Footwear style design method based on digital three-dimensional model

Abstract

The invention relates to a footwear style design method based on a digital three-dimensional model. The method comprises the following steps: acquiring three-dimensional point cloud data of the protective shoes, and extracting distribution characteristics to construct a three-dimensional grid model after noise reduction processing; dividing the model into a sole area, a vamp area and an anti-smashing head area, and obtaining a closed boundary point sequence of each area; extracting a sole thickness distribution value, an anti-smashing head curvature value and a vamp wrinkle value based on the boundary point sequence, and generating a feature vector data set; constructing a finite element mechanical simulation model by using the feature vector data set, and generating a protection performance quantitative relation data set; and a multi-objective optimization model is established in combination with the quantitative relation data set, and an optimal structure parameter combination meeting preset safety standard constraint conditions is generated. By means of the method, the efficiency and precision of protective shoe design can be improved, a data-driven standardized process is provided for protective shoe structure design, it is effectively guaranteed that products meet the safety standard, and the comprehensive performance is optimized.

Description

Technical Field

[0001] The present invention belongs to the field of computer graphics, and particularly relates to a method for designing shoe styles based on a digital three-dimensional model. Background Art

[0002] With the development of the field of computer graphics, there has emerged a technology for designing shoe styles based on a digital three-dimensional model. Traditional style design methods rely on manual surveying and mapping, physical modeling, and empirical trial and error, and have defects such as a long design cycle, high cost, insufficient accuracy, and difficulty in balancing functionality and comfort. With the development of digital technology, technologies such as three-dimensional scanning, computer-aided design (CAD), and finite element analysis (FEA) have gradually been applied to product research and development. However, there is still a lack of a systematic solution for accurately modeling the complex curved surface structures of shoes (such as the curvature of the anti-smashing head, the thickness distribution of the sole, and the pleat degree of the upper), and for optimizing the performance under the coupling action of multiple physical fields (mechanical impact, electric field shielding, plantar pressure distribution). In the prior art, the noise interference in point cloud data processing, the ambiguity of the regional boundary in mesh model segmentation, and the technical bottleneck in the balance between safety standards and ergonomics in multi-objective optimization restrict the in-depth application of digital design in the shoe industry. Summary of the Invention

[0003] Based on this, it is necessary to provide a method for designing shoe styles based on a digital three-dimensional model that can improve the design efficiency and comprehensive performance of shoe products and achieve digital precision modeling of protective shoes for the above technical problems.

[0004] In a first aspect, the present application provides a method for designing shoe styles based on a digital three-dimensional model, including: Obtaining three-dimensional point cloud data of a protective shoe, performing noise reduction processing on the three-dimensional point cloud data, and extracting distribution features to construct a three-dimensional mesh model.

[0005] Dividing the three-dimensional mesh model into a sole region, an upper region, and an anti-smashing head region to obtain a closed boundary point sequence for each region.

[0006] Extracting the sole thickness distribution value, the anti-smashing head curvature value, and the upper pleat degree value according to the closed boundary point sequence to generate a feature vector data set.

[0007] Constructing a finite element mechanics simulation model based on the feature vector data set to generate a quantitative relationship data set of protective performance.

[0008] Establishing a multi-objective optimization model according to the quantitative relationship data set to generate an optimal structural parameter combination that satisfies the preset safety standard constraint conditions.

[0009] In one embodiment, obtaining three-dimensional point cloud data of a protective shoe, performing noise reduction processing on the three-dimensional point cloud data, and extracting distribution features to construct a three-dimensional mesh model includes: Obtain the three-dimensional point cloud data of the protective shoes; the three-dimensional point cloud data includes point cloud density information.

[0010] Determine the noise reduction parameter threshold for controlling the neighborhood search radius of the point cloud according to the point cloud density information.

[0011] Perform outlier filtering on the three-dimensional point cloud data based on the noise reduction parameter threshold to obtain the noise-reduced three-dimensional point cloud data.

[0012] Extract the surface normal vector distribution characteristics of the noise-reduced three-dimensional point cloud data; the surface normal vector distribution characteristics include curvature continuity information.

[0013] Generate a triangular patch topological structure according to the curvature continuity information, and perform Poisson reconstruction on the noise-reduced three-dimensional point cloud data to obtain an initial three-dimensional mesh model.

[0014] Calculate the vertex curvature gradient of the initial three-dimensional mesh model and perform subdivision optimization on the initial three-dimensional mesh model to generate the final three-dimensional mesh model.

[0015] In one embodiment, the vertex curvature gradient is calculated by the following formula: ; Where ) represents the vertex curvature gradient of the th mesh, represents the vertex in the th mesh, represents the set of first-order neighborhood vertices of vertex , represents the number of neighborhood vertices, and represent the principal curvature values of vertex and vertex , represents the unit direction vector from vertex to vertex , represents the weighting coefficient, represents the scaling parameter for controlling the influence of curvature difference.

[0016] In one embodiment, divide the three-dimensional mesh model into a sole area, an upper area, and an anti-smashing head area to obtain the closed boundary point sequences of each area, including: Obtain the vertex coordinate data and triangular patch topological data of the three-dimensional mesh model.

[0017] Calculate the curvature characteristic parameters of each vertex according to the vertex coordinate data to obtain a curvature distribution map.

