A virtual pre-assembly method for shield segments using laser point cloud technology

Abstract

The present invention discloses a virtual pre-assembly method for shield segments using laser point cloud technology, comprising the following steps: obtaining segment point cloud data of different forms, obtaining characteristic dimensions of the segments based on the pre-processed segment point cloud data after pre-processing; assembling the segments based on a plurality of pre-processed segment point clouds of different forms based on a single-ring segment standard model, extracting the calculated positions of vertical gaps between the single-ring segments, and obtaining characteristic dimensions between the single-ring segments; assembling the pre-processed segment point clouds of different forms based on a three-ring segment standard model, extracting the calculated positions of vertical and horizontal gaps between the three-ring segments, and obtaining characteristic dimensions between the three-ring segments; judging whether the automated assembly has achieved a preset accuracy based on the segment characteristic dimensions, and if so, using equipment to scan the three-ring scanning data of the point cloud of the segments to be assembled to assemble the segments to be assembled, and measuring the segment gaps after assembly to complete the pre-assembly of the segments.

Description

Technical Field

[0001] The present invention relates to the field of tunnel construction, and in particular to a virtual pre-assembly method for shield tunnel segments using laser point cloud technology. Background Art

[0002] Shield tunneling is widely used in modern subway construction projects due to its high degree of automation, precise construction, fast excavation speed, and environmental adaptability. A significant advantage of this technology is the efficient assembly of precast shield concrete segments, which primarily form circular linings known as segment rings. However, numerous tunnel projects have proven that quality control during the shield segment assembly process is extremely challenging, with issues such as segment ring misalignment, deformation, and even leakage occurring. Timely and accurate inspection of segment assembly is crucial to ensure uniform tunnel appearance and avoid safety risks throughout the tunnel's service life.

[0003] According to traditional testing methods, pipe segments must undergo a three-ring pre-assembly test before leaving the factory to check whether the dimensional accuracy and shape and position deviation of the pipe segments meet the design and specification requirements. However, on-site assembly is difficult, inefficient, and the test data is incomplete. It also requires operators, takes up a lot of space, and increases resource investment. Therefore, developing a virtual assembly technology to replace physical pre-assembly is a new development trend. Summary of the Invention

[0004] The purpose of the present invention is to overcome the deficiencies of the prior art and provide a method for virtual pre-assembly of shield segments using laser point cloud technology, comprising the following steps:

[0005] Obtain segment point cloud data of different forms and pre-process the segment point cloud data. After pre-processing, convert the segment point cloud data of different forms into standard part point cloud data with the centroid located at the origin and the main direction consistent. Obtain the characteristic dimensions of the segment based on the pre-processed segment point cloud data.

[0006] Based on the standard model of a single-ring segment, multiple pre-processed segment point clouds of different shapes are assembled to extract the calculated positions of the vertical gaps between the single-ring segments and obtain the characteristic dimensions between the single-ring segments.

[0007] Based on the standard model of three-ring segments, the pre-processed segment point clouds of different shapes are assembled to extract the calculated positions of the vertical and horizontal gaps between the three-ring segments and obtain the characteristic dimensions between the three-ring segments.

[0008] Whether the automated assembly has achieved the preset accuracy is determined based on the characteristic dimensions of the segments. If so, the equipment is used to scan the three-ring scanning data of the point cloud of the segments to be assembled, and the segments to be assembled are assembled. After the assembly is completed, the segment gap is measured to complete the pre-assembly of the segments.

