A method for generating shell element meshes for bridge components based on principal plane projection of 3D point clouds
By projecting the three-dimensional point cloud of the outer surface of the bridge component onto the main plane and extracting the nodes layer by layer, a regular shell element mesh is generated, which solves the problems of unstable node extraction and shell element connection in the prior art and realizes efficient finite element analysis preparation.
Patent Information
- Application Number
- CN202611419871.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-09-14
- Publication Date
- 2026-11-17
AI Technical Summary
Existing finite element meshing methods for directly generating bridge shell components from 3D point clouds struggle to balance node extraction with regular distribution when dealing with complex boundaries. The lack of deterministic shell element connection relationships leads to unstable generated meshes and low automation.
The 3D point cloud of the outer surface of the bridge component is projected onto the main plane. Nodes are extracted layer by layer and adjacent nodes are connected based on parametric coordinates to generate planar quadrilateral units. Finally, the mesh is back-projected into 3D space to ensure the regularity of the mesh shape and the geometric fit.
The process of node extraction and mesh generation is simplified, and the generated shell element mesh has high morphological regularity and geometric fidelity, making it suitable for finite element analysis and enhancing the workflow from 3D scanning to finite element analysis.
Smart Images

Figure CN123118969A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the fields of digitalization of the building industry, digital twins of bridges, 3D point cloud processing, and finite element mesh generation, and particularly to a method for generating shell element meshes for bridge components based on the principal plane projection of 3D point clouds. Background Technology
[0002] 3D laser scanning can acquire high-density 3D point cloud data of bridge component surfaces. To use this point cloud data for finite element analysis, it is typically necessary to generate a finite element mesh composed of nodes and elements from the point cloud. Existing methods typically involve first reconstructing the geometric surface, CAD model, or BIM model from the 3D point cloud, and then generating the finite element mesh based on the reconstructed geometric model. This type of method requires an intermediate geometric reconstruction process and is easily affected by surface fitting accuracy, manual modeling assumptions, and parameter settings. Especially when dealing with bridge slab and shell components with irregular boundaries, local deformations, or construction deviations, the modeling process is cumbersome and makes rapid conversion from point cloud to finite element mesh difficult.
[0003] To reduce intermediate geometric reconstruction steps, existing methods have attempted to directly generate finite element meshes from 3D point clouds. For linear components such as beams, rods, and trusses, direct modeling is relatively easy because their finite element topology can usually be determined by centerlines, endpoints, or feature nodes. However, for plate and shell components such as bridge decks, box girder webs, and the outer surface of arch ribs, simply determining node positions is insufficient to form a shell element mesh usable for analysis; stable 2D element connections must also be established on the component surface. Since measured point clouds often exhibit irregular boundaries, non-convex profiles, local missing elements, and uneven sampling density, directly generating plate and shell meshes from 3D point clouds remains quite challenging.
[0004] Existing point cloud node placement methods mainly include spatial interpolation, local region partitioning, and adaptive node generation methods based on iterative optimization or exclusion mechanisms. While these methods can achieve relatively uniform node distribution to a certain extent, they are prone to insufficient boundary fitting when dealing with complex or non-convex boundaries. Some iterative node generation methods are also susceptible to the influence of initial conditions, random processes, and parameter selection, leading to unstable node distribution results and making it difficult to simultaneously meet the requirements of boundary consistency, node uniformity, and overall topological regularity.
[0005] Besides node arrangement, generating shell element connections is also a key issue in direct meshing of 3D point clouds. Existing 3D surface mesh generation methods, such as the sphere rotation method, Poisson surface reconstruction method, and Delaunay-type methods, are typically geared towards general surface reconstruction tasks. For open plate shell surfaces or non-closed outer surfaces in bridge components, these methods are prone to topological inconsistencies. For example, implicit surface reconstruction methods may generate closed or redundant surfaces, leading to extra shell elements near the boundaries; explicit surface reconstruction methods are easily affected by point cloud density and sampling uniformity, potentially resulting in holes, missing shell elements, or incomplete meshes when the point cloud distribution is uneven. These problems often require subsequent manual repair or semi-automatic restoration, reducing the automation and stability of direct finite element mesh generation from 3D point clouds.
