Automatic production line digital model reconstruction method and device

Abstract

The embodiment of the invention provides an automatic production line digital model reconstruction method and device, geometric semantic features are extracted through hierarchical slicing and a deep convolutional network, and an implicit curved surface reconstruction framework with physical constraints is innovatively constructed. The system adopts an octree structure and a parallel computing framework to realize efficient neighbor point search, and performs feature extraction and noise filtering in combination with principal component analysis and covariance matrix decomposition. Through moving least square fitting and an error compensation mechanism of a self-adaptive basis function, the accuracy and continuity of curved surface reconstruction are ensured, and seamless integration with a collision detection function of an automatic production line is realized. According to the method, the limitation of traditional model reconstruction is broken through, and an efficient and reliable solution is provided for industrial digital twinning.

Description

Technical Field

[0001] This application relates to the field of data processing, and particularly to a method and device for reconstructing a digital model of an automated production line. Background Art

[0002] Traditional digital model reconstruction methods mainly rely on a single geometric reconstruction algorithm, making it difficult to handle the requirements for model integrity and accuracy in complex industrial scenarios. Existing technologies lack a systematic method in feature extraction and semantic understanding, and have low computational efficiency when dealing with large-scale point cloud data.

[0003] At the same time, existing systems have obvious deficiencies in surface reconstruction and topology preservation. Traditional methods often ignore physical constraint conditions, resulting in a deviation between the reconstructed model and the actual working conditions. The system is also relatively simple in surface stitching and error compensation, and fails to achieve high-quality model reconstruction.

[0004] In addition, existing technologies also have limitations in parallel computing optimization and industrial application integration. There is a lack of an efficient nearest neighbor search strategy and an adaptive surface fitting mechanism, and the actual application requirements in the industrial environment are not fully considered. Solving these problems is of great significance for improving the efficiency and accuracy of digital model reconstruction. Summary of the Invention

[0005] In view of the problems in the prior art, this application provides a method and device for reconstructing a digital model of an automated production line, which can break through the limitations of traditional model reconstruction and provide an efficient and reliable solution for industrial digital twins.

[0006] To solve at least one of the above problems, this application provides the following technical solutions:

[0007] In a first aspect, this application provides a method for reconstructing a digital model of an automated production line, including:

[0008] Slice the digital model of the automated production line along the main coordinate axes, extract the sequence of feature points of the slice contour line at different resolutions, perform geometric semantic annotation on the sequence of feature points, construct a variational level set function from the annotated sequence of feature points, extract a hierarchical feature map from the variational level set function using a deep convolutional network, establish an implicit surface reconstruction framework with physical constraint terms based on the hierarchical feature map, and embed the physical constraint terms as regularization conditions into the implicit surface reconstruction process;

[0009] Under the implicit surface reconstruction framework, an octree structure is used to perform spatial partitioning on the sequence of feature points, a nearest neighbor search algorithm based on the spatial partitioning is constructed, the nearest neighbor search algorithm is mapped into a parallel computing framework to establish an acceleration strategy, a local coordinate system is constructed according to the nearest neighbor search result, the principal component analysis method is used to calculate the orthogonal basis vectors of the feature points in the local coordinate system, an orthogonal basis space is established, the feature point projection matrix in the orthogonal basis space is decomposed into a three-dimensional covariance matrix, and the three-dimensional covariance matrix is input into a noise filtering module for data enhancement;

[0010] Based on the eigenvectors of the three-dimensional covariance matrix, a local surface patch is constructed, the local surface patch is subjected to moving least squares fitting using an adaptive basis function, topological constraints and smoothness constraints are added during the fitting process, the fitted surface patch is geometrically matched with the slice contour line, the surface patches are assembled and spliced according to the geometric matching result, the assembled and spliced model is iteratively refined, an error compensation mechanism is established to eliminate surface seams, and the compensated reconstruction model is imported into the three-dimensional environment of the automated production line for collision detection.

[0011] Further, the method of slicing the digital model of the automated production line along the main coordinate axes, extracting the sequence of feature points of the slice contour line at different resolutions, and performing geometric semantic annotation on the sequence of feature points, and constructing a variational level set function from the annotated sequence of feature points includes:

[0012] The slicing direction of the main coordinate axis is calculated using the hierarchical scanning algorithm of the three-dimensional grid model, a slice plane parallel to the coordinate axis is constructed based on the slicing direction, the slice plane intersects with the boundary surface of the digital model to generate a slice contour line, the grid division of the slice contour line is performed to calculate the spatial coordinates of the feature points, and an adaptive resolution sampling mechanism is established to screen the feature points to obtain an ordered sequence of feature points;

[0013] A curvature tensor analysis model is established for the sequence of feature points, the geometric attributes of the feature points are calculated based on the principal directions of the curvature tensor, the geometric attributes are used as label information to endow the feature points, a variational objective functional with weight coefficients is constructed, and the mathematical expression of the level set function is obtained by solving the variational objective functional, thereby establishing a mapping relationship from the sequence of feature points to the level set function.

[0014] Further, the method of using a deep convolutional network to extract hierarchical feature maps from the variational level set function, and establishing an implicit surface reconstruction framework with physical constraint terms, and embedding the physical constraint terms as regularization conditions into the implicit surface reconstruction process includes:

[0015] Construct a multi-scale convolutional neural network structure, set convolutional kernels of different sizes in the convolutional layer to extract the spatial features of the variational level set function, perform dimensionality reduction and compression on the feature map through pooling operations, input the dimensionality-reduced feature map into the transposed convolutional layer for upsampling, and splice and fuse the upsampling result with the feature map of the skip connection to generate a hierarchical feature map;

[0016] Based on the hierarchical feature map, construct an implicit surface equation, add physical constraint terms for surface smoothness and shape preservation to the implicit surface equation, transform the physical constraint terms into regularization conditions in the form of Lagrange multipliers, establish an objective function with a regularization term, and use the variational method to solve the objective function to obtain an expression for implicit surface reconstruction.

[0017] Further, under the implicit surface reconstruction framework, use an octree structure to perform spatial partitioning on the sequence of feature points, construct a nearest neighbor search algorithm based on the spatial partitioning, and map the nearest neighbor search algorithm to a parallel computing framework to establish an acceleration strategy, including:

[0018] Calculate the spatial bounding box of the sequence of feature points, divide the spatial bounding box into eight subspaces along the coordinate axes, calculate the density distribution of feature points in each subspace, recursively subdivide the subspaces based on the density distribution threshold, establish an octree index structure to store the spatial position information of feature points, and record the spatial position information of the feature points in the corresponding leaf nodes;

[0019] Construct a nearest neighbor search path in the octree index structure, calculate the spatial distance between the feature point to be searched and the leaf nodes, mark the leaf nodes with a spatial distance less than the search radius as candidate nodes, establish parallel search tasks for the candidate nodes, allocate the parallel search tasks to computing units to perform nearest neighbor search calculations, and summarize the search results to obtain the neighborhood information of the feature points.

[0020] Further, construct a local coordinate system according to the nearest neighbor search result, use the principal component analysis method to calculate the orthogonal basis vectors of the feature points in the local coordinate system, establish an orthogonal basis space, decompose the feature point projection matrix in the orthogonal basis space into a three-dimensional covariance matrix, and input the three-dimensional covariance matrix into a noise filtering module for data enhancement, including:

[0021] Use the nearest neighbor search result to calculate the barycentric coordinates of the feature point neighborhood, translate the barycentric coordinates to the origin to establish the origin of the local coordinate system, calculate the divergence matrix of the feature point neighborhood, perform eigenvalue decomposition on the divergence matrix to obtain an eigenvector matrix, sort the eigenvector matrix according to the eigenvalue magnitude to determine the three basis vectors of the local coordinate system, and construct a rotation matrix of the local coordinate system based on the basis vectors;

[0022] Transform the feature points from the global coordinate system to the local coordinate system, calculate the coordinate components of the feature points in the local coordinate system, construct the projection matrix of the feature points, perform singular value decomposition on the projection matrix to obtain the three-dimensional covariance matrix, calculate the noise threshold based on the eigenvalue distribution of the three-dimensional covariance matrix, and input the noise threshold as a filtering parameter into the anisotropic diffusion filter.

[0023] Further, construct a local surface patch based on the eigenvectors of the three-dimensional covariance matrix, perform moving least squares fitting on the local surface patch using an adaptive basis function, and add topological constraints and smoothness constraints during the fitting process. Geometric matching of the fitted surface patch with the slice contour line includes:

[0024] Construct the parametric equation of the local surface patch using the eigenvectors of the three-dimensional covariance matrix as the normal vectors, calculate the principal curvature and Gaussian curvature of the local surface patch, select the order of the adaptive basis function based on the curvature information, construct a moving least squares equation with a weight function, and introduce the topological continuity constraint and curvature smoothness constraint of the surface into the moving least squares equation to establish a weighted least squares objective function with constraint terms;

[0025] Construct an analytical solution for the weighted least squares objective function, numerically solve the objective function using the domain decomposition method, substitute the solution result into the parametric equation of the local surface patch to obtain the fitted surface, calculate the distance field distribution between the fitted surface and the slice contour line, construct a geometric matching criterion based on the distance field, and use the geometric matching criterion as an evaluation index to screen the optimal fitted surface.

[0026] Further, assemble and splice the surface patches according to the geometric matching result, perform iterative refinement on the assembled and spliced model, establish an error compensation mechanism to eliminate the surface seam, and import the compensated reconstructed model into the three-dimensional environment of the automated production line for collision detection, including:

[0027] Calculate the splicing boundary of adjacent surface patches according to the geometric matching result, construct a transition region at the splicing boundary, establish a surface splicing equation with the geometric parameters of the transition region as the constraint conditions, solve the surface splicing equation using an iterative optimization algorithm to obtain the continuity parameters of the surface, construct an error compensation function based on the continuity parameters, and apply the error compensation function to the surface seam region to eliminate the splicing gap;

[0028] Establish a grid topological structure for the assembled reconstructed model, calculate the geometric quality evaluation index of the grid topological structure, determine the region to be refined based on the geometric quality evaluation index, construct a grid refinement criterion in the refinement region, perform local refinement on the reconstructed model using an adaptive grid refinement algorithm, and convert the refined reconstructed model into a three-dimensional solid model and import it into the automated production line environment.