[0018] Extract high-curvature regions based on the curvature distribution map, and combine the regional connectivity constraints to merge adjacent triangular patches to generate initial segmentation regions.

[0019] Perform dilation and erosion operations on the boundaries of the initial segmentation regions according to the preset segmentation threshold to determine the set of boundary points for the sole region, the upper region, and the anti-smashing head region.

[0020] Perform topological relationship analysis on the set of boundary points, remove isolated points, and then match with the triangular patch topological data to obtain the closed boundary point sequence for each region.

[0021] In one embodiment, extract the sole thickness distribution value, the anti-smashing head curvature value, and the upper wrinkle degree value according to the closed boundary point sequence to generate a feature vector data set, including: Extract the boundary point curvature continuity parameter and the thickness gradient change rate according to the closed boundary point sequence.

[0022] Calculate the change rate of the anti-smashing head curvature radius according to the curvature continuity parameter to generate a curvature radius distribution matrix.

[0023] Combine the thickness gradient change rate and the curvature radius distribution matrix to generate a three-dimensional feature mapping table.

[0024] Extract the wrinkle region density distribution parameter from the three-dimensional feature mapping table, and use the principal component analysis method to perform dimensionality reduction processing on the parameter to obtain a low-dimensional feature vector set.

[0025] Construct a feature vector similarity matrix according to the low-dimensional feature vector set; the similarity matrix is used to judge the correlation between the upper wrinkle degree and the anti-smashing head curvature.

[0026] If the correlation exceeds the preset threshold, update the curvature radius distribution weight coefficient in the three-dimensional feature mapping table.

[0027] Adjust the dimensionality distribution of the low-dimensional feature vector set according to the updated weight coefficient, and output the optimized feature vector data set.

[0028] In one embodiment, construct a finite element mechanical simulation model based on the feature vector data set to generate a quantitative relationship data set for the protection performance, including: Construct a finite element mesh model according to the material properties and geometric parameters in the feature vector data set; the finite element mesh model contains node coordinates and element topological relationships.

[0029] Call the mechanical solver to calculate the stress distribution data of the finite element mesh model.

[0030] Synchronously call the electric field solver to calculate the electric potential gradient distribution of the finite element mesh model.

[0031] Couple the stress distribution data with the electric potential gradient distribution for analysis, and extract the anti-smashing performance index value; the anti-smashing performance index value includes the impact resistance coefficient and the electric field shielding rate.

[0032] Generate a dataset of quantitative relationships of protection performance based on the anti-smashing performance index value and geometric parameters.

[0033] In one embodiment, establish a multi-objective optimization model according to the dataset of quantitative relationships, and generate an optimal combination of structural parameters that meet the preset safety standard constraint conditions, including: Extract the plantar pressure distribution matrix and the shear force gradient matrix from the dataset of quantitative relationships; the pressure distribution matrix contains the pressure intensity data of each region; the shear force gradient matrix contains the dynamic shear change amount.

[0034] Calculate the safety evaluation function value by weighted fusion of the pressure distribution matrix and the shear force gradient matrix.

[0035] Judge whether the safety evaluation function value exceeds the preset threshold. If it does not exceed, dynamically adjust the structural parameter step size based on the gradient descent algorithm.

[0036] Update the structural parameter combination based on the structural parameter step size to generate a new sequence of parameter combinations.

[0037] Input the sequence of parameter combinations into the multi-objective optimization model to obtain a set of objective function values.

[0038] Screen the optimal parameter identification numbers that meet the constraints from the set of objective function values to obtain the optimal combination of structural parameters; the optimal parameter identification numbers correspond to the double-objective balance point of pressure and shear force.

[0039] In one embodiment, the safety evaluation function value is calculated by the following formula: ; Among them, represents the safety evaluation function value, represents the plantar pressure distribution matrix, represents the pressure intensity value in the th row and th column of the pressure distribution matrix, represents the shear force gradient matrix, represents the dynamic shear change amount in the th row and th column of the shear force gradient matrix, represents the regional weight coefficient, and represent the weighting coefficients, represents the non-linear mapping function of pressure and shear force.

[0040] In a second aspect, the present application further provides a computer device, including a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, the foregoing method is implemented.

[0041] In a third aspect, the present application further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the foregoing method is implemented.

[0042] The foregoing method for designing a footwear style based on a digital three-dimensional model, computer device, and storage medium acquire three-dimensional point cloud data of a protective shoe, extract distribution features after noise reduction processing to construct a three-dimensional mesh model; divide the model into a sole, a shoe upper, and an anti-smashing head area, and acquire a sequence of closed boundary points for each area; extract the sole thickness distribution value, the anti-smashing head curvature value, and the shoe upper fold degree value based on the boundary point sequence to generate a feature vector data set; use the feature vector data set to construct a finite element mechanics simulation model to generate a quantitative relationship data set of protective performance; combine the quantitative relationship data set to establish a multi-objective optimization model to generate an optimal structural parameter combination that meets the preset safety standard constraint conditions. Through three-dimensional point cloud processing and mesh modeling, digital modeling of the protective shoe is realized. By means of area division and feature extraction, accurate analysis of key structural parameters is achieved. Using finite element simulation and multi-objective optimization, quantitative evaluation of protective performance and optimization of structural parameters are realized. This method solves the problems of relying on manual operations and empirical judgments in traditional designs, improves the efficiency and accuracy of protective shoe design, provides a data-driven standardized process for protective shoe structure design, and effectively ensures that the product meets safety standards and optimizes comprehensive performance. Description of the Drawings