[0006] Therefore, existing methods for direct mesh generation from 3D point clouds have the following main shortcomings in applications involving plate and shell components: First, the node extraction process struggles to balance complex boundary fit with a regular overall node distribution; second, the shell element connection relationships lack deterministic constraints, easily leading to problems such as over-generation, under-generation, or inconsistent normals of shell elements. In summary, existing finite element node extraction and mesh generation methods based on 3D point clouds are highly sensitive to point cloud density, boundary morphology, and parameter settings, making it difficult to directly and stably generate shell element meshes suitable for finite element analysis. Summary of the Invention
[0007] To address the shortcomings of existing technologies, this invention proposes a method for generating shell element meshes for bridge components based on the projection of a 3D point cloud onto a principal plane. This method projects the 3D point cloud of the bridge component's outer surface onto a principal plane, thereby reducing the 3D mesh generation problem to node arrangement and shell element connection within a planar domain. This effectively reduces geometric complexity and ensures that the final generated shell element mesh possesses high morphological regularity and geometric fidelity.
[0008] The specific technical solution is as follows: A method for generating shell element meshes for bridge components based on principal plane projection of 3D point clouds includes the following steps: S1: Project the 3D point cloud of the outer surface of the bridge component onto the main plane, and record each projection point in the projected point cloud. The coordinates; S2: Extract grid nodes layer by layer from the outside to the inside based on the projected point cloud to obtain the coordinates of all nodes; this is specifically achieved through the following sub-steps: S201: Set the mesh size to a×b, a≤b, and the total number of nodes to (a+1)×(b+1); each node is denoted according to its topological position in the mesh. ,in, , Divide all nodes into m layers from the outside in. The outermost layer includes 4 corner points and 2a+2b-4 edge points. The (a-1)×(b-1) nodes in the remaining layers are all internal points. S202: Based on the concave hull of the projected point cloud, extract the coordinates of 4 corner points and 2a+2b-4 edge points from the projected point cloud; S203: Extract the coordinates of the nodes in the innermost layer from the projected point cloud based on the coordinates of the outermost nodes and the shape of the enclosed polygon. S204: Extract the coordinates of each node from the outside to the inside according to the method of S203, until the extraction of all m-layer nodes is completed; S3: Node-based The parameter coordinates [j,k] are used to connect four adjacent nodes to create a planar quadrilateral element, thus achieving mesh generation; S4: Back projection of planar nodes and shell elements: based on the coordinate mapping relationship of node parameters Planar nodes Recovering the planar quadrilateral element to its coordinates in three-dimensional space And shell unit connections.
[0009] Furthermore, S1 is specifically implemented through the following sub-steps: S101: For any 3D point cloud on the outer surface of a bridge component, calculate its 3D centroid. and three main directions , , ; S102: Based on the three-dimensional centroid As a point within the principal plane, with the third principal direction The normal direction of the principal plane is used to construct the principal plane of the 3D point cloud. ; S103: On the main plane Internally, it is constructed with a three-dimensional centroid. Taking the origin as the starting point and the first principal direction as the starting point With the u-axis as the axis and the second principal direction as the axis, A Cartesian coordinate system for the v-axis; S104: Calculate any point in a 3D point cloud On the main plane Coordinates of the projection point within .
[0010] Furthermore, in S101, the three principal directions are calculated using principal component analysis; in S104, the coordinates of the projection points... It is a point Vertical projection onto the principal plane Two-dimensional coordinates on the surface, where the coordinate values are... and pass and Calculate the index of the points retained before and after projection. Consistent.
[0011] Furthermore, in S2, the total number of layers m is determined by... Calculate, where This indicates rounding up to the nearest integer.
[0012] Furthermore, in S202, the concave hull of the projected point cloud is capable of enclosing all projected points. The smallest concave polygon is used to represent the outline of the projected point cloud. The extraction method of the four corner points is as follows: calculate the smallest bounding rectangle of the concave hull with the principal axis and the u axis at a 45° angle, and extract the vertices of the concave hull located on the four sides of the smallest bounding rectangle as corner points.
[0013] Furthermore, the extraction of the 2a+2b-4 edge points in S202 specifically includes the following steps: S2021: Based on the four extracted corner points, divide the concave hull of the projected point cloud into four edge lines; S2022: For each edge line, a set of positioning points are set on the edge line according to the ratio of the total length of the edge line to the number of shell elements in that direction. S2023: Extract the actual point in the projected point cloud that is closest to each positioning point, and use it as the edge point corresponding to each edge line.