[0029] In a second aspect, the present application provides an automated production line digital model reconstruction device, including:

[0030] A model slicing module, configured to slice the automated production line digital model along the main coordinate axes, extract a sequence of feature points of the slice contour line at different resolutions, perform geometric semantic annotation on the sequence of feature points, construct a variational level set function from the annotated sequence of feature points, extract a hierarchical feature map from the variational level set function using a deep convolutional network, establish an implicit surface reconstruction framework with physical constraint terms based on the hierarchical feature map, and embed the physical constraint terms as regularization conditions into the implicit surface reconstruction process;

[0031] A space partitioning module, configured to, under the implicit surface reconstruction framework, partition the sequence of feature points using an octree structure, construct a nearest neighbor search algorithm based on the space partitioning, map the nearest neighbor search algorithm into a parallel computing framework to establish an acceleration strategy, construct a local coordinate system according to the nearest neighbor search results, calculate the orthogonal basis vectors of the feature points in the local coordinate system using the principal component analysis method, establish an orthogonal basis space, decompose the feature point projection matrix in the orthogonal basis space into a three-dimensional covariance matrix, and input the three-dimensional covariance matrix into a noise filtering module for data enhancement;

[0032] A model reconstruction module, configured to construct local surface patches based on the eigenvectors of the three-dimensional covariance matrix, perform moving least squares fitting on the local surface patches using an adaptive basis function, add topological constraints and smoothness constraints during the fitting process, geometrically match the fitted surface patches with the slice contour line, assemble and splice the surface patches according to the geometric matching results, iteratively refine the assembled and spliced model, establish an error compensation mechanism to eliminate surface seams, and import the compensated reconstruction model into the three-dimensional environment of the automated production line for collision detection.

[0033] In a third aspect, the present application provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, and when the processor executes the program, the steps of the automated production line digital model reconstruction method are implemented.

[0034] In a fourth aspect, the present application provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the automated production line digital model reconstruction method are implemented.

[0035] In a fifth aspect, the present application provides a computer program product, including a computer program / instructions, and when the computer program / instructions are executed by a processor, the steps of the automated production line digital model reconstruction method are implemented.

[0036] As can be seen from the above technical solutions, this application provides a method and device for reconstructing a digital model of an automated production line. By performing hierarchical slicing and using a deep convolutional network to extract geometric semantic features, an implicit surface reconstruction framework with physical constraints is innovatively constructed. The system uses an octree structure and a parallel computing framework to achieve efficient nearest neighbor search, and combines principal component analysis and covariance matrix decomposition for feature extraction and noise filtering. Through the moving least squares fitting with adaptive basis functions and an error compensation mechanism, the accuracy and continuity of surface reconstruction are ensured, and seamless integration with the collision detection function of the automated production line is achieved. This method breaks through the limitations of traditional model reconstruction and provides an efficient and reliable solution for industrial digital twins. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of this application. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0038] Figure 1 It is one of the flow diagrams of the method for reconstructing the digital model of the automated production line in the embodiments of this application;

[0039] Figure 2 It is another flow diagram of the method for reconstructing the digital model of the automated production line in the embodiments of this application;

[0040] Figure 3 It is yet another flow diagram of the method for reconstructing the digital model of the automated production line in the embodiments of this application;

[0041] Figure 4 It is still another flow diagram of the method for reconstructing the digital model of the automated production line in the embodiments of this application;

[0042] Figure 5 It is another flow diagram of the method for reconstructing the digital model of the automated production line in the embodiments of this application;

[0043] Figure 6 It is yet another flow diagram of the method for reconstructing the digital model of the automated production line in the embodiments of this application;

[0044] Figure 7 It is still another flow diagram of the method for reconstructing the digital model of the automated production line in the embodiments of this application;

[0045] Figure 8 It is the structural diagram of the device for reconstructing the digital model of the automated production line in the embodiments of this application;

[0046] Figure 9Schematic diagram of the structure of the electronic device in the embodiment of the present application.

[0047] Reference numerals:

[0048] Electronic device 9600, central processing unit 9100, memory 9140, communication module 9110, input unit 9120, audio processor 9130, display 9160, power supply 9170, buffer memory 9141, application / function storage unit 9142, data storage unit 9143, driver program storage unit 9144, antenna 9111, speaker 9131, microphone 9132. Detailed implementation manners

[0049] To make the objectives, technical solutions, and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application. Apparently, the described embodiments are some, but not all, of the embodiments of the present application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts shall fall within the protection scope of the present application.

[0050] In the technical solutions of the present application, the acquisition, storage, use, processing, etc. of data all comply with the relevant provisions of national laws and regulations.

[0051] Considering the problems existing in the prior art, the present application provides an automated production line digital model reconstruction method and device. By performing layer slicing and using a deep convolutional network to extract geometric semantic features, an implicit surface reconstruction framework with physical constraints is innovatively constructed. The system adopts an octree structure and a parallel computing framework to achieve efficient nearest neighbor search, and combines principal component analysis and covariance matrix decomposition for feature extraction and noise filtering. Through the moving least squares fitting with adaptive basis functions and an error compensation mechanism, the accuracy and continuity of surface reconstruction are ensured, and seamless integration with the collision detection function of the automated production line is achieved. This method breaks through the limitations of traditional model reconstruction and provides an efficient and reliable solution for industrial digital twins.

[0052] In order to be able to break through the limitations of traditional model reconstruction and provide an efficient and reliable solution for industrial digital twins, the present application provides an embodiment of an automated production line digital model reconstruction method. Refer to Figure 1 The automated production line digital model reconstruction method specifically includes the following contents:

[0053] Step S101: Slice the digital model of the automated production line along the main coordinate axes, extract the sequence of feature points of the slice contour line at different resolutions, perform geometric semantic annotation on the sequence of feature points, construct a variational level set function from the annotated sequence of feature points, extract hierarchical feature maps from the variational level set function using a deep convolutional network, establish an implicit surface reconstruction framework with physical constraint terms based on the hierarchical feature maps, and embed the physical constraint terms as regularization conditions into the implicit surface reconstruction process;

[0054] Optionally, in the process of slicing the digital model in this embodiment, an adaptive slicing strategy is adopted. For different types of equipment models in the automated production line, the optimal slicing direction is determined by a three-dimensional principal direction analysis algorithm. First, calculate the principal curvature distribution of the surface mesh of the model, construct a curvature tensor field, and determine the local coordinate system based on the eigenvectors of the curvature tensor. For the joint structure of the robotic arm, the slicing direction parallel to the joint rotation axis is preferentially selected; for the conveyor belt system, the slicing direction is determined according to the normal vector of the conveying plane; for regular equipment such as control cabinets, the slicing direction parallel to the main boundary surface is selected.

[0055] In this embodiment, when determining the slice spacing, a spacing adjustment algorithm based on geometric features is developed. By analyzing the local curvature change rate and shape complexity of the model, an adaptive spacing function is constructed. When processing precision components such as the end effector of the robotic arm, when the local curvature exceeds a preset threshold, the system automatically reduces the slice spacing to improve the sampling accuracy; while for gradually changing structures such as conveyor belts, the slice spacing is appropriately increased to optimize the calculation efficiency.

[0056] In the link of extracting the slice contour line in this embodiment, a multi-resolution feature sampling mechanism is realized. First, construct a quadtree structure to partition the space of the slice plane, and calculate the curvature and gradient information in each leaf node. For areas with high-precision requirements such as the end effector of the robotic arm, the system automatically subdivides the grid until the preset precision requirement is met; for the planar area of the control cabinet, a larger grid size is maintained to reduce data redundancy.

[0057] In the process of geometric semantic annotation in this embodiment, a classification algorithm based on local features is developed. By constructing the local neighborhood of the feature points, geometric descriptors including principal curvature, Gaussian curvature, mean curvature, etc. are calculated. For the moving joints of the robotic arm, the axial features and mating surface features are mainly annotated; for the guiding structure of the conveyor belt, the boundary features and transition features are emphasized; for the mounting surface of the control cabinet, the flatness and perpendicularity features are concerned.

[0058] In this embodiment, when constructing the variational level set function, a target functional with weights is adopted. The target functional consists of two parts: a data term and a regularization term. The data term measures the distance between the reconstructed surface and the sampling points, and the regularization term controls the smoothness of the surface. When processing precision mating parts such as robotic arm flanges, the reconstruction accuracy is ensured by increasing the weight of the data term; when processing the transition area of the conveyor belt, the weight of the regularization term is appropriately increased to ensure the smooth transition of the surface.

[0059] In the design of the deep convolutional network in this embodiment, a multi-scale feature extraction mechanism is implemented. The network structure adopts the U-Net architecture, which includes 5 layers of encoders and 5 layers of decoders. The encoder gradually extracts features through 3×3 convolutional layers and max pooling layers, and the decoder restores the spatial resolution through transposed convolutional layers. The input data is the normalized level set function value, and the intermediate features are fused with different scale information through skip connections. The output is a 32-channel feature map, and each channel corresponds to different scale geometric features.

[0060] In the post-processing of the feature map in this embodiment, a feature fusion strategy enhanced by an attention mechanism is developed. By calculating the correlation matrix between features at different levels, the weight distribution of the features is adaptively adjusted. For complex structures such as robotic arm bearing seats, the system can accurately identify and retain key geometric features; for continuous features such as conveyor belt guide grooves, the continuity of the surface is emphasized.

[0061] In the construction process of the physical constraint term in this embodiment, a constraint model based on material mechanics is adopted. The constraint conditions include the bending energy, tensile energy, and torsional energy of the surface, and the physical rationality of the reconstruction result is ensured by minimizing the total energy. When processing load-bearing structures such as robotic arm supports, the stress distribution characteristics of the material are particularly considered; when processing conveyor belt brackets, it is ensured that the structure meets the stiffness requirements.

[0062] In the embedding of the regularization condition in this embodiment, a weight adaptive mechanism based on geometric features is implemented. By analyzing the change rate of the local curvature and the complexity of the features, a dynamic weight function is constructed. When processing smooth areas such as robotic arm fillet transitions, the weight of the smoothing term is increased; when processing feature areas such as sensor mounting grooves, the weight of the feature-preserving term is increased.

[0063] Through the above implementation scheme, this embodiment constructs a complete model reconstruction framework. This scheme shows excellent performance in practical applications and can accurately reconstruct the models of various automated production line equipment including robotic arms, conveyor belts, control cabinets, etc. The reconstruction results not only maintain geometric accuracy but also ensure the rationality of physical properties, providing a reliable digital model basis for subsequent dynamic simulation and path planning.

[0064] In the digital upgrade project of the production line, the solution of this embodiment has been successfully applied to the equipment modeling of multiple automated production lines. Through precise geometric reconstruction and reasonable physical constraints, the reliability of the digital twin model has been significantly improved, effectively supporting the optimization design and transformation and upgrading of the production line.

[0065] Step S102: Under the implicit surface reconstruction framework, use an octree structure to partition the feature point sequence in space, construct a nearest neighbor search algorithm based on the spatial partition, map the nearest neighbor search algorithm into a parallel computing framework to establish an acceleration strategy, construct a local coordinate system according to the nearest neighbor search result, use the principal component analysis method to calculate the orthogonal basis vectors of the feature points in the local coordinate system, establish an orthogonal basis space, decompose the feature point projection matrix in the orthogonal basis space into a three-dimensional covariance matrix, and input the three-dimensional covariance matrix into a noise filtering module for data enhancement;

[0066] Optionally, in the process of constructing the octree structure in this embodiment, an adaptive spatial partition strategy is adopted. First, calculate the spatial bounding box of the feature point sequence, and determine the initial division scale by analyzing the point cloud density distribution. For complex structure areas such as robotic arm joints, when the local point density exceeds a preset threshold, the system automatically subdivides the spatial unit until a predetermined spatial resolution is reached; for simple feature areas such as the conveyor belt body, a larger spatial unit size is maintained to improve the calculation efficiency.