[0009] Furthermore, after the automated assembly is completed, a feature point data set is established through a KDtree search algorithm, and the characteristic dimensions of the segment are obtained through the relationship between points in the feature point data set, including:

[0010] S1, align point cloud B with point cloud A, where point cloud A is the point cloud data converted from the CAD standard model of the segment, and point cloud B is the point cloud data generated by scanning the segment;

[0011] The cross product of two non-parallel vectors in a plane is the normal vector of the plane. According to the order of multiplying the two vectors, two normal vectors with opposite directions are calculated, which can be expressed as follows:

[0012] Two non-parallel normal vectors in the plane: a, b

[0013] a×b=c

[0014] c is the normal vector, get the normal vector c of point cloud A and point cloud B respectively A and c B , compare them, and use the highest correlation value as the registration principle for registration;

[0015] S2, create kdtree t1 of point cloud B for fast search of points,

[0016] Select (x; y; z) (1) As the coordinate axis, all the data in the training set (x; y; z) (1) The median of the coordinates is used as the dividing point to cut the super rectangular area into two sub-areas; the dividing point is used as the root node, and the root node generates left and right child nodes with a depth of 1. The left node corresponds to (x; y; z) (1) The coordinates are less than the split point, and the right node corresponds to (x; y; z) (1) The coordinates are greater than the split point;

[0017] For a node with depth j, select (x; y; z) (1) To split the coordinate axis, l = j (mod k) + 1 is the training data (x; y; z) in the node area (1) The median of the coordinates is used as the dividing point to divide the region into two sub-regions, and generate left and right child nodes with a depth of j+1. The left node corresponds to (x; y; z) (1) The coordinates are less than the split point, and the right node corresponds to (x; y; z) (1) The coordinate is greater than the split point; stop when there is no data in the two sub-areas, and complete the construction of the point cloud data tree;

[0018] S3, calculate the unit vector v1 where p2>p1, and search downward based on the coordinates of the point cloud p1 of the standard model A and the nodes of the kdtree, that is, if the node of the tree is (x; y; z) (1) = c, then if the coordinate of p1 is less than c, go to the left child node, otherwise go to the right child node;

[0019] When S4 reaches a leaf node, it marks it as visited. If there are less than k points in the point set S, the node is added to S. If the point set S is not empty and the distance between the current node and point p1 is less than the longest distance in S, the point farthest from p1 in S is replaced by the current node. If the current node is not the root node, execute S3. Otherwise, obtain the unit vector v1:

[0020] S4, set the search radius r, continue to use the search algorithm, find all points in t1 with a distance less than r from p1, get the point set S1, traverse S1, set any point in it as p s1 , calculate each p2>p s1 The unit vector v s1 ;

[0021] S5, set the threshold δ, if v s1 The dot product of and v1 ≥ (1-δ), then the current p s1 Save to point set s v1 In , traverse s v1 , find the point p that is farthest from p1 1b , as the corresponding point of p1 in point cloud B, similarly we get p 2b ;

[0022] S6, calculate p 1b and p 2b The distance d 12 As the characteristic size f.

[0023] Furthermore, the unit vector v1 is:

[0024]

[0025] p1=(p 1x , p 1y , p 1z ) is the specific point p1 of the standard model A point cloud, (x1, y1, z1) is the KDtree node 1 of the scanned model B point cloud, (x e ,y e , z e ) is the current KDtree node of the scan model B.

[0026] Furthermore, the unit vector v s1 for:

[0027]

[0028] P2=(p 2x , p 2y , p 2z is a specific point p2 of the point cloud of the standard model A; p sl =(p s1x , p s1y , p s1z ) is the coordinate of any point in the point set S1; (x2, y2, z2) is the KDtree node 2 of the scan model B; (x s1 ,y s1 , z s1 ) is the same as p sl The corresponding KDtree node of scan model B.

[0029] Furthermore, the point set s v1 for:

[0030]

[0031] s v1 To traverse S1, each p2>p s1 The unit vector when p2>p1, v1 is the unit vector when p2>p1.