[0014] Furthermore, S203 is specifically implemented through the following sub-steps: S2031: Calculate the inward contraction direction of each point on a given edge, taking two reference directions on adjacent edges. , ,but The inward contraction direction is , ; S2032: Calculate the intersection direction, if... a If it is an odd number, then when At that time, the intersecting direction Calculated using equation ①, when At that time, the intersecting direction Calculated using equation ②; if a is even, then when At that time, the intersecting direction Calculated using equation ①, when At that time, the intersecting direction The middle node is calculated using equation ②. The two intersecting directions are calculated using equations ① and ②. and ; ① ; ② ; S2033: Calculate the internal positioning points, from both sides towards the middle; if a If it is an odd number, initially when or At that time, the coordinates of the internal positioning point , ;when At that time, the coordinates of the internal positioning point Calculated using equation ③, when hour, Calculated using formula ④; ③ ; ④ ; In the formula, This represents the intersection point formed by the trajectory of point P1 moving along direction D1 and the trajectory of point P2 moving along direction D2. If a is even, then when At that time, the coordinates of the positioning point The calculation method is the same as for odd numbers; when j=(a+2) / 2, the coordinates of the two internal positioning points can be obtained through equations ③ and ④ respectively. and Then the coordinates of the final internal positioning point are taken as and The midpoint, i.e. ; S2034: Following the steps of S2031 to S2033, calculate the internal positioning points of the edge points on the other three edge lines after shrinking them inward by one layer, merge the internal positioning points with the same coordinates, and extract the actual points in the projected point cloud that are closest to each internal positioning point as the precise coordinates of the corresponding nodes in that layer.
[0015] Furthermore, in step S204, when extracting the innermost node, if a is even, then the node is selected from the innermost node. and Two sets of internal positioning points were obtained by performing inward contraction calculations, respectively. and The final coordinates of the internal positioning points are taken as follows: and The midpoint.
[0016] Furthermore, S3 is specifically implemented through the following sub-steps: S301: For each extracted node Assign it the corresponding parameter coordinates [j,k]. The mapping relationship between these parameter coordinates and the original projection point index i is as follows: ; S302: Connect four adjacent nodes clockwise or counterclockwise to create a planar quadrilateral element. A planar quadrilateral element created by clockwise connections is represented as: The planar quadrilateral element created by connecting counterclockwise is represented as: .
[0017] A device for generating shell element meshes for bridge components based on the principal plane projection of three-dimensional point clouds includes a memory and one or more processors. The memory stores executable code, and when the one or more processors execute the executable code, they are used to implement the method for generating shell element meshes for bridge components based on the principal plane projection of three-dimensional point clouds.
[0018] A device for generating shell element meshes for bridge components based on the principal plane projection of three-dimensional point clouds includes a memory and one or more processors. The memory stores executable code, and when the one or more processors execute the executable code, they are used to implement the method for generating shell element meshes for bridge components based on the principal plane projection of three-dimensional point clouds.
[0019] A computer-readable storage medium having a program stored thereon, which, when executed by a processor, implements the method for generating shell element meshes for bridge components based on the principal plane projection of three-dimensional point clouds.
[0020] The beneficial effects of this invention are: (1) This invention simplifies and clarifies node extraction and mesh generation by planarizing the three-dimensional point cloud. Although this process reduces the dimensionality of the data, the original point cloud data is still retained and associated with the projected point cloud through the consistency of the serial numbers. After completing the node extraction and mesh generation in the plane, the three-dimensional shape of each node and mesh can still be restored through the consistency of the serial numbers, ensuring that the generated mesh has a high degree of geometric fit with the original point cloud.
[0021] (2) In addressing the node extraction problem, this invention employs a layered approach: first, corner points are extracted to determine the basic orientation of the mesh; then, boundary points are extracted to control the fit of the contour; and finally, internal points are extracted to ensure uniform shell unit size and regular shape. Furthermore, in extracting internal nodes, a progressive point placement strategy is adopted, proceeding "from the outside in, from both sides to the center." The extracted nodes and hull shape of the outer layer are used to determine the positions of the inner layer nodes, ensuring that the arrangement of all nodes is regular and shape-adaptive, enabling adaptive adjustment of node positions for different point cloud contours.
[0022] (3) In dealing with the problem of mesh generation, the present invention adopts a mesh generation method based on node parameter coordinates. According to the topological position of the node in the mesh, each point is assigned parameter coordinates, and shell elements are established by connecting four nodes corresponding to adjacent parameter coordinates in a clockwise (or counterclockwise) direction. This makes the entire mesh generation process clear and unambiguous, and avoids the absence or redundancy of shell elements.