[0043] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the following will briefly introduce the drawings required for use in the description of the embodiments or related technologies. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0044] Figure 1 It is a flowchart of a method for designing a footwear style based on a digital three-dimensional model provided by an embodiment of the present invention; Figure 2 It is a flowchart of acquiring three-dimensional point cloud data of a protective shoe, performing noise reduction processing on the three-dimensional point cloud data, and extracting distribution features to construct a three-dimensional mesh model provided by an embodiment of the present invention; Figure 3 It is a flowchart of constructing a finite element mechanics simulation model based on a feature vector data set to generate a quantitative relationship data set of protective performance provided by an embodiment of the present invention. Detailed Embodiments

[0045] To make the objectives, technical solutions and advantages of this application more clear and understandable, the following further elaborates on this application in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely for explaining this application and are not intended to limit this application.

[0046] In one of the embodiments, as Figure 1 shown, this application provides a method for designing footwear styles based on a digital three-dimensional model, which may include the following steps: Step S101: Obtain the three-dimensional point cloud data of the protective shoes, perform noise reduction processing on the three-dimensional point cloud data, and extract distribution features to construct a three-dimensional mesh model.

[0047] Specifically, to obtain the three-dimensional point cloud data of the protective shoes (including shoes with different functions such as anti-smashing, anti-puncturing, insulation, anti-slip, oil resistance, acid and alkali resistance, etc.), laser scanning or structured light scanning technology can be used to collect the spatial coordinate information on the surface of the shoes, forming the original data containing a dense point set. When performing noise reduction processing on this data, the neighborhood search radius threshold is calculated based on the point cloud density information, and the noise points are removed through a statistical outlier filtering algorithm, while retaining the real surface features. Subsequently, the normal vector distribution and curvature continuity features of the point cloud are extracted for subsequent mesh model construction to ensure the accurate expression of the complex curved surfaces of the shoes (such as the sole arc surface and the shoe upper bending area).

[0048] Step S102: Divide the three-dimensional mesh model into a sole area, a shoe upper area, and an anti-smashing head area to obtain the closed boundary point sequences of each area.

[0049] The preprocessed three-dimensional mesh model is divided into a sole, a shoe upper, and an anti-smashing head area, and automatic segmentation is achieved by using a region growing algorithm based on curvature distribution combined with connectivity constraints. First, the curvature characteristic parameters of the mesh vertices are calculated to generate a curvature distribution map, and based on this, the high-curvature boundary regions are identified; then, through region merging and boundary optimization, the closed boundary point sequences of each functional region are determined.

[0050] Step S103: Extract the sole thickness distribution value, the anti-smashing head curvature value, and the shoe upper fold degree value according to the closed boundary point sequences to generate a feature vector data set.

[0051] Specifically, key structure parameters are extracted based on the closed boundary point sequences: the sole thickness distribution value is obtained by calculating the distance between the corresponding points on the upper and lower surfaces of the sole, the anti-smashing head curvature value is derived by analyzing the curvature continuous parameters of the boundary points, and the shoe upper fold degree value is determined by identifying the density distribution parameters of the local curvature change on the surface. These parameters are integrated into a multi-dimensional feature vector, and after dimensionality reduction processing by principal component analysis, a feature vector data set containing geometric and mechanical characteristics is generated.

[0052] Step S104: Build a finite element mechanical simulation model based on the feature vector dataset to generate a quantitative relationship dataset of the protection performance.

[0053] Furthermore, build a finite element mesh model using the material properties and geometric parameters in the feature vector dataset, and define the model space discretization form by setting the node coordinates and element topological relationships. Invoke the mechanical solver to calculate the stress distribution data of the model under impact loads, synchronously analyze the characteristics of the electric field distribution, and obtain the protection performance index values (such as the impact resistance coefficient and the electric field shielding rate) through multi-physics field coupling calculations. Associate these indices with the geometric parameters to generate a quantitative dataset reflecting the relationship between the structural parameters and the protection performance.

[0054] Step S105: Establish a multi-objective optimization model based on the quantitative relationship dataset to generate an optimal combination of structural parameters that meets the preset safety standard constraint conditions.

[0055] Specifically, establish a multi-objective optimization model based on the quantitative relationship dataset, with the protection performance indices, the plantar pressure distribution, and the shear force change value as the objective functions, and the preset safety standard as the constraint condition. Use the gradient descent algorithm to dynamically adjust the step size of the structural parameters, generate a sequence of parameter combinations and input them into the optimization model, and find the double-objective balance point of the pressure and the shear force through iterative calculations. Finally, screen out the parameter combination that meets the safety constraints and has the optimal comprehensive performance, realizing the quantitative optimization and standardized verification of the protective shoe structure design.