[0032] Furthermore, the calculation p 1b and p 2b The distance d 12 As the characteristic size f:

[0033]

[0034] According to the method for virtual pre-assembly of shield segments using laser point cloud technology, a virtual pre-assembly platform for shield segments using laser point cloud technology includes a data processing module, a segment data acquisition module, a segment assembly detection module, and a segment virtual assembly module; the segment data acquisition module, the segment assembly detection module, and the segment virtual assembly module are respectively connected to the data processing module;

[0035] The segment data acquisition module is used to collect point cloud data of the segments to be assembled, and the segment virtual assembly module is used to virtually assemble the segments; the segment assembly detection module is used to perform digital quality inspection on the segment laser point cloud model; multiple segments are assembled into a ring, the longitudinal seam gap between the segments in the ring and the circumferential seam gap between the rings are measured, and a corresponding pre-assembly monitoring report is given.

[0036] The beneficial effects of the present invention are: the present invention can realize the automatic assembly of incomplete pipe segments and automatic detection of characteristic dimensions, and can also realize automatic detection of pipe segment gaps of incomplete pipe segments, which can greatly improve the accuracy and efficiency of pipe segment assembly, effectively save labor and machinery costs, and avoid safety risks. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 It is a virtual pre-assembly method for shield segments using laser point cloud technology;

[0038] Figure 2 It is a virtual pre-assembly platform for shield segments using laser point cloud technology;

[0039] Figure 3 This is a schematic diagram of the segment feature size detection process;

[0040] Figure 4 Schematic diagram of the automatic measurement process of segment characteristic dimensions. DETAILED DESCRIPTION

[0041] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings, but the protection scope of the present invention is not limited to the following.

[0042] In order to make the objectives, technical solutions, and advantages of the present invention more clearly understood, the present invention is further described in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only intended to explain the present invention and are not intended to limit the present invention. That is, the embodiments described herein are only some embodiments of the present invention, not all embodiments. Generally, the components of the embodiments of the present invention described and illustrated in the drawings herein may be arranged and designed in various different configurations.

[0043] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making any creative work shall fall within the scope of protection of the present invention. It should be noted that relational terms such as "first" and "second" are merely used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations.

[0044] Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not preclude the presence of additional identical elements in the process, method, article, or apparatus that includes the element.

[0045] The features and performance of the present invention are further described in detail below with reference to the embodiments.

[0046] like Figure 1 As shown in FIG, a virtual pre-assembly method of shield segments using laser point cloud technology includes the following steps:

[0047] Obtaining segment point cloud data of different forms and preprocessing the segment point cloud data of different forms respectively. After preprocessing, converting the segment point cloud data of different forms into standard part point cloud data with the centroid located at the origin and the main direction consistent; obtaining the characteristic size of the segment based on the preprocessed segment point cloud data; the point cloud data preprocessing includes rotating and aligning the initial point cloud data, and performing noise reduction processing on the point cloud data to remove bad points and redundant data points;

[0048] Based on the standard model of a single-ring segment, multiple pre-processed segment point clouds of different shapes are assembled to extract the calculated positions of the vertical gaps between the single-ring segments and obtain the characteristic dimensions between the single-ring segments.

[0049] Based on the standard model of three-ring segments, the pre-processed segment point clouds of different shapes are assembled to extract the calculated positions of the vertical and horizontal gaps between the three-ring segments and obtain the characteristic dimensions between the three-ring segments.

[0050] Whether the automated assembly has achieved the preset accuracy is determined based on the characteristic dimensions of the segments. If so, the equipment is used to scan the three-ring scanning data of the point cloud of the segments to be assembled, and the segments to be assembled are assembled. After the assembly is completed, the segment gap is measured to complete the pre-assembly of the segments.