[0023] (4) The shell element mesh generated by the method described in this invention can be directly saved and imported into the finite element analysis software for subsequent structural performance analysis, which enhances the continuity of the process from 3D scanning to finite element analysis. Attached Figure Description
[0024] Figure 1 This is a flowchart of a method for generating shell element meshes for bridge components based on the principal plane projection of three-dimensional point clouds, as described in an embodiment of the present invention.
[0025] Figure 2 This is a schematic diagram of a preferred three-dimensional point cloud P and its principal plane according to an embodiment of the present invention, wherein... It is the center of mass of P. It is the principal plane of P. , , These are the first, second, and third principal directions of P, respectively.
[0026] Figure 3 In this embodiment of the invention, the three-dimensional point cloud P is on the principal plane. A schematic diagram of the projected point cloud Q, in which yes The origin of the internal local coordinate system, , They are The u-axis and v-axis of the internal local coordinate system, and B is the concave hull of the 3D point cloud P.
[0027] Figure 4 This is a schematic diagram of the mesh and node settings applicable to a 3D point cloud P in an embodiment of the present invention, wherein the number of shell elements in the u direction is 17, the number of shell elements in the v direction is 6, and the mesh nodes are represented by dots. The nodes are divided into four layers from the outside in.
[0028] Figure 5 This is a schematic diagram illustrating the extraction process of the outermost corner points and edge points in an embodiment of the present invention. ReT is the smallest bounding rectangle of a concave hull B whose principal axis forms a 45° angle with the u-axis. B1, B2, B3, and B4 are the four edge lines of the concave hull B, divided by the four corner points. Tar These are the node's location points, marked in red. mark.
[0029] Figure 6 This is the extraction process of each internal point in the second layer in this embodiment of the invention, wherein Dir It is the direction of inward contraction. Dis It's a cross direction.
[0030] Figure 7 This is the extraction result of all 4 layers of nodes in this embodiment of the invention.
[0031] Figure 8 This is a shell unit connection scheme applicable to this embodiment, wherein [ j , k ] is a node The parameter coordinates.
[0032] Figure 9 This is a three-dimensional back projection of the planar nodes and shell elements in the embodiments of the present invention, wherein... It is a one-to-one correspondence between the parameter coordinates [j,k] and the projection point index i.
[0033] Figure 10 This is a schematic diagram of the structure of the bridge component shell unit mesh generation device based on the three-dimensional point cloud principal plane projection in an embodiment of the present invention. Detailed Implementation
[0034] The present invention will be described in detail below with reference to the accompanying drawings and preferred embodiments. The objectives and effects of the present invention will become clearer as a result. The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
[0035] like Figure 1 As shown, a method for generating shell element meshes for bridge components based on the principal plane projection of 3D point clouds includes the following steps: S1: Perform principal plane projection of the 3D point cloud on the outer surface of the bridge components and record the coordinates of each projection point. This is achieved through the following sub-steps: S101: As Figure 2 As shown, for any three-dimensional point cloud P on the outer surface of a bridge component, calculate the three-dimensional centroid of the point cloud. and three main directions , , The three-dimensional centroid of a point cloud is calculated using the following formula:
[0036] In the formula, It is a 3D point cloud with the sequence number as The three-dimensional coordinates (x, y, and z coordinates) of the points are given, and N is the total number of points in the point cloud. In this embodiment, N = 63835.
[0037] main direction , , Calculated using Principal Component Analysis (PCA), it characterizes the dispersion of P's distribution in a single direction. First Principal Direction The point cloud distribution is the sparsest on the upper part, and the shape it exhibits is the most elongated; the third principal direction The point cloud distribution is densest on the upper part, and the shape it exhibits is the narrowest. Three main directions. , , The two pairs are perpendicular to each other. In this embodiment, the three-dimensional centroid is taken. First main direction Second main direction Third main direction .
[0038] S102: Based on the three-dimensional centroid As a point within the principal plane, with the third principal direction The principal plane of the 3D point cloud P is constructed using the normal direction of the principal plane. .
[0039] S103: On the main plane Internally, it is constructed with a three-dimensional centroid. Taking the origin as the starting point and the first principal direction as the starting point With the u-axis as the axis and the second principal direction as the axis, Let v be a Cartesian coordinate system.