[0056] The above method for designing the shoe style based on the digital three-dimensional model obtains the three-dimensional point cloud data of the protective shoe, extracts the distribution features after noise reduction processing to build a three-dimensional mesh model; divides the model into the sole, the upper, and the anti-smashing head area, and obtains the sequence of closed boundary points in each area; extracts the sole thickness distribution value, the anti-smashing head curvature value, and the upper wrinkle degree value based on the boundary point sequence to generate a feature vector dataset; builds a finite element mechanical simulation model using the feature vector dataset to generate a quantitative relationship dataset of the protection performance; combines the quantitative relationship dataset to establish a multi-objective optimization model to generate an optimal combination of structural parameters that meets the preset safety standard constraint conditions. Through three-dimensional point cloud processing and mesh modeling, the digital modeling of the protective shoe is realized. By means of area division and feature extraction, the accurate analysis of the key structural parameters is achieved. Using finite element simulation and multi-objective optimization, the quantitative evaluation of the protection performance and the optimization of the structural parameters are realized. This method solves the problems of relying on manual operations and empirical judgments in traditional designs, improves the efficiency and accuracy of the protective shoe design, provides a data-driven standardized process for the protective shoe structure design, and effectively ensures that the product meets the safety standards and optimizes the comprehensive performance.

[0057] In one embodiment, such as Figure 2As shown, to obtain the three-dimensional point cloud data of the protective shoes, perform noise reduction processing on the three-dimensional point cloud data, and extract distribution features to construct a three-dimensional mesh model, the following steps may be included: Step S201, obtain the three-dimensional point cloud data of the protective shoes; the three-dimensional point cloud data contains point cloud density information.

[0058] Step S202, determine the noise reduction parameter threshold for controlling the point cloud neighborhood search radius according to the point cloud density information.

[0059] Step S203, perform outlier filtering on the three-dimensional point cloud data based on the noise reduction parameter threshold to obtain the noise-reduced three-dimensional point cloud data.

[0060] Step S204, extract the surface normal vector distribution features of the noise-reduced three-dimensional point cloud data; the surface normal vector distribution features contain curvature continuity information.

[0061] Step S205, generate a triangular patch topology structure according to the curvature continuity information, and perform Poisson reconstruction on the noise-reduced three-dimensional point cloud data to obtain an initial three-dimensional mesh model.

[0062] Step S206, calculate the vertex curvature gradient of the initial three-dimensional mesh model, perform subdivision optimization on the initial three-dimensional mesh model, and generate the final three-dimensional mesh model.

[0063] Specifically, obtain the three-dimensional point cloud data of the protective shoes containing point cloud density information, determine the noise reduction parameter threshold for controlling the point cloud neighborhood search radius according to the point cloud density information, perform outlier filtering on the three-dimensional point cloud data based on this threshold to obtain the noise-reduced point cloud data; extract the surface normal vector distribution features (including curvature continuity information) of the noise-reduced point cloud data, generate a triangular patch topology structure according to the curvature continuity, and obtain the initial three-dimensional mesh model through Poisson reconstruction; calculate the vertex curvature gradient of the initial model and perform subdivision optimization on it to generate the final three-dimensional mesh model.

[0064] In this embodiment, the noise reduction parameters are adaptively determined by the point cloud density, realizing precise noise filtering of the three-dimensional point cloud data of the protective shoes and retaining the true geometric features of the shoe body surface; based on Poisson reconstruction and vertex curvature gradient subdivision optimization of curvature continuity, the three-dimensional mesh model can accurately reflect the details of the complex structure of the protective shoes (such as the curved surface radian of the anti-smashing head and the sole thickness change area), avoiding problems such as patch breakage or feature blurring in the model. This method improves the automation degree and modeling accuracy of point cloud processing, provides a high-quality three-dimensional model basis for subsequent regional segmentation and mechanical property simulation of the protective shoes, and enhances the reliability and efficiency of the digital design process.

[0065] In one of the embodiments, the vertex curvature gradient can be calculated by the following formula: ; Among them, ) represents the vertex curvature gradient of the th grid, represents the th vertex in the grid, represents the vertex 's first-order neighborhood vertex set, represents the number of neighborhood vertices, and represent the principal curvature values of vertex and vertex , represents the unit direction vector pointing from vertex to vertex , represents the weighting coefficient, represents the scaling parameter that controls the influence of curvature difference.

[0066] Preferably, the principal curvature of the vertex can be calculated by the following steps: Calculate the covariance matrix of the neighborhood of vertex : : ; Among them, represents the centroid of the neighborhood vertices.

[0067] Perform eigenvalue decomposition on the covariance matrix to obtain three eigenvalues , and their corresponding eigenvectors , , .

[0068] The normal vector of the vertex is the eigenvector corresponding to the minimum eigenvalue .

[0069] The principal curvature of the vertex can be approximated as: ; In this embodiment, in the area with a drastic curvature change (such as the junction of the anti-impact head arc and the shoe upper of the safety shoe), the curvature difference of neighboring vertices automatically reduces the influence of vertices with a large curvature difference at a long distance, focuses on the neighboring vertices with significant local features, and drives the mesh subdivision algorithm to generate denser patches in the key area, accurately depicting the details of the surface turning. By suppressing the ineffective participation of vertices at a long distance through geometric distance weighting, it avoids the waste of computing resources caused by global uniform subdivision, reduces redundant patches while ensuring the model accuracy, and improves the operation efficiency of subsequent mechanical simulation and optimization. This method solves the problems of insufficient characterization or over-subdivision in the feature area in traditional mesh optimization, and provides theoretical support for the high-precision modeling of the three-dimensional model of the safety shoe.

[0070] In one embodiment, the three-dimensional mesh model is divided into a sole area, a shoe upper area, and an anti-impact head area to obtain the closed boundary point sequences of each area, which may include the following steps: Step S301, obtain the vertex coordinate data and triangular patch topology data of the three-dimensional mesh model.