[0051] After the automated assembly is completed, a feature point data set is established using the KDtree search algorithm. The feature size of the segment is obtained through the relationship between points in the feature point data set, including:

[0052] S1, align point cloud B with point cloud A, where point cloud A is the point cloud data converted from the CAD standard model of the segment, and point cloud B is the point cloud data generated by scanning the segment;

[0053] The cross product of two non-parallel vectors in a plane is the normal vector of the plane. According to the order of multiplying the two vectors, two normal vectors with opposite directions are calculated, which can be expressed as follows:

[0054] Two non-parallel normal vectors in the plane: a, b

[0055] a×b=c

[0056] c is the normal vector, get the normal vector c of point cloud A and point cloud B respectively A and c B , compare them, and use the highest correlation value as the registration principle for registration;

[0057] S2, create kdtree t1 of point cloud B for fast search of points,

[0058] Select (x; y; z) (1) As the coordinate axis, all the data in the training set (x; y; z) (1) The median of the coordinates is used as the dividing point to cut the super rectangular area into two sub-areas; the dividing point is used as the root node, and the root node generates left and right child nodes with a depth of 1. The left node corresponds to (x; y; z) (1) The coordinates are less than the split point, and the right node corresponds to (x; y; z) (1) The coordinates are greater than the split point;

[0059] For a node with depth j, select (x; y; z) (1) To split the coordinate axis, l = j (mod k) + 1 is the training data (x; y; z) in the node area (1) The median of the coordinates is used as the dividing point to divide the region into two sub-regions, and generate left and right child nodes with a depth of j+1. The left node corresponds to (x; y; z) (1) The coordinates are less than the split point, and the right node corresponds to (x; y; z) (1) The coordinate is greater than the split point; stop when there is no data in the two sub-areas, and complete the construction of the point cloud data tree;

[0060] S3, calculate the unit vector v1 where p2>p1, and search downward based on the coordinates of the point cloud p1 of the standard model A and the nodes of the kdtree, that is, if the node of the tree is (x; y; z) (1) = c, then if the coordinate of p1 is less than c, go to the left child node, otherwise go to the right child node;

[0061] When S4 reaches a leaf node, it marks it as visited. If there are less than k points in the point set S, the node is added to S. If the point set S is not empty and the distance between the current node and point p1 is less than the longest distance in S, the point farthest from p1 in S is replaced by the current node. If the current node is not the root node, execute S3. Otherwise, obtain the unit vector v1:

[0062] S4, set the search radius r, continue to use the search algorithm, find all points in t1 with a distance less than r from p1, get the point set S1, traverse S1, set any point in it as p s1 , calculate each p2>p s1 The unit vector v s1 ;

[0063] S5, set the threshold δ, if v s1 The dot product of and v1 ≥ (1-δ), then the current p s1 Save to point set s v1 In , traverse s v1 , find the point p that is farthest from p1 1b , as the corresponding point of p1 in point cloud B, similarly we get p 2b ;

[0064] S6, calculate p 1b and p 2b The distance d 12 As the characteristic size f.

[0065] The unit vector v1 is:

[0066]

[0067] p1=(p 1x , p 1y , p 1z ) is the specific point p1 of the standard model A point cloud, (x1, y1, z1) is the KDtree node 1 of the scanned model B point cloud, (x e ,y e , z e ) is the current KDtree node of the scan model B.

[0068] The unit vector v s1 for:

[0069]

[0070] P2=(p 2x , p 2y , p 2z is a specific point p2 of the point cloud of the standard model A; p sl =(p s1x , p s1y, p s1z ) is the coordinate of any point in the point set S1; (x2, y2, z2) is the KDtree node 2 of the scan model B; (x s1 ,y s1 , z s1 ) is the same as p sl The corresponding KDtree node of scan model B.

[0071] The point set s v1 for:

[0072]

[0073] s v1 To traverse S1, each p2>p s1 The unit vector when p2>p1, v1 is the unit vector when p2>p1.