[0040] S104: As Figure 3 As shown, the 3D point cloud P is on the principal plane. Let Q be the projected point cloud, and let B be the calculated hull of Q. To determine the specific location of each point in the projected point cloud Q, calculate the hull of any point in the 3D point cloud. On the main plane Coordinates of the projection point within Projection point coordinates It is a point Vertical projection onto the principal plane Two-dimensional coordinates on the surface, where the coordinate values are... and pass and Calculate the index of the points retained before and after projection. Consistent.
[0041] The concave hull B is the smallest concave polygon that can enclose the entire projected point cloud Q, representing the outline of the projected point cloud Q. It can be calculated using the "boundary" function in MATLAB. The shrinkage factor of the projected point cloud Q in this embodiment is 0.116.
[0042] S2: Mesh node extraction based on projected point cloud. This is implemented through the following sub-steps: S201: Set the mesh size to a×b (a≤b), and the total number of nodes to (a+1)×(b+1); each node is denoted according to its topological position in the mesh. j and k represent the positions of the node in the j-th row and k-th column of all nodes, respectively; divide all nodes into m layers from the outside in, with the total number of layers m determined by the formula. Calculate, where This indicates rounding up; the outermost layer of nodes includes 4 corner points and 2a+2b-4 edge points, while the remaining (a-1)×(b-1) nodes in the inner layer are all internal points.
[0043] like Figure 4 As shown, in this embodiment, a=6, b=17, and the corresponding total number of nodes is 7×18. All nodes are divided into m=4 layers according to the order from the outside to the inside, with the outermost layer (i.e., the first layer) containing the nodes. The following conditions must be met: when j=1 or 7, k∈[1,18]; and when k=1 or 18, j∈[1,7]. Starting from the outermost node, proceed to the node in the next innermost layer (i.e., the second layer). The following conditions must be met: when j = 2 or 6, k ∈ [2, 17]; and when k = 2 or 17, j ∈ [2, 6]. The nodes in the next innermost layer (i.e., the third layer) satisfy these conditions. The following conditions must be met: when j=3 or 5, k∈[3,16]; and when k=3 or 16, j∈[3,5]. The innermost (i.e., the fourth layer) node. The condition is satisfied that j=4 and k∈[4,15]. The outermost nodes include 4 corner nodes. There are 42 edge points; all nodes except the outermost ones are internal points.
[0044] S202: Based on the concave hull of the projected point cloud, extract 4 corner points and 2a+2b-4 edge points from the projected point cloud.
[0045] The specific method for extracting the four corner points is as follows: Calculate the minimum bounding rectangle ReT of the concave hull B, whose principal axis forms a 45° angle with the u-axis; extract the vertices of the concave hull located on the four sides of this bounding rectangle as corner points. For example... Figure 5 As shown, in this embodiment, the vertices of the concave hull B located on the four sides of rectangle ReT are extracted. , , , and take them as corner points respectively. The precise coordinates.
[0046] The extraction of 2a+2b-4 edge points is achieved through the following operations: S2021: Based on the four extracted corner points, divide the concave hull into four edge lines; for example... Figure 5 As shown, in this embodiment, the concave hull B can be divided into four side lines (B1, B2, B3, B4).
[0047] S022: For each edge line, based on its total length and the number of shell elements a (or b) in that direction, according to the spacing (or Set up a set of positioning points along the edge line; for example Figure 5 As shown, in this embodiment, taking B3 as an example, the total length of this edge line is... The required number of shell elements, a=6, is determined by the spacing between the points. A set of positioning points was set up on line B3. .
[0048] S2023: Extract the actual point in the projected point cloud that is closest to each positioning point, and use it as the edge point corresponding to that edge. For example... Figure 5 As shown in this embodiment, the point in Q that is closest to the five positioning points is extracted. , , , , and take them as corner points respectively. The precise coordinates.
[0049] S203: Based on the coordinates of the outermost nodes and the polygon shape enclosed by these nodes, further extract the nodes of the innermost layer (i.e., the second layer) from the projected point cloud. For example... Figure 6 As shown, this is achieved through the following sub-steps: S2031: Calculate the inward contraction direction of each point on a given edge. Using a set of points on edge B1... For example, take two reference directions on adjacent sides. , ,but The inward contraction direction is t=(j-1) / a. In this embodiment, there are two reference directions. , , Inward contraction direction , .