[0071] Step S302, calculate the curvature characteristic parameters of each vertex according to the vertex coordinate data to obtain a curvature distribution map.

[0072] Preferably, the local bending degree of the mesh surface is quantified through geometric calculation. Specifically, based on the vertex coordinate data of the three-dimensional mesh model, first construct the local neighborhood of each vertex (usually select the first-order or second-order neighborhood), fit the local tangent plane of the neighborhood point set by the least squares method, and then calculate the deviation of the vertex relative to the tangent plane. Commonly used curvature calculation methods include Mean Curvature and Gaussian Curvature. Among them, the mean curvature reflects the average bending degree of the surface in two principal curvature directions. After completing the curvature calculation by traversing all vertices, map the curvature values of each vertex to the corresponding spatial coordinates to generate a visual curvature distribution map, providing a quantitative geometric feature basis for subsequent region segmentation. This calculation process does not rely on subjective judgment, objectively reflects the bending characteristics of the mesh surface through mathematical methods, and lays a foundation for accurately identifying high-curvature areas such as the sole bend and anti-impact head arc surface of the safety shoe.

[0073] Step S303, extract the high-curvature areas based on the curvature distribution map, and combine the regional connectivity constraints to merge adjacent triangular patches to generate initial segmentation regions.

[0074] Step S304, perform dilation and erosion operations on the boundaries of the initial segmentation regions according to a preset segmentation threshold to determine the boundary point sets of the sole area, the shoe upper area, and the anti-impact head area.

[0075] Step S305: Perform topological relation analysis on the set of boundary points. After removing isolated points, match with the topological data of triangular patches to obtain the closed boundary point sequences of each region.

[0076] Specifically, obtain the vertex coordinate data and triangular patch topological data of the 3D mesh model, calculate the curvature characteristic parameters of each vertex based on the vertex coordinates, and generate a curvature distribution map; extract high-curvature regions according to the curvature distribution map, and combine regional connectivity constraints to perform regional merging on adjacent triangular patches to generate initial segmentation regions; perform dilation and erosion operations on the boundaries of the initial regions through a preset segmentation threshold to determine the set of boundary points for the sole, upper, and anti-impact head regions; perform topological relation analysis on the set of boundary points, and after removing isolated points, match with the triangular patch topological data to obtain the closed boundary point sequences of each region.

[0077] In this embodiment, the characteristic regions of the 3D mesh model of the protective shoe are identified through the curvature characteristic parameters and the curvature distribution map, and the geometric integrity of the segmentation result is ensured by using regional connectivity constraints and topological analysis. The extraction of high-curvature regions combined with dilation and erosion operations can accurately define the boundaries of functional regions such as the sole, upper, and anti-impact head, avoiding the subjectivity and errors of traditional manual segmentation. The generation of the closed boundary point sequences provides an accurate geometric boundary basis for subsequent extraction of the structural parameters of each region (such as the sole thickness and the curvature of the anti-impact head), improving the efficiency and accuracy of multi-region feature decoupling in the digital design of protective shoes.

[0078] In one embodiment, according to the closed boundary point sequences, extract the sole thickness distribution values, anti-impact head curvature values, and upper wrinkle degree values to generate a feature vector data set, which may include the following steps: Step S401: Extract the boundary point curvature continuity parameter and thickness gradient change rate according to the closed boundary point sequences.

[0079] Step S402: Calculate the change rate of the anti-impact head curvature radius according to the curvature continuity parameter to generate a curvature radius distribution matrix.

[0080] Step S403: Combine the thickness gradient change rate and the curvature radius distribution matrix to generate a three-dimensional feature mapping table.

[0081] Step S404: Extract the density distribution parameter of the wrinkled area from the three-dimensional feature mapping table, and perform dimensionality reduction processing on the parameter using the principal component analysis method to obtain a low-dimensional feature vector set.

[0082] Step S405: Construct a feature vector similarity matrix according to the low-dimensional feature vector set; the similarity matrix is used to judge the correlation between the upper wrinkle degree and the anti-impact head curvature.

[0083] Step S406: If the correlation exceeds a preset threshold, update the curvature radius distribution weight coefficient in the three-dimensional feature mapping table.

[0084] Step S407: Adjust the dimensional distribution of the low-dimensional feature vector set according to the updated weight coefficients, and output the optimized feature vector data set.

[0085] Specifically, first extract the boundary point curvature continuity parameter and the thickness gradient change rate from the closed boundary point sequence, calculate the change rate of the anti-smashing head curvature radius using the curvature continuity parameter to form a curvature radius distribution matrix; then combine the thickness gradient change rate with the curvature radius distribution matrix to generate a three-dimensional feature mapping table; subsequently, extract the density distribution parameter of the wrinkled area from the mapping table and reduce the dimension through principal component analysis to obtain a low-dimensional feature vector set; then construct a similarity matrix based on the low-dimensional feature vector set to judge the correlation between the shoe upper wrinkle degree and the anti-smashing head curvature; if the correlation exceeds the preset threshold, update the curvature radius distribution weight coefficient of the three-dimensional feature mapping table, and accordingly adjust the dimensional distribution of the low-dimensional feature vector set, and output the optimized feature vector data set.