[0074] The calculation p 1b and p 2b The distance d 12 As the characteristic size f:

[0075]

[0076] According to the method for virtual pre-assembly of shield segments using laser point cloud technology, a virtual pre-assembly platform for shield segments using laser point cloud technology is characterized by comprising a data processing module, a segment data acquisition module, a segment assembly detection module, and a segment virtual assembly module; the segment data acquisition module, the segment assembly detection module, and the segment virtual assembly module are respectively connected to the data processing module;

[0077] Specifically, there is a standard model point cloud A. Assume that there are two specific points p1 and p2 on A. The distance between the two points is a characteristic size f of the standard part represented by the point cloud.

[0078] Objective: For any identical scan point cloud B, calculate the key eigenvalue f1 of B.

[0079] Algorithm flow:

[0080] 1. Align point cloud B with point cloud A. Point cloud A is the point cloud data converted from the standard CAD model of the segment, and point cloud B is the point cloud data generated by scanning the actual segment. The point cloud registration is completed by comparing the normal vectors of the special front face.

[0081] That is, the cross product (also called "outer product, vector product, cross product") of two non-parallel vectors in a plane is calculated as the normal vector of the plane. Based on the order in which the two vectors are multiplied, two normal vectors with opposite directions are calculated. This can be expressed as:

[0082] Two non-parallel normal vectors in the plane: a, b (c is the normal vector)

[0083] a×b=c

[0084] Get the normal vector c of point cloud A and B respectively A and c B , compare them, and use the highest correlation value as the registration principle for registration.

[0085] 2. Create a KDtree t1 of point cloud B for fast point search.

[0086] 2.1 Select (x; y; z) (1) As the coordinate axis, all the data in the training set (x; y; z) (1) The median of the coordinates is used as the dividing point to cut the hyperrectangular area into two sub-areas. The dividing point is used as the root node, and the root node generates left and right child nodes with a depth of 1. The left node corresponds to (x; y; z) (1) The coordinates are less than the split point, and the right node corresponds to (x; y; z) (1) Coordinates greater than the split point

[0087] 2.2 For the node with depth j, select (x; y; z) (1) To split the coordinate axis, l = j (mod k) + 1 is the training data (x; y; z) in the node area (1) The median of the coordinates is used as the dividing point to divide the region into two sub-regions, and generate left and right child nodes with a depth of j+1. The left node corresponds to (x; y; z) (1) The coordinates are less than the split point, and the right node corresponds to (x; y; z) (1) Coordinates greater than the split point

[0088] Repeat 2.2 until there is no data in the two sub-regions, thus completing the construction of the point cloud data tree.

[0089] 3. Calculate the unit vector v1 where p2>p1

[0090] 3.1 Search downwards based on the coordinates of point cloud p1 located in the standard model A and the nodes of kdtree, that is, if the node of the tree is (x; y; z) (1) = c, then if the coordinate of p1 is less than c, go to the left child node, otherwise go to the right child node.

[0091] 3.2 When a leaf node is reached, it is marked as visited. If the point set S contains fewer than k points, the node is added to S. If the point set S is not empty and the distance between the current node and point p1 is less than the longest distance in S, the point farthest from p1 in S is replaced by the current node.

[0092] 3.3 If the current node is not the root node, execute 3.1; otherwise, end the algorithm to obtain the unit vector v1:

[0093]

[0094] p1=(p 1x , p 1y , p 1z )————Specific point p1 of the standard model A point cloud

[0095] (x1, y1, z1)————Scan KDtree node 1 of model B

[0096] (x e ,y e , z e )————Scan the current KDtree node of model B

[0097] 4. Set the search radius r, continue to use the search algorithm, find all points in t1 whose distance to p1 is less than r, and get the point set S1

[0098] 5. Traverse S1 and let any point be p s1 , calculate each p2>p s1 The unit vector v s1

[0099]

[0100] P2=(p 2x , p 2y , p 2z )————Specific point p2 of the standard model A point cloud

[0101] p sl =(p s1x , p s1y , p s1z )————the coordinates of any point in the point set S1

[0102] (x2, y2, z2)————Scan KDtree node 2 of model B

[0103] (x s1 ,y s1 , z s1 )————with p sl Corresponding KDtree node of scan model B