[0067] In the construction of the spatial index in this embodiment, a dynamic balance tree structure optimization algorithm is developed. By calculating the spatial distribution characteristics of the feature points in each node, including the point density change rate and direction distribution, etc., the splitting and merging strategies of the nodes are adaptively adjusted. When processing precision components such as the end effector of the robotic arm, the system dynamically adjusts the depth of the tree according to the complexity of the local geometric features; for regular structures such as the control cabinet shell, a relatively uniform tree layer structure is maintained.

[0068] In the nearest neighbor search process of this embodiment, a multi-threaded parallel search mechanism is implemented. First, divide the feature points into multiple spatially continuous data blocks, and each data block is assigned to an independent computing thread. For the moving joints of the robotic arm, focus on analyzing the topological structure of the local neighborhood; for the guiding structure of the conveyor belt, pay attention to the continuous distribution characteristics of the feature points. GPU acceleration is achieved through the CUDA architecture, significantly improving the search efficiency.

[0069] In the local coordinate system construction link of this embodiment, a coordinate system optimization strategy based on geometric features is adopted. By analyzing the spatial distribution of the nearest neighbor point set, calculate the centroid position and main direction of the point set. For the mating surface of the robotic arm, preferentially select the coordinate axes that are consistent with the assembly direction; for the moving surface of the conveyor belt, align the coordinate axes with the moving direction. The system ensures the stability of the coordinate system through iterative optimization.

[0070] In this embodiment, a robust feature extraction algorithm is developed in the principal component analysis calculation. By constructing the divergence matrix of neighboring points and using SVD decomposition to obtain the principal direction vector. When dealing with precision mating structures such as the flange of the robotic arm, the reliability of the principal direction is determined by analyzing the eigenvalue distribution; for planar features such as the sensor mounting surface, the accuracy of the normal vector is emphasized.

[0071] In this embodiment, an iterative optimization orthogonalization method is adopted in the calculation process of orthogonal basis vectors. By analyzing the angle relationship between the principal direction vectors, an orthogonalization constraint condition is constructed. For complex structures such as the bearing seat of the robotic arm, the system ensures the orthogonality of the basis vectors through multiple iterative optimizations; for regular structures such as the control cabinet mounting surface, the orthogonal basis is directly constructed using the principal direction.

[0072] In this embodiment, a projection strategy based on feature weights is implemented when constructing the projection matrix. By analyzing the distribution characteristics of feature points in different directions, a weight coefficient is assigned to each projection direction. When dealing with smooth regions such as the fillet transition of the robotic arm, the weight of the tangential component is increased; when dealing with feature regions such as the sensor mounting groove, the weight of the normal component is increased.

[0073] In this embodiment, an adaptive feature analysis algorithm is developed in the calculation link of the three-dimensional covariance matrix. By analyzing the distribution law of eigenvalues, the geometric feature type of the local area is identified. For the special-shaped surface of the robotic arm, the principal curvature characteristics of the surface are judged by the eigenvalue ratio; for the guiding groove of the conveyor belt, the significance of the feature is determined by the eigenvalue size.

[0074] In this embodiment, an adaptive filtering strategy based on geometric features is adopted during the noise filtering process. By analyzing the eigenvalue distribution of the covariance matrix, an anisotropic filtering kernel function is constructed. When dealing with precision parts of the robotic arm, a smaller filtering radius is used to retain detailed features; when dealing with the conveyor belt body, a larger filtering radius is used to improve the smoothing effect.

[0075] Through the above implementation scheme, this embodiment constructs an efficient and stable feature analysis framework. This scheme shows excellent performance in practical applications, and can accurately extract and enhance the geometric features of various automated production line equipment. The acceleration strategy based on parallel computing significantly improves the processing efficiency and provides strong support for the digital reconstruction of large-scale production line equipment.

[0076] In the digital twin project of the production line, the scheme of this embodiment is successfully applied to the feature extraction and optimization of complex industrial scenarios. Through precise geometric analysis and efficient data processing, the quality of the reconstructed model is significantly improved, laying a solid foundation for subsequent simulation analysis and optimization design.

[0077] Step S103: Construct local surface patches based on the eigenvectors of the three-dimensional covariance matrix, perform moving least squares fitting on the local surface patches using adaptive basis functions, add topological constraints and smoothness constraints during the fitting process, geometrically match the fitted surface patches with the slice contour lines, assemble and splice the surface patches according to the geometric matching results, iteratively refine the assembled and spliced model, establish an error compensation mechanism to eliminate surface seams, and import the compensated reconstructed model into the three-dimensional environment of the automated production line for collision detection.

[0078] Optionally, in the process of constructing local surface patches in this embodiment, the main direction of the surface is determined based on the eigenvectors of the three-dimensional covariance matrix. For different functional components in the automated production line, such as the joint connection of the robotic arm, the guiding groove of the conveyor belt, and the mounting surface of the control cabinet, the local parameterization method of the surface is determined by analyzing the direction distribution of the eigenvectors. For the spherical joint of the robotic arm, spherical coordinate parameterization is used; for the cylindrical structure of the conveyor belt, cylindrical coordinate parameterization is used; for the planar structure of the control cabinet, rectangular coordinate parameterization is used.

[0079] When selecting the adaptive basis function in this embodiment, a construction strategy of the basis function based on geometric features is developed. By analyzing the curvature distribution and topological features of the local surface, the type and order of the basis function are dynamically adjusted. When dealing with complex surfaces such as the end effector of the robotic arm, high-order B-spline basis functions are used to ensure the fitting accuracy; for the regular surface of the conveyor belt, low-order polynomial basis functions are selected to improve the calculation efficiency.

[0080] In the process of moving least squares fitting in this embodiment, an adaptive weight allocation mechanism is realized. By constructing a distance-based weight function, different influence factors are assigned to the sampling points in the neighborhood. When dealing with the precision mating surface of the robotic arm, the weight of the near neighbor points is increased to ensure the local fitting accuracy; when dealing with the transition region of the conveyor belt, a larger influence radius is used to ensure the smooth transition of the surface.

[0081] In the link of constructing topological constraints in this embodiment, a constraint model based on boundary matching is developed. By analyzing the connection relationship between adjacent surface patches, position continuity and tangential continuity constraints are established. For the complex assembly surface of the robotic arm, ensure the precise docking of adjacent surface patches; for the guiding structure of the conveyor belt, ensure the continuous transition of the surface.

[0082] In the design of smoothness constraints in this embodiment, a multi-objective optimization strategy is adopted. By comprehensively considering the mean curvature, Gaussian curvature and torsional energy of the surface, a smoothness evaluation function is constructed. During the optimization process, the constraint weights are dynamically adjusted according to the functional requirements of different regions to ensure the geometric quality of the surface.

[0083] In this embodiment, during the geometric matching process, a feature-based matching algorithm is implemented. By analyzing the geometric features of surface patches and slice contour lines, feature descriptors and similarity metrics are constructed. For precision parts of the robotic arm, strict matching criteria are adopted; for conventional components of the conveyor belt, the matching conditions are appropriately relaxed to improve the splicing efficiency.

[0084] In this embodiment, during the surface patch splicing process, an adaptive transition region generation algorithm is developed. By analyzing the geometric features of adjacent surface patches, the range and shape of the transition region are dynamically determined. When dealing with the key joints of the robotic arm, a narrower transition region is constructed to ensure precision; when dealing with general transitions of the conveyor belt, a wider transition region is adopted to ensure smoothness.

[0085] In this embodiment, during the model iteration and refinement process, a mesh optimization strategy based on error analysis is adopted. By calculating the distance field between the reconstructed model and the original data, regions that need to be refined are identified. For precision parts of the robotic arm, local mesh encryption is performed; for flat areas of the conveyor belt, a moderate mesh density is maintained.

[0086] In this embodiment, during the design of the error compensation mechanism, an adaptive compensation algorithm is implemented. By analyzing the geometric discontinuities at the surface seams, a local deformation field is constructed. During the compensation process, material properties and mechanical constraints are considered to ensure the physical rationality of the compensation results.

[0087] Through the above implementation solutions, this embodiment establishes a complete surface reconstruction and optimization framework. This solution can accurately reconstruct various functional components in the automated production line and ensure that the reconstruction results meet the actual application requirements. In the collision detection link, by establishing an efficient spatial index structure, fast interference analysis is achieved, providing a reliable geometric basis for the layout optimization and path planning of the production line.

[0088] In practical applications, the solution of this embodiment has successfully handled various complex production line equipment reconstruction problems, such as special-shaped joints of robotic arms and guiding structures of conveyor belts, significantly improving the accuracy and usability of digital models, and effectively supporting the intelligent upgrading and transformation of production lines.

[0089] As can be seen from the above description, the automated production line digital model reconstruction method provided by the embodiments of this application can extract geometric semantic features through layer-by-layer slicing and deep convolutional networks, and innovatively constructs an implicit surface reconstruction framework with physical constraints. The system uses an octree structure and a parallel computing framework to achieve efficient nearest neighbor search, and combines principal component analysis and covariance matrix decomposition for feature extraction and noise filtering. Through the moving least squares fitting of adaptive basis functions and the error compensation mechanism, the accuracy and continuity of surface reconstruction are ensured, and seamless integration with the collision detection function of the automated production line is achieved. This method breaks through the limitations of traditional model reconstruction and provides an efficient and reliable solution for industrial digital twins.

[0090] In an embodiment of the method for reconstructing the digital model of the automated production line of the present application, refer to Figure 2 , and it may specifically include the following content:

[0091] Step S201: Calculate the slicing direction of the main coordinate axis by using the hierarchical scanning algorithm of the three-dimensional grid model. Based on the slicing direction, construct a slicing plane parallel to the coordinate axis. Intersect the slicing plane with the boundary surface of the digital model to generate a slicing contour line. Perform grid division on the slicing contour line to calculate the spatial coordinates of the feature points, and establish an adaptive resolution sampling mechanism to screen the feature points to obtain an ordered sequence of feature points;

[0092] Step S202: Establish a curvature tensor analysis model for the sequence of feature points. Calculate the geometric attributes of the feature points based on the main direction of the curvature tensor. Assign the geometric attributes as label information to the feature points. Construct a variational objective functional with weight coefficients, and obtain the mathematical expression of the level set function by solving the variational objective functional. Establish a mapping relationship from the sequence of feature points to the level set function.

[0093] Optionally, in the process of analyzing the three-dimensional grid model in this embodiment, a hierarchical scanning strategy is adopted to determine the optimal slicing direction. For the robotic arm in the automated production line, first analyze its joint motion characteristics, and identify the main motion axes by calculating the normal vector distribution of the local surface; for the conveyor belt system, determine the slicing direction based on the conveying direction and the installation reference plane; for equipment such as control cabinets, determine the appropriate slicing direction by analyzing the boundary characteristics of its main structure.

[0094] In the construction of the slicing plane in this embodiment, an adaptive slicing algorithm based on geometric features is developed. By analyzing the geometric complexity of the model in different directions, the slicing spacing is dynamically adjusted. When processing precision components such as the end effector of the robotic arm, a smaller slicing spacing is adopted to ensure the sampling accuracy; for regular structures such as the conveyor belt body, a larger slicing spacing is used to improve the calculation efficiency.