[0086] In this embodiment, by quantitatively analyzing the curvature and thickness characteristics of the boundaries of each region of the protective shoe, the systematic extraction and integration of the anti-smashing head curvature change, the sole thickness distribution, and the shoe upper wrinkle characteristics are realized. The dimension reduction by principal component analysis and the feature correlation judgment can remove redundant information, optimize the feature vector structure, ensure that the data can fully reflect the geometric and mechanical characteristics of the product, and reduce the computational complexity of subsequent finite element simulation and multi-objective optimization. This method provides accurate and efficient feature data support for the design of protective shoes, which helps to improve the accuracy and efficiency of product performance analysis and structural optimization.

[0087] In one of the embodiments, as Figure 3 shown, based on the feature vector data set, a finite element mechanical simulation model is constructed to generate a quantitative relationship data set of protective performance, which may include the following steps: Step S501: Construct a finite element mesh model according to the material properties and geometric parameters in the feature vector data set; the finite element mesh model includes node coordinates and element topological relationships.

[0088] Preferably, extract the material properties (such as elastic modulus, Poisson's ratio, conductivity, etc.) of each part of the protective shoe from the feature vector data set, and the geometric parameters (such as sole thickness, anti-smashing head curvature radius, etc.) obtained based on the three-dimensional mesh model. Discretize the protective shoe solid model into a finite element mesh including node coordinates and element topological relationships through finite element preprocessing technology. The node coordinates accurately describe the position of the mesh in three-dimensional space, while the element topological relationship defines the connection method between nodes (such as tetrahedron and hexahedron elements), ensuring that the model can accurately reflect the geometric shape and structural characteristics of the protective shoe. This process needs to balance the mesh density and computational efficiency, and use encrypted meshes at key parts (such as the connection between the anti-smashing head and the sole) to capture stress concentration phenomena, while using sparse meshes in flat areas to reduce the amount of calculation.

[0089] Step S502: Invoke a mechanical solver to calculate the stress distribution data of the finite element mesh model.

[0090] Step S503: Synchronously invoke an electric field solver to calculate the electric potential gradient distribution of the finite element mesh model.

[0091] Step S504: Perform a coupled analysis on the stress distribution data and the electric potential gradient distribution, and extract the anti-impact performance index values; the anti-impact performance index values include the anti-impact coefficient and the electric field shielding rate.

[0092] Step S505: Generate a quantitative relationship data set of the protection performance according to the anti-impact performance index values and the geometric parameters.

[0093] Specifically, construct a finite element mesh model according to the material properties and geometric parameters in the eigenvector data set. The model includes node coordinates and element topological relationships; invoke a mechanical solver to calculate the stress distribution data of the model, and synchronously invoke an electric field solver to calculate the electric potential gradient distribution; perform a coupled analysis on the stress distribution and the electric potential gradient distribution, and extract anti-impact performance index values such as the anti-impact coefficient and the electric field shielding rate; generate a quantitative relationship data set of the protection performance according to the performance index values and the geometric parameters.

[0094] In this embodiment, the materials and geometric features of the safety shoes are integrated through the finite element mesh model, realizing the coupled simulation of multiple physical fields of mechanics and electric fields, and accurately quantifying the anti-impact performance index. The coupled analysis of the stress and electric potential distributions breaks through the limitations of single physical field evaluation, and can comprehensively reflect the structural strength and electric field protection ability of the safety shoes under impact loads. The generated quantitative relationship data set provides data support for subsequent multi-objective optimization, enabling the design process to shift from experience-driven to data-driven, improving the scientificity and optimization efficiency of the safety shoe performance evaluation, ensuring that the product meets safety standards and achieving multi-dimensional performance balance.

[0095] In one of the embodiments, a multi-objective optimization model is established according to the quantitative relationship data set, and an optimal structural parameter combination that meets the preset safety standard constraint conditions can include the following steps: Step S601: Extract the plantar pressure distribution matrix and the shear force gradient matrix according to the quantitative relationship data set; the pressure distribution matrix includes the pressure intensity data of each region; the shear force gradient matrix includes the dynamic shear change amount.

[0096] Step S602: Calculate the safety evaluation function value by weighted fusion of the pressure distribution matrix and the shear force gradient matrix.

[0097] Specifically, based on the pressure intensity data of each region in the plantar pressure distribution matrix and the dynamic shear change amount in the shear force gradient matrix in the quantitative relationship dataset, a regional weight coefficient and a weighting coefficient are introduced for comprehensive calculation. The regional weight coefficient is set differently according to the sensitivity of different parts of the sole (such as the arch, heel, and toes) to pressure and shear force, reflecting the importance of each region in the safety assessment; the weighting coefficient is used to adjust the relative proportion of pressure and shear force in the assessment. At the same time, the original pressure and shear force data are normalized using a non-linear mapping function, converting them into comparable values and eliminating the influence of dimensional differences. Through weighted summation, the pressure and shear force data are fused into a single safety assessment function value, which can comprehensively and objectively reflect the adaptability and safety of the protective shoes to the plantar mechanical environment during actual use.

[0098] Step S603, determine whether the safety assessment function value exceeds a preset threshold. If it does not exceed, dynamically adjust the step size of the structural parameters based on the gradient descent algorithm.

[0099] Step S604, update the structural parameter combination based on the step size of the structural parameters to generate a new sequence of parameter combinations.