[0050] S2032: Calculate the intersection direction. If a is odd, then when At that time, the intersecting direction Calculated using equation ①, when At that time, the intersecting direction Calculated using equation ②; if a is even, then when At that time, the intersecting direction Calculated using equation ①, when At that time, the intersecting direction The middle node is calculated using equation ②. It has two intersecting directions (Obtained through equation ①) and (Obtained through equation ②).
[0051] ① ; ② ; This embodiment a Since it is an even number, therefore when At that time, the intersecting direction ;when At that time, the intersecting direction The middle node It has two intersecting directions and ; S2033: Calculate the internal positioning points, from both sides towards the middle; if a is odd, initially calculate the coordinates of the internal positioning points when j=2 or j=a. , ;when At that time, the coordinates of the internal positioning point Calculated using equation ③, when hour, Calculated using formula ④.
[0052] ③ ; ④ ; In the above formula, This represents the intersection point formed by the trajectory of point P1 moving along direction D1 and the trajectory of point P2 moving along direction D2.
[0053] If a is even, then when At that time, the coordinates of the positioning point The calculation method is the same as for odd numbers; when j=(a+2) / 2, the coordinates of the two internal positioning points can be obtained through equations ③ and ④ respectively. and Then the coordinates of the final internal positioning point are taken as and The midpoint, i.e. .
[0054] In this embodiment, 'a' is an even number, initially when... or At that time, the coordinates of the internal positioning point , ,when or hour, , ,when At that time, the coordinates of the two internal positioning points can be obtained, respectively and Then the coordinates of the final internal positioning point are taken as and The midpoint, i.e. .
[0055] S2034: Following the steps of S2031 to S2033, calculate the internal positioning points of the edge points on the other three edge lines after shrinking them inward by one layer, merge the internal positioning points with the same coordinates, and extract the actual point in the projected point cloud Q that is closest to each internal positioning point as the accurate coordinates of the corresponding node of this layer (i.e., the second layer).
[0056] S204: Using the method in S203, based on the coordinates of the nodes in the innermost layer (i.e., the second layer) and the polygon shape enclosed by these nodes, further extract the nodes in the next innermost layer (i.e., the third layer) from the projected point cloud; repeat this process until all m layers of nodes have been extracted. Specifically, when extracting the innermost layer of nodes, if a is even, then the nodes are... and Two sets of internal positioning points are obtained by contracting inwards respectively. and Then the coordinates of the final internal positioning point are taken as and The midpoint, i.e. m is the total number of floors.
[0057] like Figure 7 As shown, in this embodiment, when extracting the innermost (i.e., the fourth layer) nodes, since 'a' is an even number in this embodiment, the nodes are selected from the data. and Performing inward contraction calculations separately will yield two sets of internal positioning points, respectively. and Then the coordinates of the final internal positioning point are taken as and The midpoint, i.e. .
[0058] S3: Mesh the projected point cloud based on node parameter coordinates, and establish planar quadrilateral elements (i.e., shell elements). This is achieved through the following sub-steps: S301: For each extracted node Assign it the corresponding parameter coordinates [j,k], which are related to the original projection point index i. One-to-one correspondence, This represents a mapping function.
[0059] S302: Connect four adjacent nodes clockwise or counterclockwise to create a planar quadrilateral element. A planar quadrilateral element created by clockwise connections is represented as: The planar quadrilateral element created by connecting counterclockwise is represented as: ,like Figure 8 As shown. Connect clockwise. Connect counterclockwise There are two shell unit connection schemes available, but consistency should be maintained within the same embodiment, and these two schemes should not be mixed.
[0060] S4: Back projection of planar nodes and shell elements. Specifically: planar nodes and planar quadrilateral unit or According to the mapping relationship And the consistent index i before and after projection, to restore its coordinates in three-dimensional space. and shell unit connection ,like Figure 9 As shown.
[0061] Corresponding to the aforementioned embodiments of the bridge component shell element mesh generation method based on the principal plane projection of three-dimensional point cloud, the present invention also provides embodiments of the bridge component shell element mesh generation device based on the principal plane projection of three-dimensional point cloud.
[0062] like Figure 10 As shown in the figure, an embodiment of the present invention provides a bridge component shell element mesh generation device based on the principal plane projection of a three-dimensional point cloud, including a memory and one or more processors, for implementing the bridge component shell element mesh generation method based on the principal plane projection of a three-dimensional point cloud in the above embodiment.
