation module, the 3D to 2D conversion and seam segmentation module being used to implement 3D to 2D processing, automatic segmentation of two-dimensional cut pieces, and the addition of predetermined seams; and a cutting optimization module, the cutting optimization module being used to implement automatic cutting optimization function. Data output module, the data output module is used to output production data packets; The management update module is used to manage project processes and control data accumulation and model update mechanisms.

7. The system as described in claim 5, characterized in that, The database module also includes a suture parameter library, which stores parameter information for different types of sutures and is associated with the visual effect mapping library to support visual preview of suture effects.

8. The system as described in claim 5, characterized in that, The cutting optimization module also includes a manual intervention optimization module. The manual intervention optimization module compares the production of physical objects with virtual generation, and makes a judgment on whether the data meets the requirements based on the data differences. The optimized data is then stored in the system for improvement, or the data that meets the requirements is stored directly.

Images