[0100] Step S605, input the sequence of parameter combinations into the multi-objective optimization model to obtain a set of objective function values.

[0101] Step S606, screen the optimal parameter identification numbers that meet the constraints according to the set of objective function values to obtain the optimal structural parameter combination; the optimal parameter identification number corresponds to the dual-objective balance point of pressure and shear force.

[0102] Furthermore, obtain the plantar pressure distribution matrix containing the pressure intensity data of each region and the shear force gradient matrix containing the dynamic shear change amount, and calculate the safety assessment function value through weighted fusion; determine whether the function value exceeds the preset threshold. If it does not exceed, dynamically adjust the step size of the structural parameters based on the gradient descent algorithm, update the parameter combination to generate a new sequence; input the new sequence into the multi-objective optimization model to obtain a set of objective function values, and screen the optimal parameter identification numbers that meet the constraints from it, corresponding to the optimal structural parameter combination at the dual-objective balance point of pressure and shear force.

[0103] In this embodiment, a safety assessment system is constructed by fusing plantar pressure and shear force data to achieve a quantitative assessment of the ergonomic performance of protective shoes. Based on the parameter adjustment based on gradient descent and the multi-objective optimization model, the balance solution of pressure and shear force can be automatically searched under the preset safety standard, avoiding the blindness of subjective parameter adjustment in traditional designs. This method incorporates the protection performance and wearing comfort into a unified optimization framework. The generated optimal parameter combination can not only meet safety indicators such as anti-smashing and electrical insulation, but also optimize the plantar force distribution, improve the comprehensive performance and market adaptability of the product, and promote the development of protective shoe design towards intelligence and precision.

[0104] In one of the embodiments, the safety evaluation function value can be calculated by the following formula: ; wherein, represents the safety evaluation function value, represents the plantar pressure distribution matrix, represents the th row and th column of the pressure distribution matrix, represents the shear force gradient matrix, represents the dynamic shear change of the th row and th column of the shear force gradient matrix, represents the regional weight coefficient, and represent the weighting coefficients, represents the non-linear mapping function of pressure and shear force.

[0105] Preferably, the non-linear mapping function of pressure and shear force can be expressed by the following formula: ; wherein, and respectively represent the optimal values of pressure and shear force, and represent the parameters controlling the function shape.

[0106] In this embodiment, by quantifying and fusing plantar pressure and shear force data, a systematic safety evaluation system is constructed to avoid the one-sidedness of single-index evaluation. The introduction of non-linear mapping and weighting mechanisms can accurately reflect the differential impacts of pressure intensity and dynamic shear changes in different plantar regions on safety, making the evaluation results more in line with the actual usage scenarios. This data-driven evaluation method provides a scientific quantitative basis for the multi-objective optimization of protective shoes, helps to quickly identify the relationship between structural parameters and safety performance during the design stage, improves the efficiency and reliability of product design, and promotes the development of the safety performance evaluation of protective shoes towards standardization and precision.

[0107] It should be understood that although the steps in the flowcharts involved in the above-described embodiments are shown in sequence according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless there is a clear indication in this article, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-described embodiments may include multiple steps or multiple stages. These steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be executed alternately or in turn with at least a part of other steps or steps or stages in other steps.

[0108] In one embodiment, a computer device is provided, including a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, the steps of a method for designing footwear styles based on a digital three-dimensional model as described above are implemented.

[0109] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0110] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can refer to the partial descriptions of the method embodiments. The device embodiments described above are only illustrative. The components described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the present disclosure solution. A person of ordinary skill in the art can understand and implement it without creative work.

[0111] The above-described embodiments only represent several implementation manners of the embodiments of the present application. The description is relatively specific and detailed, but it should not be construed as a limitation on the patent scope of the embodiments of the application. It should be noted that for those of ordinary skill in the art, without departing from the concept of the embodiments of the present application, several modifications and improvements can still be made, and these all belong to the protection scope of the embodiments of the present application.

Claims

1. A footwear style design method based on a digital three-dimensional model, characterized in that, The method includes: Obtaining the three-dimensional point cloud data of the protective shoes, performing noise reduction processing on the three-dimensional point cloud data, and extracting distribution features to construct a three-dimensional mesh model; Dividing the three-dimensional mesh model into a sole area, an upper area, and an anti-impact head area to obtain a closed boundary point sequence for each area; Extracting the sole thickness distribution value, the anti-impact head curvature value, and the upper fold degree value according to the closed boundary point sequence to generate a feature vector data set; Constructing a finite element mechanical simulation model based on the feature vector data set to generate a quantitative relationship data set for protective performance; Establishing a multi-objective optimization model according to the quantitative relationship data set to generate an optimal structural parameter combination that meets the preset safety standard constraint conditions.

2. The method according to claim 1, wherein The obtaining the three-dimensional point cloud data of the protective shoes, performing noise reduction processing on the three-dimensional point cloud data, and extracting distribution features to construct a three-dimensional mesh model includes: Obtaining the three-dimensional point cloud data of the protective shoes; the three-dimensional point cloud data includes point cloud density information; Determining a noise reduction parameter threshold for controlling the neighborhood search radius of the point cloud according to the point cloud density information; Performing outlier filtering on the three-dimensional point cloud data based on the noise reduction parameter threshold to obtain the noise-reduced three-dimensional point cloud data; Extracting the surface normal vector distribution features of the noise-reduced three-dimensional point cloud data; the surface normal vector distribution features include curvature continuity information; Generating a triangular patch topological structure according to the curvature continuity information, and performing Poisson reconstruction on the noise-reduced three-dimensional point cloud data to obtain an initial three-dimensional mesh model; Calculating the vertex curvature gradient of the initial three-dimensional mesh model and performing subdivision optimization on the initial three-dimensional mesh model to generate the final three-dimensional mesh model.