[0063] The embodiment of the bridge component shell element mesh generation device based on the principal plane projection of three-dimensional point cloud of the present invention can be applied to any device with data processing capabilities, such as a computer. The device embodiment can be implemented in software, hardware, or a combination of both. Taking software implementation as an example, as a logical device, it is formed by the processor of any data processing device loading the corresponding computer program instructions from non-volatile memory into memory for execution. From a hardware perspective, besides... Figure 10 In addition to the processor, memory, network interface, and non-volatile memory shown, any data processing device in the embodiment may also include other hardware depending on the actual function of the data processing device, which will not be described in detail here.
[0064] The specific implementation process of the functions and roles of each unit in the above device can be found in the implementation process of the corresponding steps in the above method, and will not be repeated here.
[0065] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to in the description of the method embodiments. The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of the present invention according to actual needs. Those skilled in the art can understand and implement this without creative effort.
[0066] This invention also provides a computer-readable storage medium storing a program that, when executed by a processor, implements the bridge component shell element mesh generation method based on the principal plane projection of three-dimensional point cloud described in the above embodiments.
[0067] The computer-readable storage medium can be an internal storage unit of any data processing device described in any of the foregoing embodiments, such as a hard disk or memory. The computer-readable storage medium can also be an external storage device, such as a plug-in hard disk, smart media card (SMC), SD card, flash card, etc., equipped on the device. Furthermore, the computer-readable storage medium can include both internal storage units of any data processing device and external storage devices. The computer-readable storage medium is used to store the computer program and other programs and data required by the data processing device, and can also be used to temporarily store data that has been output or will be output.
[0068] It will be understood by those skilled in the art that the above descriptions are merely preferred examples of the invention and are not intended to limit the invention. Although the invention has been described in detail with reference to the foregoing examples, those skilled in the art can still modify the technical solutions described in the foregoing examples or make equivalent substitutions for some of the technical features. All modifications and equivalent substitutions made within the spirit and principles of the invention should be included within the scope of protection of the invention.
Claims
1. A bridge component shell element mesh generation method based on three-dimensional point cloud principal plane projection, characterized in that, Includes the following steps: S1: Project the 3D point cloud of the outer surface of the bridge component onto the main plane, and record each projection point in the projected point cloud. The coordinates; S2: Extract grid nodes layer by layer from the outside to the inside based on the projected point cloud to obtain the coordinates of all nodes; this is specifically achieved through the following sub-steps: S201: Set the mesh size to a×b, a≤b, and the total number of nodes to (a+1)×(b+1); each node is denoted according to its topological position in the mesh. ,in, , Divide all nodes into m layers from the outside in. The outermost layer includes 4 corner points and 2a+2b-4 edge points. The (a-1)×(b-1) nodes in the remaining layers are all internal points. S202: Based on the concave hull of the projected point cloud, extract the coordinates of 4 corner points and 2a+2b-4 edge points from the projected point cloud; S203: Extract the coordinates of the nodes in the innermost layer from the projected point cloud based on the coordinates of the outermost nodes and the shape of the enclosed polygon. S204: Extract the coordinates of each node from the outside to the inside according to the method of S203, until the extraction of all m layers of nodes is completed; S3: Node-based The parameter coordinates [j,k] are used to connect four adjacent nodes to create a planar quadrilateral element, thus achieving mesh generation; S4: Back projection of planar nodes and shell elements: based on the coordinate mapping relationship of node parameters Planar nodes Recovering the planar quadrilateral element to its coordinates in three-dimensional space And shell unit connections.
2. The method for generating shell element meshes for bridge components based on principal plane projection of three-dimensional point clouds according to claim 1, characterized in that, S1 is specifically implemented through the following sub-steps: S101: For any 3D point cloud on the outer surface of a bridge component, calculate its 3D centroid. and three main directions , , ; S102: Based on the three-dimensional centroid As a point within the principal plane, with the third principal direction The normal direction of the principal plane is used to construct the principal plane of the 3D point cloud. ; S103: On the main plane Internally, it is constructed with a three-dimensional centroid. Taking the origin as the starting point and the first principal direction as the starting point With the u-axis as the axis and the second principal direction as the axis, A Cartesian coordinate system for the v-axis; S104: Calculate any point in a 3D point cloud On the main plane Coordinates of the projection point within .
3. The method for generating shell element meshes for bridge components based on the principal plane projection of three-dimensional point clouds according to claim 2, characterized in that, In step S101, the three principal directions are calculated using principal component analysis; in step S104, the coordinates of the projection points are... It is a point Vertical projection onto the principal plane Two-dimensional coordinates on the surface, where the coordinate values are... and pass and Calculate the index of the points retained before and after projection. Consistent.