[0095] In the link of generating the slicing contour line in this embodiment, an accurate intersection calculation mechanism is realized. The intersection detection is accelerated by constructing a spatial bounding box, and an accurate numerical algorithm is used to calculate the intersection line of the slicing plane and the model boundary surface. For the special-shaped surface of the robotic arm, special attention is paid to the contour features of the surface; for the guiding structure of the conveyor belt, the key geometric features are mainly retained.

[0096] In the grid division process of this embodiment, an adaptive grid generation strategy is adopted. By analyzing the curvature change of the contour line, grid nodes are encrypted in the high-curvature area, and the grid density is appropriately reduced in the flat area. For the precision mating surface of the robotic arm, high-density division is used to ensure the reconstruction accuracy; for the planar structure of the control cabinet, coarser grid division is used to improve efficiency.

[0097] In the feature point screening process of this embodiment, a multi-criterion evaluation mechanism is developed. By comprehensively considering the spatial distribution, curvature characteristics, and geometric importance of points, a priority ranking of feature points is established. When dealing with the complex curved surface at the robotic arm joint, more feature points are retained; for the regular structure of the conveyor belt, the number of sampling points is appropriately reduced.

[0098] In the process of curvature tensor analysis of this embodiment, a robust curvature estimation algorithm is implemented. By constructing a local quadratic surface to fit the neighborhood of feature points, the principal curvature and principal direction are calculated. For the spherical joint of the robotic arm, the spherical feature is accurately extracted; for the cylindrical structure of the conveyor belt, the bus direction is accurately calculated.

[0099] In the geometric attribute calculation process of this embodiment, a multi-scale feature extraction strategy is adopted. By analyzing the variation law of the curvature tensor at different scales, a hierarchical geometric descriptor is constructed. For the precision parts of the robotic arm, geometric features at the detail level are extracted; for the conventional components of the conveyor belt, macroscopic geometric features are concerned.

[0100] In the construction of the variational objective functional of this embodiment, a weight assignment scheme based on physical meaning is developed. By analyzing the importance of different geometric features, reasonable weight coefficients are assigned to each item of the objective functional. When dealing with the functional surface of the robotic arm, the weight of the conformal term is increased; when dealing with the transition region of the conveyor belt, the weight of the smoothing term is increased.

[0101] In the process of solving the level set function of this embodiment, an efficient numerical calculation method is implemented. By constructing a multi-grid solver, an adaptive time step is used to control the convergence of the solution. For the complex structure of the robotic arm, a fine solution grid is used; for the simple structure of the conveyor belt, a coarser solution grid is used.

[0102] In the establishment of the feature mapping of this embodiment, a non-linear mapping strategy is adopted. By analyzing the distribution characteristics of feature points, a highly adaptable mapping function is constructed. When dealing with the irregular surface of the robotic arm, a locally adaptive mapping method is adopted; for the regular surface of the conveyor belt, a globally unified mapping method is used.

[0103] Through the above implementation solution, this embodiment constructs a complete geometric feature extraction and analysis framework. This solution can accurately capture various geometric features of automated production line equipment and establish reasonable mathematical expressions. In practical applications, this solution has successfully processed various equipment models including robotic arms, conveyor belts, control cabinets, etc., providing reliable feature data support for subsequent reconstruction processes.

[0104] In the digital upgrade project of the production line, the solution of this embodiment significantly improves the accuracy and efficiency of geometric feature extraction, lays a solid foundation for the precise modeling of production line equipment, and effectively supports subsequent simulation analysis and optimization design work.

[0105] In an embodiment of the method for reconstructing the digital model of an automated production line of this application, refer to Figure 3 , it may specifically include the following content:

[0106] Step S301: Construct a multi-scale convolutional neural network structure. Set convolutional kernels of different sizes in the convolutional layer to extract the spatial features of the variational level set function. Perform dimensionality reduction and compression on the feature map through pooling operations. Input the dimensionality-reduced feature map into the deconvolution layer for upsampling. Concatenate and fuse the upsampling result with the feature map of the skip connection to generate a hierarchical feature map;

[0107] Step S302: Based on the hierarchical feature map, construct an implicit surface equation. Add physical constraint terms for surface smoothness and shape preservation to the implicit surface equation. Convert the physical constraint terms into regularization conditions in the form of Lagrange multipliers, establish an objective function with regularization terms, and use the variational method to solve the objective function to obtain the expression for implicit surface reconstruction.

[0108] Optionally, in the process of constructing the multi-scale convolutional neural network in this embodiment, a hierarchical feature extraction strategy is adopted. For the geometric features of different types of equipment in the automated production line, a U-Net structure including five-layer encoders and five-layer decoders is designed. In the encoder, through the combination of convolutional kernels of three sizes: 3×3, 5×5, and 7×7, the local details of the precision parts of the robotic arm, the medium-scale features of the conveyor belt guiding structure, and the large-scale geometric features of the control cabinet shell are captured respectively.

[0109] In the design of the convolutional layer of this embodiment, an adaptive feature extraction mechanism is realized. By setting deformable convolutional modules at different levels, the adaptability of the network to complex geometric shapes is enhanced. When dealing with the complex curved surface at the robotic arm joint, the deformable convolution can automatically adjust the shape of the receptive field and accurately capture local features; for the regular structure of the conveyor belt, a standard rectangular receptive field is maintained to improve the calculation efficiency.

[0110] In the pooling operation design of this embodiment, an adaptive pooling strategy based on geometric features is developed. By analyzing the geometric information distribution in the feature map, different pooling strategies are adopted in different regions. For the precision mating surfaces of the robotic arm, a smaller pooling window is used to retain detailed information; for the planar structure of the control cabinet, a larger pooling window is adopted to reduce the computational complexity.

[0111] In the implementation of the transposed convolution layer in this embodiment, a multi-stage feature recovery mechanism is adopted. By designing a learnable upsampling kernel, geometric details are maintained during the feature recovery process. When reconstructing precision components such as the end effector of the robotic arm, a fine upsampling strategy is adopted; for conventional structures such as the conveyor belt body, standard bilinear interpolation is used to improve efficiency.

[0112] In the feature fusion stage of this embodiment, a feature integration strategy enhanced by the attention mechanism is implemented. By calculating the correlation matrix between features at different levels, the weight distribution of features is adaptively adjusted. For the irregular surfaces of the robotic arm, local geometric features are emphasized; for the continuous curved surfaces of the conveyor belt, the global shape features are emphasized.

[0113] When constructing the implicit surface equation in this embodiment, a constraint model based on physical laws is developed. By introducing surface energy functionals, including bending energy, stretching energy, and torsion energy, the physical rationality of the reconstruction result is ensured. When dealing with load-bearing structures such as the support base of the robotic arm, the mechanical properties of the material are particularly considered; when dealing with the conveyor belt support, it is ensured that the structure meets the stiffness requirements.

[0114] In the process of converting physical constraints in this embodiment, an adaptive constraint weight adjustment strategy is adopted. By analyzing the complexity of local geometric features, the weight coefficients of constraint terms are dynamically adjusted. For the functional surfaces of the robotic arm, the weight of the shape preservation term is increased; for the transition regions of the conveyor belt, the weight of the smoothness constraint is increased.

[0115] In the construction of the regularization conditions in this embodiment, a constraint framework for multi-objective optimization is implemented. By constructing constraint conditions in the form of Lagrange multipliers, the reconstruction accuracy and surface quality are balanced. When dealing with precision parts of the robotic arm, strong regularization constraints are adopted to ensure surface quality; for conventional components of the conveyor belt, the constraints are appropriately relaxed to improve the reconstruction efficiency.

[0116] In the process of solving the objective function in this embodiment, an iterative optimization solution strategy is developed. By constructing a gradient descent optimizer and adopting an adaptive step size control algorithm, the convergence of the solution is ensured. For the complex structures of the robotic arm, a smaller optimization step size is used to ensure accuracy; for the simple structures of the conveyor belt, a larger step size is adopted to accelerate convergence.

[0117] Through the above implementation solution, this embodiment constructs a complete deep learning reconstruction framework. This solution can accurately reconstruct various device models in the automated production line and ensure that the reconstruction results meet the requirements of engineering applications. In practical applications, this solution has successfully processed the reconstruction tasks of various complex devices including robotic arms, conveyor belts, control cabinets, etc., significantly improving the accuracy and reliability of the digital model.

[0118] In the project of digital upgrading of the production line, the solution of this embodiment effectively improves the quality and efficiency of model reconstruction through the combination of deep learning and physical constraints, provides a reliable geometric basis for the intelligent transformation of the production line, and supports subsequent simulation optimization and intelligent control work.

[0119] In an embodiment of the method for reconstructing the digital model of the automated production line of this application, refer to Figure 4 , it may specifically include the following content:

[0120] Step S401: Calculate the spatial bounding box of the feature point sequence, divide the spatial bounding box into eight sub-spaces according to the coordinate axis directions, calculate the density distribution of feature points for each sub-space, recursively subdivide the sub-spaces based on the density distribution threshold, establish an octree index structure to store the spatial position information of feature points, and record the spatial position information of the feature points into the corresponding leaf nodes;

[0121] Step S402: Construct a nearest neighbor search path in the octree index structure, calculate the spatial distance between the feature point to be searched and the leaf node, mark the leaf nodes with a spatial distance less than the search radius as candidate nodes, establish a parallel search task for the candidate nodes, allocate the parallel search task to the computing unit to perform the nearest neighbor search calculation, and summarize the search results to obtain the neighborhood information of the feature points.

[0122] Optionally, in the process of calculating the spatial bounding box in this embodiment, an adaptive boundary analysis strategy is adopted. For different types of device models in the automated production line, first calculate the spatial range of the feature point set. For complex structures such as robotic arms, determine the boundaries of the bounding box by analyzing the joint motion space; for conveyor belt systems, determine the spatial range based on the conveying path and guiding structure; for regular devices such as control cabinets, directly use the main body contour to determine the boundaries.

[0123] In this embodiment, an adaptive partitioning mechanism based on geometric features is implemented during subspace partitioning. By analyzing the distribution characteristics of feature points in each region, the partitioning strategy is dynamically adjusted. When processing precision components such as the end effector of a robotic arm, a smaller partitioning scale is used to accurately capture local features; for regular structures such as the conveyor belt body, a larger partitioning scale is used to improve efficiency.

[0124] In the density distribution calculation section of this embodiment, a multi-scale density estimation algorithm is developed. By constructing a kernel density estimator, the aggregation degree of points is analyzed at different spatial scales. For the complex curved surface of the robotic arm, a smaller kernel function radius is adopted to improve the local accuracy of density estimation; for the planar structure of the control cabinet, a larger kernel function radius is used to reduce the computational cost.

[0125] In the recursive subdivision process of this embodiment, an adaptive termination criterion is implemented. By analyzing the geometric feature complexity of the subspace, the subdivision depth is dynamically adjusted. When dealing with the precision mating surface of the robotic arm, deeper recursive levels are allowed; when dealing with the regular structure of the conveyor belt, the recursive depth is appropriately controlled to balance accuracy and efficiency.