3. The method according to claim 2, wherein The vertex curvature gradient is calculated by the following formula: ； Among them, represents the vertex curvature gradient of the th grid, represents the vertex in the th grid, represents the set of first-order neighborhood vertices of vertex , represents the number of neighborhood vertices, and represent the principal curvature values of vertex and vertex , represents the unit direction vector pointing from vertex to vertex , represents the weighting coefficient, represents the scaling parameter that controls the influence of curvature difference.

4. The method according to claim 1, characterized in that The dividing the three-dimensional mesh model into a sole area, an upper area, and an anti-impact head area to obtain a closed boundary point sequence for each area includes: Obtaining the vertex coordinate data and triangular patch topological data of the three-dimensional mesh model; Calculating the curvature characteristic parameters of each vertex according to the vertex coordinate data to obtain a curvature distribution map; Extracting high-curvature regions based on the curvature distribution map, and combining regional connectivity constraints to merge adjacent triangular patches to generate initial segmentation regions; Performing dilation and erosion operations on the boundaries of the initial segmentation regions according to a preset segmentation threshold to determine the boundary point sets of the sole area, the upper area, and the anti-impact head area; Performing topological relationship analysis on the boundary point sets, removing isolated points, and then matching them with the triangular patch topological data to obtain a closed boundary point sequence for each area.

5. The method according to claim 1, wherein The extracting the sole thickness distribution value, the anti-impact head curvature value, and the upper fold degree value according to the closed boundary point sequence to generate a feature vector data set includes: Extracting the boundary point curvature continuous parameter and the thickness gradient change rate according to the closed boundary point sequence; Calculating the change rate of the anti-impact head curvature radius according to the curvature continuous parameter to generate a curvature radius distribution matrix; Combining the thickness gradient change rate and the curvature radius distribution matrix to generate a three-dimensional feature mapping table; Extract the density distribution parameters of the wrinkle area from the three-dimensional feature mapping table, and use the principal component analysis method to reduce the dimension of the parameters to obtain a low-dimensional feature vector set; Construct a feature vector similarity matrix according to the low-dimensional feature vector set; the similarity matrix is used to judge the correlation between the shoe upper wrinkle degree and the anti-smashing head curvature; If the correlation exceeds the preset threshold, update the curvature radius distribution weight coefficient in the three-dimensional feature mapping table; Adjust the dimension distribution of the low-dimensional feature vector set according to the updated weight coefficient, and output an optimized feature vector data set.

6. The method according to claim 1, characterized in that Construct a finite element mechanical simulation model based on the feature vector data set to generate a quantitative relationship data set of protection performance, including: Construct a finite element mesh model according to the material properties and geometric parameters in the feature vector data set; the finite element mesh model includes node coordinates and element topological relationships; Invoke a mechanical solver to calculate the stress distribution data of the finite element mesh model; Synchronously invoke an electric field solver to calculate the electric potential gradient distribution of the finite element mesh model; Perform a coupling analysis on the stress distribution data and the electric potential gradient distribution, and extract the anti-smashing performance index value; the anti-smashing performance index value includes an anti-impact coefficient and an electric field shielding rate; Generate a quantitative relationship data set of protection performance according to the anti-smashing performance index value and the geometric parameters.

7. The method according to claim 1, characterized in that, Establish a multi-objective optimization model according to the quantitative relationship data set to generate an optimal structural parameter combination that meets the preset safety standard constraint conditions, including: Extract the plantar pressure distribution matrix and the shear force gradient matrix from the quantitative relationship data set; the pressure distribution matrix includes the pressure intensity data of each region; the shear force gradient matrix includes the dynamic shear change amount; Calculate the safety evaluation function value by weighted fusion of the pressure distribution matrix and the shear force gradient matrix; Judge whether the safety evaluation function value exceeds the preset threshold. If it does not exceed, dynamically adjust the structural parameter step size based on the gradient descent algorithm; Update the structural parameter combination based on the structural parameter step size to generate a new parameter combination sequence; Input the parameter combination sequence into the multi-objective optimization model to obtain a set of objective function values; Screen the optimal parameter identification numbers that meet the constraints according to the set of objective function values to obtain an optimal structural parameter combination; the optimal parameter identification number corresponds to the double-objective balance point of pressure and shear force.

8. The method according to claim 7, characterized in that The safety evaluation function value is calculated by the following formula: ； Among them, represents the safety evaluation function value, represents the plantar pressure distribution matrix, represents the pressure intensity value at the th row and the th column in the pressure distribution matrix, represents the shear force gradient matrix, represents the dynamic shear change amount at the th row and the th column in the shear force gradient matrix, represents the regional weight coefficient, and represent the weighting coefficient, represents the non-linear mapping function of pressure and shear force.

9. A computer device, comprising a memory and a processor, the memory storing a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the method according to any one of claims 1 to 8.

Images