[0104] 6. Set the threshold δ, if v s1 The dot product of v1 is ≥ (1-δ), that is, the angle is close to 0 degrees, then the current p s1 Save to point set s v1 middle

[0105]

[0106] v s1 ————Traverse S1, each p2>p s1 The unit vector at

[0107] v1————unit vector when p2>p1

[0108] 7. Traverse s v1 , find the point p that is farthest from p1 1b , as the corresponding point of p1 in point cloud B

[0109] 8. Repeat steps 3-8 to calculate p 2b

[0110] 9. Calculate p 1b and p 2b The distance d 12 As the characteristic size f

[0111]

[0112] The segment data acquisition module is used to collect point cloud data of the segments to be assembled, and the segment virtual assembly module is used to virtually assemble the segments; the segment assembly detection module is used to perform digital quality inspection on the segment laser point cloud model; multiple segments are assembled into a ring, the longitudinal seam gap between the segments in the ring and the circumferential seam gap between the rings are measured, and a corresponding pre-assembly monitoring report is given.

[0113] The foregoing description is merely a preferred embodiment of the present invention. It should be understood that the present invention is not limited to the form disclosed herein and should not be construed as excluding other embodiments. Rather, the present invention can be used in various other combinations, modifications, and environments and can be modified within the scope of the concept described herein through the above teachings or techniques or knowledge in the relevant field. Modifications and variations made by those skilled in the art that do not depart from the spirit and scope of the present invention are intended to be protected by the appended claims.

Claims

1. A virtual pre-assembly method for shield segments using laser point cloud technology, characterized in that: The process includes the following steps: Obtain segment point cloud data of different forms and pre-process the segment point cloud data of different forms respectively. After pre-processing, convert the segment point cloud data of different forms into standard part point cloud data with the centroid located at the origin and the main direction consistent; According to the pre-processed segment point cloud data, the characteristic size of the segment is obtained; Based on the standard model of a single-ring segment, multiple pre-processed segment point clouds of different shapes are assembled to extract the calculated positions of the vertical gaps between the single-ring segments and obtain the characteristic dimensions between the single-ring segments. Based on the standard model of three-ring segments, the pre-processed segment point clouds of different shapes are assembled to extract the calculated positions of the vertical and horizontal gaps between the three-ring segments and obtain the characteristic dimensions between the three-ring segments. Whether the automated assembly has achieved the preset accuracy is determined based on the characteristic dimensions of the segments. If so, the equipment is used to scan the three-ring scanning data of the point cloud of the segments to be assembled, and the segments to be assembled are assembled. After the assembly is completed, the segment gap is measured to complete the pre-assembly of the segments.