[0126] In the octree construction section of this embodiment, a hierarchical node management strategy is adopted. By designing an efficient parent-child node link structure, the storage and access efficiency of the tree is optimized. For the irregular surface of the robotic arm, local feature information is mainly saved; for the continuous curved surface of the conveyor belt, the overall shape features are mainly recorded.

[0127] When constructing the nearest neighbor search path in this embodiment, a heuristic search strategy is developed. By analyzing the spatial relationship between the target point and the node center, a priority queue is constructed to guide the search order. When dealing with the complex structure at the robotic arm joint, the area with high geometric similarity is preferentially searched; for the regular structure of the conveyor belt, a simple distance-first criterion is adopted.

[0128] In the spatial distance calculation process of this embodiment, an adaptive metric mechanism is implemented. By considering the geometric characteristics of the local surface, an anisotropic distance metric is constructed. For the spherical joint of the robotic arm, a spherical distance metric is adopted; for the cylindrical structure of the conveyor belt, a cylindrical distance metric is used.

[0129] In the parallel task assignment section of this embodiment, a dynamic load balancing strategy is adopted. By analyzing the computational complexity of the candidate nodes, the computing resources are reasonably allocated. When dealing with the precision parts of the robotic arm, more computing units are allocated to the high-complexity areas; for the conventional components of the conveyor belt, a uniform allocation strategy is adopted.

[0130] In the search result summarization process of this embodiment, a filtering mechanism based on feature similarity is developed. By analyzing the geometric features of the neighborhood points, unreasonable search results are eliminated. For the functional surface of the robotic arm, the geometric consistency of the neighborhood points is strictly controlled; for the transition area of the conveyor belt, the screening conditions are appropriately relaxed.

[0131] Through the above implementation solutions, this embodiment constructs an efficient and stable spatial index structure and a nearest neighbor search framework. This solution can quickly and accurately identify the local feature associations in the automated production line equipment model, providing reliable neighborhood information support for subsequent geometric reconstruction.

[0132] In the digital upgrade project of the production line, the solution of this embodiment significantly improves the efficiency and accuracy of feature analysis. Through the combination of parallel computing and intelligent indexing, it effectively supports the digital reconstruction of large-scale production line equipment, providing important technical support for the intelligent transformation of the production line.

[0133] In an embodiment of the method for reconstructing the digital model of an automated production line in this application, referring to Figure 5 , it may specifically include the following content:

[0134] Step S501: Calculate the centroid coordinates of the feature point neighborhood using the nearest neighbor search result, translate the centroid coordinates to the origin to establish the origin of the local coordinate system, calculate the divergence matrix of the feature point neighborhood, perform eigenvalue decomposition on the divergence matrix to obtain the eigenvector matrix, sort the eigenvector matrix according to the eigenvalue magnitude to determine the three basis vectors of the local coordinate system, and construct the rotation matrix of the local coordinate system based on the basis vectors;

[0135] Step S502: Transform the feature point from the global coordinate system to the local coordinate system, calculate the coordinate components of the feature point in the local coordinate system, construct the projection matrix of the feature point, perform singular value decomposition on the projection matrix to obtain the three-dimensional covariance matrix, calculate the noise threshold based on the eigenvalue distribution of the three-dimensional covariance matrix, and input the noise threshold as the filtering parameter into the anisotropic diffusion filter.

[0136] Optionally, in the process of calculating the centroid coordinates in this embodiment, a weighted average calculation strategy is adopted. For the feature point distributions of different types of equipment in the automated production line, the weight coefficients are determined by analyzing the geometric importance of the neighborhood points. For the precision mating surface of the robotic arm, different weights are assigned according to the curvature change; for the guiding structure of the conveyor belt, the weights are determined based on the spatial distribution characteristics of the points; for the planar structure of the control cabinet, uniform weights are used to simplify the calculation.

[0137] When establishing the local coordinate system in this embodiment, a coordinate system optimization mechanism based on geometric features is realized. By analyzing the distribution characteristics of the neighborhood points, it is ensured that the coordinate axis directions correspond to the physical features. When dealing with the spherical surface at the robotic arm joint, the main axis is aligned with the center of the sphere direction; for the cylindrical surface structure of the conveyor belt, the main axis is aligned with the generatrix direction.

[0138] In the divergence matrix calculation link of this embodiment, a robust feature extraction algorithm is developed. By constructing a high-order matrix to describe the spatial distribution of the neighborhood points, the local geometric features are accurately captured. For the irregular surface of the robotic arm, high-order divergence features are calculated; for the regular surface of the conveyor belt, low-order divergence is used for description.

[0139] In this embodiment, an adaptive numerical calculation strategy is implemented during the eigenvalue decomposition process. By analyzing the condition number of the matrix, a suitable decomposition algorithm is selected. When processing precision parts of the robotic arm, a high-precision iterative algorithm is adopted; for the conventional structure of the conveyor belt, a fast direct solution method is used.

[0140] In the basis vector sorting step of this embodiment, a multi-criterion evaluation mechanism is adopted. By comprehensively considering the eigenvalue magnitude and geometric meaning, the priority of the basis vectors is determined. For the functional surface of the robotic arm, special attention is paid to the accuracy of the normal vector; for the guiding structure of the conveyor belt, emphasis is placed on ensuring the direction of the tangent vector.

[0141] When constructing the rotation matrix in this embodiment, an optimization strategy based on physical constraints is developed. By introducing orthogonality and right-hand system constraints, the rationality of the coordinate transformation is ensured. When dealing with the assembly surface of the robotic arm, the coordinate axes are guaranteed to be consistent with the assembly direction; for the moving surface of the conveyor belt, the coordinate axes are made to correspond to the moving direction.

[0142] In the coordinate transformation process of this embodiment, a high-precision numerical calculation mechanism is implemented. By designing a stable transformation algorithm, the accumulation of numerical errors is reduced. For precision components of the robotic arm, double-precision calculation is used; for the conventional structure of the conveyor belt, single-precision calculation is used to improve efficiency.

[0143] In the projection matrix construction step of this embodiment, an adaptive projection strategy is adopted. By analyzing the geometric characteristics of the local surface, the optimal projection direction is determined. For the spherical joint of the robotic arm, spherical projection is used; for the cylindrical structure of the conveyor belt, cylindrical projection is adopted.

[0144] In the singular value decomposition process of this embodiment, a robust calculation method is developed. By designing an iterative optimization strategy, the accuracy of the decomposition result is ensured. When dealing with complex surfaces of the robotic arm, an accurate solution algorithm is adopted; for simple surfaces of the conveyor belt, a fast approximation algorithm is used.

[0145] In the noise threshold calculation of this embodiment, an adaptive parameter estimation mechanism is implemented. By analyzing the eigenvalue distribution of the covariance matrix, the filtering intensity is dynamically determined. For precision parts of the robotic arm, a smaller filtering threshold is used to retain details; for conventional components of the conveyor belt, a larger threshold is used to improve the smoothing effect.

[0146] Through the above implementation scheme, this embodiment constructs a complete local feature analysis framework. This scheme can accurately extract the local geometric features of the automated production line equipment model and effectively suppress the influence of noise through anisotropic filtering. In practical applications, this scheme has successfully processed various complex equipment models including robotic arms, conveyor belts, control cabinets, etc., significantly improving the quality of geometric feature extraction.

[0147] In the digital upgrade project of the production line, the solution of this embodiment provides reliable feature support for the accurate modeling of production line equipment through precise geometric analysis and effective noise suppression, promoting the intelligent transformation process of the production line.

[0148] In an embodiment of the method for reconstructing the digital model of an automated production line of the present application, referring to Figure 6 , the following specific content may also be included:

[0149] Step S601: Use the eigenvectors of the three-dimensional covariance matrix as normal vectors to construct the parametric equation of the local surface patch, calculate the principal curvature and Gaussian curvature of the local surface patch, select the order of the adaptive basis function based on the curvature information, construct the moving least squares equation with a weight function, introduce the topological continuity constraint and curvature smoothness constraint of the surface into the moving least squares equation, and establish a weighted least squares objective function with constraint terms;

[0150] Step S602: Construct an analytical solution for the weighted least squares objective function, numerically solve the objective function using the domain decomposition method, substitute the solution result into the parametric equation of the local surface patch to obtain the fitted surface, calculate the distance field distribution between the fitted surface and the slice contour line, construct a geometric matching criterion based on the distance field, and use the geometric matching criterion as an evaluation index to screen the optimal fitted surface.

[0151] Optionally, in the process of constructing the local surface patch in this embodiment, the principal direction of the surface is determined based on the eigenvectors of the three-dimensional covariance matrix. For the robotic arm joints in the automated production line, the eigenvector corresponding to the largest eigenvalue is selected as the normal vector to ensure that the surface is adapted to the motion characteristics; for the guiding structure of the conveyor belt, the tangential direction is determined based on the eigenvector corresponding to the second largest eigenvalue to ensure consistency with the conveying direction.

[0152] In the curvature calculation link of this embodiment, a multi-scale curvature estimation strategy is adopted. By calculating the principal curvature and Gaussian curvature in different neighborhood ranges, local geometric features are accurately captured. For the spherical joints of the robotic arm, the uniformity of the curvature is emphasized; for the transition region of the conveyor belt, the change trend of the curvature is mainly analyzed.

[0153] When selecting the basis function in this embodiment, an adaptive order adjustment mechanism is implemented. By analyzing the complexity of the local surface, the type and order of the basis function are dynamically determined. When dealing with the precision mating surface of the robotic arm, high-order B-spline basis functions are used to ensure the fitting accuracy; for the planar structure of the control cabinet, low-order polynomial basis functions are used to improve efficiency.

[0154] In the process of constructing the weight function in this embodiment, a weight assignment strategy based on geometric features is developed. By analyzing the spatial distribution and geometric importance of points, reasonable weight coefficients are determined. For the irregular surfaces of the robotic arm, the weights of key feature points are increased; for the regular surfaces of the conveyor belt, a distance attenuation-based weight assignment is adopted.

[0155] In the topological constraint design of this embodiment, multi-level continuity requirements are adopted. By establishing position continuity and tangential continuity constraints, smooth transitions between adjacent surface patches are ensured. When dealing with the assembly surfaces of the robotic arm, topological constraints are strengthened to guarantee assembly accuracy; for the connection parts of the conveyor belt, the constraints are appropriately relaxed to improve modeling efficiency.

[0156] In the construction of the fairness constraint in this embodiment, a constraint mechanism based on physical meaning is realized. By introducing the surface energy functional, the fairness degree of the surface is controlled. When dealing with the functional surfaces of the robotic arm, the fairness constraint is increased to ensure surface quality; for the conventional components of the conveyor belt, the constraint strength is appropriately reduced to improve efficiency.

[0157] In the process of solving the objective function in this embodiment, a block iterative calculation strategy is adopted. By decomposing the region, the problem scale is reduced and the solution efficiency is improved. For the complex structure of the robotic arm, a detailed block scheme is adopted; for the regular structure of the conveyor belt, a coarser block strategy is used.