4. The method for generating shell element meshes for bridge components based on principal plane projection of three-dimensional point clouds according to claim 1, characterized in that, In S2, the total number of layers m is obtained through... Calculate, where This indicates rounding up to the nearest integer.
5. The method for generating shell element meshes for bridge components based on principal plane projection of three-dimensional point clouds according to claim 1, characterized in that, In step S202, the concave hull of the projected point cloud is capable of enclosing all projected points. The smallest concave polygon is used to represent the outline of the projected point cloud. The extraction method of the four corner points is as follows: calculate the smallest bounding rectangle of the concave hull with the principal axis and the u axis at a 45° angle, and extract the vertices of the concave hull located on the four sides of the smallest bounding rectangle as corner points.
6. The method for generating shell element meshes for bridge components based on principal plane projection of three-dimensional point clouds according to claim 1, characterized in that, In step S202, the extraction of the 2a+2b-4 edge points specifically includes the following steps: S2021: Based on the four extracted corner points, divide the concave hull of the projected point cloud into four edge lines; S2022: For each edge line, a set of positioning points are set on the edge line according to the ratio of the total length of the edge line to the number of shell elements in that direction. S2023: Extract the actual point in the projected point cloud that is closest to each positioning point, and use it as the edge point corresponding to each edge line.
7. The method for generating shell element meshes for bridge components based on principal plane projection of three-dimensional point clouds according to claim 1, characterized in that, S203 is specifically implemented through the following sub-steps: S2031: Calculate the inward contraction direction of each point on a given edge, taking two reference directions on adjacent edges. , ,but The inward contraction direction is , ; S2032: Calculate the intersection direction, if... a If it is an odd number, then when At that time, the intersecting direction Calculated using equation ①, when At that time, the intersecting direction Calculate using equation ②; if a is even, then when At that time, the intersecting direction Calculated using equation ①, when At that time, the intersecting direction The middle node is calculated using equation ②. The two intersecting directions are calculated using equations ① and ②. and ; ① ; ② ; S2033: Calculate the internal positioning points, from both sides towards the middle; if a If it is an odd number, initially when or At that time, the coordinates of the internal positioning point , ;when At that time, the coordinates of the internal positioning point Calculated using equation ③, when hour, Calculated using formula ④; ③ ; ④ ; In the formula, This represents the intersection point formed by the trajectory of point P1 moving along direction D1 and the trajectory of point P2 moving along direction D2. If a is even, then when At that time, the coordinates of the positioning point The calculation method is the same as for odd numbers; when j=(a+2) / 2, the coordinates of the two internal positioning points can be obtained through equations ③ and ④ respectively. and Then the coordinates of the final internal positioning point are taken as and The midpoint, i.e. ; S2034: Following the steps of S2031 to S2033, calculate the internal positioning points of the edge points on the other three edge lines after shrinking them inward by one layer, merge the internal positioning points with the same coordinates, and extract the actual points in the projected point cloud that are closest to each internal positioning point as the precise coordinates of the corresponding nodes in that layer.
8. The method for generating shell element meshes for bridge components based on principal plane projection of three-dimensional point clouds according to claim 1, characterized in that, In step S204, when extracting the innermost node, if 'a' is even, then the node is used. and Two sets of internal positioning points were obtained by performing inward contraction calculations, respectively. and The final coordinates of the internal positioning points are taken as follows: and The midpoint.
9. The method for generating shell element meshes for bridge components based on principal plane projection of three-dimensional point clouds according to claim 1, characterized in that, S3 is specifically implemented through the following sub-steps: S301: For each extracted node Assign it the corresponding parameter coordinates [j,k]. The mapping relationship between these parameter coordinates and the original projection point index i is as follows: ; S302: Connect four adjacent nodes clockwise or counterclockwise to create a planar quadrilateral element. A planar quadrilateral element created by clockwise connections is represented as: The planar quadrilateral element created by connecting counterclockwise is represented as: .
10. A device for generating shell element meshes for bridge components based on principal plane projection of three-dimensional point clouds, characterized in that, The device includes a memory and one or more processors, wherein the memory stores executable code, and the one or more processors execute the executable code to implement the method for generating shell element meshes for bridge components based on the principal plane projection of three-dimensional point clouds as described in any one of claims 1 to 9.