2. The virtual pre-assembly method of shield segments using laser point cloud technology according to claim 1 is characterized in that: After the automated assembly is completed, a feature point data set is established using the KDtree search algorithm. The feature size of the segment is obtained through the relationship between points in the feature point data set, including: S1, align point cloud B with point cloud A, where point cloud A is the point cloud data converted from the CAD standard model of the segment, and point cloud B is the point cloud data generated by scanning the segment; The cross product of two non-parallel vectors in a plane is the normal vector of the plane. According to the order of multiplying the two vectors, two normal vectors with opposite directions are calculated, which can be expressed as follows: Two non-parallel normal vectors in the plane: a, b a×b=c c is the normal vector, get the normal vector c of point cloud A and point cloud B respectively A and c B , compare them, and use the highest correlation value as the registration principle for registration; S2, create kdtree t1 of point cloud B for fast search of points, Select (x; y; z) (1) As the coordinate axis, all the data in the training set (x; y; z) (1) The median of the coordinates is used as the dividing point to cut the super rectangular area into two sub-areas; the dividing point is used as the root node, and the root node generates left and right child nodes with a depth of 1. The left node corresponds to (x; y; z) (1) The coordinates are less than the split point, and the right node corresponds to (x; y; z) (1) The coordinates are greater than the split point; For a node with depth j, select (x; y; z) (1) To split the coordinate axis, l = j (mod k) + 1 is the training data (x; y; z) in the node area (1) The median of the coordinates is used as the dividing point to divide the region into two sub-regions, and generate left and right child nodes with a depth of j+1; the left node corresponds to (x; y; z) (1) The coordinates are less than the split point, and the right node corresponds to (x; y; z) (1) The coordinate is greater than the split point; stop when there is no data in the two sub-areas, and complete the construction of the point cloud data tree; S3, calculate the unit vector v1 where p2>p1, and search downward based on the coordinates of the point cloud p1 of the standard model A and the nodes of the kdtree, that is, if the node of the tree is (x; y; z) (1) = c, then if the coordinate of p1 is less than c, go to the left child node, otherwise go to the right child node; When S4 reaches a leaf node, it marks it as visited. If there are less than k points in the point set S, the node is added to S. If the point set S is not empty and the distance between the current node and point p1 is less than the longest distance in S, the point farthest from p1 in S is replaced by the current node. If the current node is not the root node, execute S3. Otherwise, obtain the unit vector v1: S4, set the search radius r, continue to use the search algorithm, find all points in t1 with a distance less than r from p1, get the point set S1, traverse S1, set any point in it as p s1 , calculate each p2>p s1 The unit vector v s1 ; S5, set the threshold δ, if v s1 The dot product of and v1 ≥ (1-δ), then the current p s1 Save to point set s v1 In , traverse s v1 , find the point p that is farthest from p1 1b , as the corresponding point of p1 in point cloud B, similarly we get p 2b ; S6, calculate p 1b and p 2b The distance d 12 As the characteristic size f.

3. The virtual pre-assembly method of shield segments using laser point cloud technology according to claim 2 is characterized in that: The unit vector v1 is: p1=(p 1x , p 1y , p 1z ) is the specific point p1 of the standard model A point cloud, (x1, y1, z1) is the KDtree node 1 of the scanned model B point cloud, (x e ,y e , z e ) is the current KDtree node of the scan model B.

4. The virtual pre-assembly method of shield segments using laser point cloud technology according to claim 3 is characterized in that: The unit vector v s1 for: P2=(p 2x , p 2y , p 2z is a specific point p2 of the point cloud of the standard model A; p sl =(p s1x , p s1y , p s1z ) is the coordinate of any point in the point set S1; (x2, y2, z2) is the KDtree node 2 of the scan model B; (x s1 ,y s1 , z s1 ) is the same as p sl The corresponding KDtree node of scan model B.

5. The method for virtual pre-assembly of shield segments using laser point cloud technology according to claim 4 is characterized in that: The point set s v1 for: s v1 To traverse S1, each p2>p s1 The unit vector when p2>p1, v1 is the unit vector when p2>p1.

6. The method for virtual pre-assembly of shield segments using laser point cloud technology according to claim 5, characterized in that: The calculation p 1b and p 2b The distance d 12 As the characteristic size f:

7. A shield segment virtual pre-assembly method using laser point cloud technology and a shield segment virtual pre-assembly platform using laser point cloud technology according to any one of claims 1 to 6, characterized in that: It includes a data processing module, a segment data acquisition module, a segment assembly detection module, and a segment virtual assembly module; the segment data acquisition module, the segment assembly detection module, and the segment virtual assembly module are respectively connected to the data processing module; The segment data acquisition module is used to collect point cloud data of the segments to be assembled, and the segment virtual assembly module is used to virtually assemble the segments; the segment assembly detection module is used to perform digital quality inspection on the segment laser point cloud model; multiple segments are assembled into a ring, the longitudinal seam gap between the segments in the ring and the circumferential seam gap between the rings are measured, and a corresponding pre-assembly monitoring report is given.

Images