[0158] In the fitting surface evaluation link of this embodiment, a multi-criterion evaluation mechanism is developed. By calculating the distance field distribution, fitting accuracy and surface quality are comprehensively considered. For the precision parts of the robotic arm, local fitting errors are focused on; for the conventional components of the conveyor belt, overall shape preservation is emphasized.

[0159] In the construction of the geometric matching criterion in this embodiment, an adaptive matching mechanism is realized. By analyzing the characteristics of the slice contour lines, the matching threshold is dynamically adjusted. When dealing with the key joints of the robotic arm, strict matching conditions are adopted; for the transition regions of the conveyor belt, the matching requirements are appropriately relaxed.

[0160] Through the above implementation scheme, this embodiment constructs a complete surface fitting and optimization framework. This scheme can accurately reconstruct the complex geometric shapes of the automated production line equipment and ensure that the reconstruction results meet the requirements of engineering applications. In practical applications, this scheme has successfully processed various equipment models including robotic arms, conveyor belts, control cabinets, etc., significantly improving the quality and efficiency of geometric reconstruction.

[0161] In the digital upgrade project of the production line, the solution of this embodiment provides reliable technical support for the accurate modeling of production line equipment through precise surface fitting and effective optimization strategies, promoting the intelligent transformation process of the production line. Through the optimized surface reconstruction results, it effectively supports subsequent simulation analysis and intelligent control, providing an important guarantee for the improvement of production line efficiency and quality.

[0162] In an embodiment of the method for reconstructing the digital model of an automated production line of the present application, refer to Figure 7 , and it may specifically include the following content:

[0163] Step S701: Calculate the splicing boundary of adjacent surface patches according to the geometric matching result, construct a transition region at the splicing boundary, establish a surface splicing equation with the geometric parameters of the transition region as constraint conditions, solve the surface splicing equation using an iterative optimization algorithm to obtain the continuity parameters of the surface, construct an error compensation function based on the continuity parameters, and apply the error compensation function to the surface seam region to eliminate the splicing gap;

[0164] Step S702: Establish a grid topology structure for the assembled reconstruction model, calculate the geometric quality evaluation index of the grid topology structure, determine the region that needs to be refined based on the geometric quality evaluation index, construct a grid refinement criterion in the refinement region, perform local refinement on the reconstruction model using an adaptive grid refinement algorithm, and convert the refined reconstruction model into a three-dimensional solid model and import it into the automated production line environment.

[0165] Optionally, in the process of calculating the splicing boundary in this embodiment, an adaptive boundary recognition strategy is adopted. For the geometric features of different equipment in the automated production line, the boundary position is determined by analyzing the curvature continuity of the surface. For the joint connection of the robotic arm, the range of the transition region is determined based on the motion characteristics; for the splicing of the conveyor belt, the boundary width is determined according to the force analysis.

[0166] In this embodiment, when constructing the transition region, a shape control mechanism based on physical constraints is realized. By establishing curvature continuity and tangential continuity constraints, the smoothness of the transition region is ensured. When dealing with the precision mating surface of the robotic arm, high-order continuity constraints are adopted; for the ordinary joint surface of the control cabinet, low-order continuity constraints are used to simplify the calculation.

[0167] In the link of constructing the surface splicing equation in this embodiment, a solution framework for multi-objective optimization is developed. By comprehensively considering geometric error and surface quality, reasonable optimization objectives are established. For the functional surface of the robotic arm, the focus is on optimizing shape preservation; for the transition region of the conveyor belt, the focus is on optimizing smoothness.

[0168] In this embodiment, a progressive optimization strategy is adopted during the calculation of continuous parameters. By designing an adaptive step-size control algorithm, the convergence of the optimization process is ensured. When dealing with the complex curved surface of the robotic arm, a smaller optimization step-size is used; for the simple curved surface of the conveyor belt, a larger step-size is adopted to accelerate convergence.

[0169] In the design of the error compensation function in this embodiment, a locally adaptive compensation mechanism is implemented. By analyzing the geometric features of the seam area, a reasonable compensation strategy is constructed. When dealing with the precision parts of the robotic arm, precise error compensation is adopted; for the conventional components of the conveyor belt, a simplified compensation scheme is used.

[0170] In the process of constructing the mesh topology in this embodiment, a hierarchical mesh generation strategy is adopted. By analyzing the geometric complexity of the model, the mesh density is dynamically adjusted. For the irregular surface of the robotic arm, a fine mesh division is used; for the planar structure of the control cabinet, a coarser mesh division is adopted to improve efficiency.

[0171] In the design of the quality evaluation index in this embodiment, a multi-dimensional evaluation system is developed. By calculating the shape quality, size uniformity, and topological correctness of the mesh, the reconstruction effect is comprehensively evaluated. For the key parts of the robotic arm, the local mesh quality is focused on; for the conventional areas of the conveyor belt, the overall uniformity is emphasized.

[0172] In the link of constructing the mesh encryption criterion in this embodiment, a feature-based adaptive mechanism is implemented. By analyzing the complexity of the geometric features, a reasonable encryption threshold is determined. When dealing with the precision mating surface of the robotic arm, a strict encryption criterion is adopted; for the transition area of the conveyor belt, the encryption requirements are appropriately relaxed.

[0173] In the process of model conversion in this embodiment, a feature-preserving conversion strategy is adopted. By maintaining the continuity and integrity of the geometric features, the conversion quality is ensured. For the functional surface of the robotic arm, the feature information is accurately maintained; for the regular structure of the control cabinet, a simplified conversion scheme is adopted to improve efficiency.

[0174] Through the above implementation scheme, this embodiment constructs a complete model reconstruction and optimization framework. This scheme can accurately reconstruct the complex equipment models in the automated production line and ensure that the reconstruction results meet the requirements of engineering applications. In practical applications, this scheme has successfully handled the reconstruction tasks of various equipment including robotic arms, conveyor belts, control cabinets, etc., significantly improving the quality and reliability of the digital models.

[0175] In the digital upgrade project of the production line, the solution of this embodiment provides reliable technical support for the accurate modeling of production line equipment through precise surface splicing and effective mesh optimization. The optimized 3D solid model can accurately reflect the geometric features of the equipment, providing a high-quality geometric basis for subsequent assembly analysis, collision detection, and motion simulation, effectively supporting the intelligent transformation and efficiency improvement of the production line.

[0176] In order to break through the limitations of traditional model reconstruction and provide an efficient and reliable solution for industrial digital twins, this application provides an embodiment of an automated production line digital model reconstruction device for implementing all or part of the content of the automated production line digital model reconstruction method. See Figure 8 The automated production line digital model reconstruction device specifically includes the following contents:

[0177] The model slicing module 10 is used to slice the automated production line digital model along the main coordinate axes, extract the feature point sequences of the slice contour lines at different resolutions, perform geometric semantic annotation on the feature point sequences, construct a variational level set function from the annotated feature point sequences, extract hierarchical feature maps from the variational level set function using a deep convolutional network, establish an implicit surface reconstruction framework with physical constraint terms based on the hierarchical feature maps, and embed the physical constraint terms as regularization conditions into the implicit surface reconstruction process;

[0178] The space partitioning module 20 is used to partition the feature point sequences in space using an octree structure under the implicit surface reconstruction framework, construct a nearest neighbor search algorithm based on the space partitioning, map the nearest neighbor search algorithm into a parallel computing framework to establish an acceleration strategy, construct a local coordinate system according to the nearest neighbor search results, calculate the orthogonal basis vectors of the feature points in the local coordinate system using the principal component analysis method, establish an orthogonal basis space, decompose the feature point projection matrix in the orthogonal basis space into a three-dimensional covariance matrix, and input the three-dimensional covariance matrix into the noise filtering module for data enhancement;

[0179] The model reconstruction module 30 is used to construct local surface patches based on the eigenvectors of the three-dimensional covariance matrix, perform moving least squares fitting on the local surface patches using an adaptive basis function, add topological constraints and smoothness constraints during the fitting process, geometrically match the fitted surface patches with the slice contour lines, assemble and splice the surface patches according to the geometric matching results, iteratively refine the assembled and spliced model, establish an error compensation mechanism to eliminate surface seams, and import the compensated reconstruction model into the automated production line three-dimensional environment for collision detection.

[0180] As can be seen from the above description, the automated production line digital model reconstruction device provided by the embodiments of the present application can extract geometric semantic features through hierarchical slicing and deep convolutional networks, and innovatively constructs an implicit surface reconstruction framework with physical constraints. The system uses an octree structure and a parallel computing framework to achieve efficient nearest neighbor search, and combines principal component analysis and covariance matrix decomposition for feature extraction and noise filtering. Through the moving least squares fitting with adaptive basis functions and the error compensation mechanism, the accuracy and continuity of surface reconstruction are ensured, and seamless integration with the collision detection function of the automated production line is achieved. This method breaks through the limitations of traditional model reconstruction and provides an efficient and reliable solution for industrial digital twins.

[0181] At the hardware level, in order to break through the limitations of traditional model reconstruction and provide an efficient and reliable solution for industrial digital twins, the present application provides an embodiment of an electronic device for implementing all or part of the content in the automated production line digital model reconstruction method. The electronic device specifically includes the following:

[0182] A processor, a memory, a communications interface, and a bus; wherein, the processor, the memory, and the communications interface communicate with each other through the bus; the communications interface is used to implement information transmission between the automated production line digital model reconstruction device and related devices such as the core business system, the user terminal, and the relevant database. The logic controller can be a desktop computer, a tablet computer, a mobile terminal, etc., and this embodiment is not limited thereto. In this embodiment, the logic controller can be implemented with reference to the embodiments of the automated production line digital model reconstruction method and the embodiments of the automated production line digital model reconstruction device, and the content is incorporated herein, and the repeated parts will not be elaborated.

[0183] It can be understood that the user terminal may include a smart phone, a tablet electronic device, an Internet set-top box, a portable computer, a desktop computer, a personal digital assistant (PDA), a vehicle-mounted device, a smart wearable device, etc. Among them, the smart wearable device may include smart glasses, a smart watch, a smart bracelet, etc.

[0184] In practical applications, part of the automated production line digital model reconstruction method can be executed on the electronic device side as described above, or all operations can be completed in the client device. Specifically, it can be selected according to the processing capacity of the client device and the limitations of the user usage scenario, etc. The present application does not make any limitation in this regard. If all operations are completed in the client device, the client device may further include a processor.

[0185] The above-mentioned client device may have a communication module (i.e., communication unit), which can communicate with a remote server to achieve data transmission with the server. The server may include a server on the task scheduling center side, and in other implementation scenarios, it may also include a server of an intermediate platform, such as a server of a third-party server platform with a communication link to the task scheduling center server. The server may include a single computer device, or a server cluster composed of multiple servers, or a server structure of a distributed device.

[0186] Figure 9 FIG. is a schematic block diagram of the system composition of the electronic device 9600 according to an embodiment of the present application. As Figure 9 shown, the electronic device 9600 may include a central processing unit 9100 and a memory 9140; the memory 9140 is coupled to the central processing unit 9100. It should be noted that this Figure 9 is exemplary; other types of structures may also be used to supplement or replace this structure to achieve telecommunication functions or other functions.

[0187] In one embodiment, the function of the automated production line digital model reconstruction method may be integrated into the central processing unit 9100. Among them, the central processing unit 9100 may be configured to perform the following controls:

[0188] Step S101: Slice the automated production line digital model along the main coordinate axis, extract the feature point sequence of the slice contour line at different resolutions, perform geometric semantic annotation on the feature point sequence, construct a variational level set function from the annotated feature point sequence, extract a hierarchical feature map from the variational level set function using a deep convolutional network, establish an implicit surface reconstruction framework with physical constraint terms based on the hierarchical feature map, and embed the physical constraint terms as regularization conditions into the implicit surface reconstruction process;

[0189] Step S102: Under the implicit surface reconstruction framework, use an octree structure to partition the space of the feature point sequence, construct a nearest neighbor search algorithm based on the space partition, map the nearest neighbor search algorithm into a parallel computing framework to establish an acceleration strategy, construct a local coordinate system according to the nearest neighbor search result, calculate the orthogonal basis vectors of the feature points in the local coordinate system using the principal component analysis method, establish an orthogonal basis space, decompose the feature point projection matrix in the orthogonal basis space into a three-dimensional covariance matrix, and input the three-dimensional covariance matrix into a noise filtering module for data enhancement;

[0190] Step S103: Construct local surface patches based on the eigenvectors of the three-dimensional covariance matrix, perform moving least squares fitting on the local surface patches using an adaptive basis function, add topological constraints and smoothness constraints during the fitting process, geometrically match the fitted surface patches with the slice contour lines, assemble and splice the surface patches according to the geometric matching results, iteratively refine the assembled and spliced model, establish an error compensation mechanism to eliminate surface seams, and import the compensated reconstructed model into the three-dimensional environment of the automated production line for collision detection.

[0191] As can be seen from the above description, the electronic device provided in the embodiment of the present application extracts geometric semantic features through layer-by-layer slicing and a deep convolutional network, and innovatively constructs an implicit surface reconstruction framework with physical constraints. The system uses an octree structure and a parallel computing framework to achieve efficient nearest neighbor search, and combines principal component analysis and covariance matrix decomposition for feature extraction and noise filtering. Through the moving least squares fitting of the adaptive basis function and the error compensation mechanism, the accuracy and continuity of surface reconstruction are ensured, and seamless integration with the collision detection function of the automated production line is achieved. This method breaks through the limitations of traditional model reconstruction and provides an efficient and reliable solution for industrial digital twins.

[0192] In another implementation, the automated production line digital model reconstruction device can be separately configured from the central processing unit 9100. For example, the automated production line digital model reconstruction device can be configured as a chip connected to the central processing unit 9100, and the functions of the automated production line digital model reconstruction method are realized through the control of the central processing unit.

[0193] As Figure 9 shown, the electronic device 9600 may further include: a communication module 9110, an input unit 9120, an audio processor 9130, a display 9160, and a power supply 9170. It should be noted that the electronic device 9600 does not necessarily have to include Figure 9 all the components shown in Figure 9 ; in addition, the electronic device 9600 may further include

[0194] components not shown in Figure 9 ; reference may be made to the prior art.

[0195] Among them, the memory 9140 can be, for example, one or more of a buffer, a flash memory, a hard drive, a removable medium, a volatile memory, a non-volatile memory, or other suitable devices. It can store the above information related to failures, and can also store programs for executing relevant information. And the central processing unit 9100 can execute the program stored in the memory 9140 to achieve information storage or processing, etc.

[0196] The input unit 9120 provides input to the central processing unit 9100. The input unit 9120 is, for example, a key or a touch input device. The power supply 9170 is used to supply power to the electronic device 9600. The display 9160 is used to display display objects such as images and texts. The display can be, for example, an LCD display, but is not limited thereto.

[0197] The memory 9140 can be a solid-state memory. For example, it can be a read-only memory (ROM), a random access memory (RAM), a SIM card, etc. It can also be a memory that stores information even when powered off, can be selectively erased, and has more data. An example of this memory is sometimes referred to as an EPROM, etc. The memory 9140 can also be some other type of device. The memory 9140 includes a buffer memory 9141 (sometimes referred to as a buffer). The memory 9140 can include an application / function storage unit 9142, which is used to store application programs and function programs or the processes for operating the electronic device 9600 through the central processing unit 9100.

[0198] The memory 9140 can also include a data storage unit 9143, which is used to store data, such as contacts, digital data, pictures, sounds, and / or any other data used by the electronic device. The driver storage unit 9144 of the memory 9140 can include various drivers of the electronic device for communication functions and / or for executing other functions of the electronic device (such as a messaging application, an address book application, etc.).

[0199] The communication module 9110 is a transmitter / receiver that transmits and receives signals via the antenna 9111. The communication module 9110 (transmitter / receiver) is coupled to the central processing unit 9100 to provide input signals and receive output signals, which can be the same as in the case of a conventional mobile communication terminal.

[0200] Based on different communication technologies, in the same electronic device, multiple communication modules 9110 can be provided, such as a cellular network module, a Bluetooth module, and / or a wireless local area network module, etc. The communication module 9110 (transmitter / receiver) is also coupled to a speaker 9131 and a microphone 9132 via an audio processor 9130 to provide an audio output via the speaker 9131 and receive an audio input from the microphone 9132, so as to implement normal telecommunication functions. The audio processor 9130 may include any suitable buffer, decoder, amplifier, etc. Additionally, the audio processor 9130 is also coupled to a central processor 9100, enabling recording on the device through the microphone 9132 and playing the sound stored on the device through the speaker 9131.

[0201] Embodiments of the present application also provide a computer-readable storage medium capable of implementing all steps in the automated production line digital model reconstruction method where the execution subject in the above embodiments is a server or a client. A computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, all steps in the automated production line digital model reconstruction method where the execution subject in the above embodiments is a server or a client are implemented. For example, when the processor executes the computer program, the following steps are implemented:

[0202] Step S101: Stratify and slice the automated production line digital model along the main coordinate axes, extract the feature point sequence of the slice contour line at different resolutions, perform geometric semantic annotation on the feature point sequence, construct a variational level set function from the annotated feature point sequence, extract a hierarchical feature map from the variational level set function using a deep convolutional network, establish an implicit surface reconstruction framework with physical constraint terms based on the hierarchical feature map, and embed the physical constraint terms as regularization conditions into the implicit surface reconstruction process;

[0203] Step S102: Under the implicit surface reconstruction framework, use an octree structure to perform spatial partitioning on the feature point sequence, construct a nearest neighbor search algorithm based on the spatial partitioning, map the nearest neighbor search algorithm into a parallel computing framework to establish an acceleration strategy, construct a local coordinate system according to the nearest neighbor search result, calculate the orthogonal basis vectors of the feature points in the local coordinate system using the principal component analysis method, establish an orthogonal basis space, decompose the feature point projection matrix in the orthogonal basis space into a three-dimensional covariance matrix, and input the three-dimensional covariance matrix into a noise filtering module for data enhancement;

[0204] Step S103: Construct local surface patches based on the eigenvectors of the three-dimensional covariance matrix, perform moving least squares fitting on the local surface patches using adaptive basis functions, add topological constraints and smoothness constraints during the fitting process, geometrically match the fitted surface patches with the slice contour lines, assemble and splice the surface patches according to the geometric matching results, iteratively refine the assembled and spliced model, establish an error compensation mechanism to eliminate surface seams, and import the compensated reconstruction model into the three-dimensional environment of the automated production line for collision detection.

[0205] As can be seen from the above description, the computer-readable storage medium provided by the embodiments of the present application extracts geometric semantic features through layer slicing and deep convolutional networks, and innovatively constructs an implicit surface reconstruction framework with physical constraints. The system uses an octree structure and a parallel computing framework to achieve efficient nearest neighbor search, and combines principal component analysis and covariance matrix decomposition for feature extraction and noise filtering. Through the moving least squares fitting with adaptive basis functions and the error compensation mechanism, the accuracy and continuity of surface reconstruction are ensured, and seamless integration with the collision detection function of the automated production line is achieved. This method breaks through the limitations of traditional model reconstruction and provides an efficient and reliable solution for industrial digital twins.

[0206] The embodiments of the present application also provide a computer program product that can implement all the steps in the automated production line digital model reconstruction method whose execution subject is a server or a client in the above embodiments. When the computer program / instructions are executed by a processor, the steps of the automated production line digital model reconstruction method are implemented. For example, the computer program / instructions implement the following steps:

[0207] Step S101: Slice the automated production line digital model along the main coordinate axes, extract the sequence of feature points of the slice contour lines at different resolutions, perform geometric semantic annotation on the sequence of feature points, construct a variational level set function from the annotated sequence of feature points, extract hierarchical feature maps from the variational level set function using a deep convolutional network, establish an implicit surface reconstruction framework with physical constraint terms based on the hierarchical feature maps, and embed the physical constraint terms as regularization conditions into the implicit surface reconstruction process;

[0208] Step S102: Under the implicit surface reconstruction framework, use an octree structure to partition the space of the sequence of feature points, construct a nearest neighbor search algorithm based on the space partition, map the nearest neighbor search algorithm into a parallel computing framework to establish an acceleration strategy, construct a local coordinate system according to the nearest neighbor search results, calculate the orthogonal basis vectors of the feature points in the local coordinate system using the principal component analysis method, establish an orthogonal basis space, decompose the feature point projection matrix in the orthogonal basis space into a three-dimensional covariance matrix, and input the three-dimensional covariance matrix into a noise filtering module for data enhancement;

[0209] Step S103: Construct local surface patches based on the eigenvectors of the three-dimensional covariance matrix, perform moving least squares fitting on the local surface patches using an adaptive basis function, add topological constraints and smoothness constraints during the fitting process, geometrically match the fitted surface patches with the slice contour lines, assemble and splice the surface patches according to the geometric matching results, iteratively refine the assembled and spliced model, establish an error compensation mechanism to eliminate surface seams, and import the compensated reconstructed model into the three-dimensional environment of the automated production line for collision detection.

[0210] As can be seen from the above description, the computer program product provided by the embodiments of the present application extracts geometric semantic features through layer slicing and deep convolutional networks, and innovatively constructs an implicit surface reconstruction framework with physical constraints. The system uses an octree structure and a parallel computing framework to achieve efficient nearest neighbor search, and combines principal component analysis and covariance matrix decomposition for feature extraction and noise filtering. Through the moving least squares fitting with an adaptive basis function and an error compensation mechanism, the accuracy and continuity of surface reconstruction are ensured, and seamless integration with the collision detection function of the automated production line is achieved. This method breaks through the limitations of traditional model reconstruction and provides an efficient and reliable solution for industrial digital twins.

[0211] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, apparatus, or computer program product. Therefore, the present invention can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0212] The present invention is described with reference to the flowcharts and / or block diagrams of methods, devices (apparatus), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, and the combination of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing devices produce a device for implementing the functions specified in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0213] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to operate in a particular manner, such that the instructions stored in the computer-readable memory produce a manufacture including an instruction device that implements the functions specified in one or more of the processes and / or blocks Figure 1 one or more of the processes and / or blocks Figure 1 specified in the block or blocks.

[0214] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, whereby the instructions executed on the computer or other programmable apparatus provide steps for implementing the functions specified in one or more of the processes and / or blocks Figure 1 one or more of the processes and / or blocks Figure 1 specified in the block or blocks.

[0215] In the present invention, specific embodiments are used to illustrate the principles and implementation manners of the present invention. The description of the above embodiments is only used to help understand the method and its core idea of the present invention; at the same time, for those of ordinary skill in the art, according to the idea of the present invention, there will be changes in the specific implementation manners and application scopes. In summary, the content of this specification should not be construed as a limitation to the present invention.

Claims

1. A method for reconstructing a digital model of an automated production line, characterized in that: The method comprises: The digital model of the automated production line is sliced ​​along the principal coordinate axis, a sequence of feature points of the slice contour is extracted at different resolutions, the sequence of feature points is geometrically annotated, a variational level set function is constructed for the annotated sequence of feature points, a deep convolutional network is used to extract a hierarchical feature map from the variational level set function, an implicit surface reconstruction framework with physical constraints is established based on the hierarchical feature map, and the physical constraints are embedded in the implicit surface reconstruction process as regularization conditions; In the implicit surface reconstruction framework, an octree structure is used to perform spatial partitioning on a feature point sequence, a neighbor point search algorithm based on the spatial partitioning is constructed, the neighbor point search algorithm is mapped to a parallel computing framework to establish an acceleration strategy, a local coordinate system is constructed according to the neighbor point search results, a principal component analysis method is used to calculate the orthogonal basis vectors of the feature points in the local coordinate system, an orthogonal basis space is established, a feature point projection matrix in the orthogonal basis space is decomposed into a three-dimensional covariance matrix, and the three-dimensional covariance matrix is ​​input into a noise filtering module for data enhancement; A local surface patch is constructed based on the eigenvector of the three-dimensional covariance matrix, and an adaptive basis function is used to perform moving least squares fitting on the local surface patch. During the fitting process, topological constraints and smoothness constraints are added, and the fitted surface patch is geometrically matched with the slice contour line. The surface patch is assembled and spliced ​​according to the geometric matching result, and the assembled and spliced ​​model is iteratively refined. An error compensation mechanism is established to eliminate surface seams, and the compensated reconstructed model is imported into the three-dimensional environment of the automated production line for collision detection.

2. The method for reconstructing the digital model of an automated production line according to claim 1, characterized in that: The method comprises: slicing the automated production line digital model along the principal coordinate axis, extracting a sequence of feature points of the slice contour at different resolutions, geometrically annotating the sequence of feature points, and constructing a variational level set function for the annotated sequence of feature points, including: The slice direction of the principal coordinate axis is calculated by using a layered scanning algorithm of a three-dimensional mesh model, a slice plane parallel to the coordinate axis is constructed based on the slice direction, the slice plane is intersected with the boundary surface of the digital model to generate a slice contour line, the slice contour line is meshed to calculate the spatial coordinates of the feature points, and an adaptive resolution sampling mechanism is established to screen the feature points to obtain an ordered sequence of feature points; A curvature tensor analysis model is established for the feature point sequence, the geometric properties of the feature points are calculated based on the main direction of the curvature tensor, the geometric properties are assigned to the feature points as label information, a variational objective functional with weight coefficients is constructed, the mathematical expression of the level set function is obtained by solving the variational objective functional, and a mapping relationship from the feature point sequence to the level set function is established.

3. The method for reconstructing the digital model of an automated production line according to claim 1, characterized in that: The method uses a deep convolutional network to extract a hierarchical feature map from the variational level set function, establishes an implicit surface reconstruction framework with a physical constraint term based on the hierarchical feature map, and embeds the physical constraint term into the implicit surface reconstruction process as a regularization condition, including: Constructing a multi-scale convolutional neural network structure, setting convolution kernels of different sizes in the convolution layer to extract the spatial features of the variational level set function, reducing the dimension of the feature map through pooling operation, inputting the reduced-dimensional feature map into the deconvolution layer for upsampling, and splicing and fusing the upsampling result with the jump-connected feature map to generate a hierarchical feature map; An implicit surface equation is constructed based on the hierarchical feature mapping, physical constraints of surface smoothness and shape preservation are added to the implicit surface equation, the physical constraints are converted into regularization conditions in the form of Lagrange multipliers, an objective function with regularization terms is established, and the objective function is solved by the variational method to obtain an expression for implicit surface reconstruction.

4. The method for reconstructing the digital model of an automated production line according to claim 1, characterized in that: In the implicit surface reconstruction framework, an octree structure is used to perform spatial partitioning on a feature point sequence, a neighbor point search algorithm based on the spatial partitioning is constructed, and the neighbor point search algorithm is mapped to a parallel computing framework to establish an acceleration strategy, including: Calculate the spatial bounding box of the feature point sequence, divide the spatial bounding box into eight subspaces along the coordinate axis direction, calculate the density distribution of the feature points for each subspace, recursively subdivide the subspace based on the density distribution threshold, establish an octree index structure to store the spatial position information of the feature points, and record the spatial position information of the feature points in the corresponding leaf nodes; A neighbor point search path is constructed in the octree index structure, the spatial distance between the feature point to be searched and the leaf node is calculated, the leaf nodes whose spatial distance is less than the search radius are marked as candidate nodes, parallel search tasks are established for the candidate nodes, the parallel search tasks are assigned to the computing units to perform neighbor point search calculations, and the search results are summarized to obtain the neighborhood information of the feature points.

5. The method for reconstructing the digital model of an automated production line according to claim 1, characterized in that: The method of constructing a local coordinate system according to the neighbor point search results, calculating the orthogonal basis vectors of the feature points in the local coordinate system by using the principal component analysis method, establishing an orthogonal basis space, decomposing the feature point projection matrix in the orthogonal basis space into a three-dimensional covariance matrix, and inputting the three-dimensional covariance matrix into a noise filtering module for data enhancement includes: Calculate the centroid coordinates of the neighborhood of the feature point using the neighbor point search results, translate the centroid coordinates to the origin to establish the origin of the local coordinate system, calculate the divergence matrix of the neighborhood of the feature point, perform eigenvalue decomposition on the divergence matrix to obtain an eigenvector matrix, sort the eigenvector matrix according to the eigenvalue size to determine the three basis vectors of the local coordinate system, and construct the rotation matrix of the local coordinate system based on the basis vectors; The feature points are transformed from a global coordinate system to a local coordinate system, the coordinate components of the feature points in the local coordinate system are calculated, a projection matrix of the feature points is constructed, a singular value decomposition is performed on the projection matrix to obtain a three-dimensional covariance matrix, a noise threshold is calculated based on the eigenvalue distribution of the three-dimensional covariance matrix, and the noise threshold is input into an anisotropic diffusion filter as a filtering parameter.

6. The method for reconstructing the digital model of an automated production line according to claim 1, characterized in that: The method of constructing a local surface patch based on the eigenvector of the three-dimensional covariance matrix, using an adaptive basis function to perform moving least squares fitting on the local surface patch, adding topological constraints and smoothness constraints during the fitting process, and geometrically matching the fitted surface patch with the slice contour line includes: The eigenvector of the three-dimensional covariance matrix is ​​used as a normal vector to construct a parametric equation of a local surface patch, the principal curvature and Gaussian curvature of the local surface patch are calculated, the order of the adaptive basis function is selected based on the curvature information, a moving least squares equation with a weight function is constructed, the topological continuity constraint and the curvature smoothness constraint of the surface are introduced into the moving least squares equation, and a weighted least squares objective function with constraint terms is established; An analytical solution is constructed for the weighted least squares objective function, and the objective function is numerically solved using the regional decomposition method. The solution result is substituted into the parametric equation of the local surface patch to obtain the fitting surface, and the distance field distribution between the fitting surface and the slice contour line is calculated. A geometric matching criterion is constructed based on the distance field, and the geometric matching criterion is used as an evaluation index to screen the optimal fitting surface.

7. The method for reconstructing the digital model of an automated production line according to claim 1, characterized in that: The method of assembling and splicing the surface pieces according to the geometric matching results, iteratively refining the assembled and spliced ​​model, establishing an error compensation mechanism to eliminate surface seams, and importing the compensated reconstructed model into the automated production line three-dimensional environment for collision detection includes: Calculate the splicing boundary of adjacent surface patches according to the geometric matching result, construct a transition area at the splicing boundary, establish a surface splicing equation by taking the geometric parameters of the transition area as constraint conditions, solve the surface splicing equation by using an iterative optimization algorithm to obtain the continuity parameters of the surface, construct an error compensation function based on the continuity parameters, and apply the error compensation function to the surface seam area to eliminate the splicing gap; A mesh topology structure is established for the assembled reconstructed model, and the geometric quality evaluation index of the mesh topology structure is calculated. The area that needs to be refined is determined based on the geometric quality evaluation index, and a mesh encryption criterion is constructed in the refined area. An adaptive mesh encryption algorithm is used to locally refine the reconstructed model, and the refined reconstructed model is converted into a three-dimensional solid model and imported into the automated production line environment.

8. An automated production line digital model reconstruction device, characterized in that: The device comprises: A model slicing module is used to slice the automated production line digital model along the principal coordinate axis, extract a sequence of feature points of the slice contour at different resolutions, perform geometric semantic annotation on the sequence of feature points, construct a variational level set function with the annotated sequence of feature points, extract a hierarchical feature map from the variational level set function using a deep convolutional network, establish an implicit surface reconstruction framework with physical constraints based on the hierarchical feature map, and embed the physical constraints as regularization conditions into the implicit surface reconstruction process; A spatial partitioning module is used to use an octree structure to perform spatial partitioning on a feature point sequence under the implicit surface reconstruction framework, construct a neighbor point search algorithm based on the spatial partitioning, map the neighbor point search algorithm to a parallel computing framework to establish an acceleration strategy, construct a local coordinate system based on the neighbor point search results, calculate the orthogonal basis vectors of the feature points under the local coordinate system using a principal component analysis method, establish an orthogonal basis space, decompose the feature point projection matrix in the orthogonal basis space into a three-dimensional covariance matrix, and input the three-dimensional covariance matrix into a noise filtering module for data enhancement; A model reconstruction module is used to construct a local surface patch based on the eigenvector of the three-dimensional covariance matrix, use an adaptive basis function to perform moving least squares fitting on the local surface patch, add topological constraints and smoothness constraints during the fitting process, geometrically match the fitted surface patch with the slice contour line, assemble and splice the surface patch according to the geometric matching results, iteratively refine the assembled and spliced ​​model, establish an error compensation mechanism to eliminate surface seams, and import the compensated reconstructed model into the three-dimensional environment of the automated production line for collision detection.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the steps of the method for reconstructing the digital model of an automated production line according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for reconstructing the digital model of an automated production line described in any one of claims 1 to 7 are implemented.

Images