Automatic assembling method for shield segment based on multi-bolt hole space constraint

Abstract

This invention belongs to the field of shield tunnel construction technology and discloses an automatic assembly method for shield tunnel segments based on spatial constraints of multiple bolt holes. The method includes the following steps: S1, collecting the inner wall point cloud of the bolt holes that correspond one-to-one between the segments to be assembled and the installed segments, and are used for splicing; S2, processing the collected inner wall point cloud of the bolt holes; S3, defining the pose transformation matrix as T; S4, decomposing the three-dimensional coordinates of the centers of at least three pairs of bolt holes in the collected bolt holes using SVD; S5, outputting the obtained optimal pose transformation matrix; S6, sending the obtained optimal pose transformation matrix to the segment assembly machine for segment assembly operation. This invention directly utilizes the inherent, high-precision machined standard bolt holes of the tunnel segments as positioning features, without the need for any additional artificial markers. It is low-cost, highly adaptable, and fully utilizes the complete geometric information of the bolt holes, achieving a solution accuracy and stability far exceeding traditional point cloud registration methods.

Description

Technical Field

[0001] This invention belongs to the field of shield tunnel construction technology, and specifically relates to an automatic assembly method for shield tunnel segments based on the spatial constraints of multiple bolt holes. Background Technology

[0002] In shield tunnel construction, segment assembly is a crucial process determining the tunnel's forming quality and structural safety. Traditional assembly methods primarily rely on manual operation or mechanical guidance based on single features (such as the mating surface or a few locating pins), which suffers from the following technical bottlenecks: First, manual assembly is characterized by low accuracy and slow efficiency, and is greatly affected by the working environment and personnel experience, making it difficult to meet the high standards required for modern tunnel construction. Second, existing automated methods typically only utilize the outer contour of the segments or a few locating holes for registration, neglecting the essential characteristic of mechanical connection between segments via multiple bolt holes. This results in insufficient pose calculation constraints and weak anti-interference capabilities, making it prone to mismatches or convergence to local optima under actual interference such as hole manufacturing errors and point cloud noise. Furthermore, bolt holes not only provide constraints on the center position, but their axial direction (normal vector) is also a key geometric element determining whether bolts can be successfully inserted. Existing technologies generally lack effective extraction and constraint utilization of the direction vector. Therefore, we propose an automated shield segment assembly method based on multi-bolt hole spatial constraints to address the above problems. Summary of the Invention

[0003] To address the aforementioned problems, this invention provides an automatic assembly method for tunnel segments based on spatial constraints of multiple bolt holes. This method solves the problems of traditional segment assembly relying on manual labor, having low precision, insufficient constraints, and being unable to simultaneously ensure the alignment of bolt hole centers and orientation.

[0004] This invention is achieved through the following scheme: an automatic assembly method for tunnel lining segments based on spatial constraints of multiple bolt holes, comprising the following steps: S1. Collect the inner wall point cloud of the bolt holes that correspond one-to-one between the pipe segments to be assembled and the installed pipe segments and are used for splicing, and sort each pair of bolt holes collected. S2. The collected point cloud of the inner wall of the bolt holes is processed to extract the features of each bolt hole, thereby obtaining the features of the bolt holes on the segments to be assembled and the bolt holes on the installed segments. The features include the three-dimensional coordinates O of the center and the direction vector V perpendicular to the plane of the bolt hole port. The features of the bolt holes on the segments to be assembled are as follows: The bolt holes on the installed segments are characterized as follows: ; S3. Define the pose transformation matrix as T, where T includes a rotation matrix R and a translation vector t. The centers of each pair of bolt holes coincide. If the direction vectors of each pair of bolt holes are parallel and in the same direction, then The total error function, consisting of the center position error and the direction vector error, is defined as follows: The The formula is: ; Where α and β are the weighting coefficients of the center position error and the direction vector error, respectively, i is the index, i=1,2……m+n, representing the i-th pair, and m+n is a natural number greater than 2; S4. Decompose the three-dimensional coordinates of the centers of at least three pairs of bolt holes in the collected bolt holes using SVD to calculate the initial values ​​of a rotation matrix R and a translation vector t. Substitute these initial values ​​of R and t into... In this study, a nonlinear optimization method is used to minimize the total error function with rotation matrix R and translation vector t as variables. ; S5. The total error function obtained from at least two consecutive iterations When the value is less than the set threshold or the maximum number of iterations is reached, the algorithm converges and outputs the optimal pose transformation matrix. ; S6. Obtain the optimal pose transformation matrix The data is sent to the kinematic model of the segment assembly machine. After the target motion quantities of each joint of the assembly machine are analyzed, the automatic segment assembly operation is performed.

[0005] A further improvement of the automatic shield tunnel segment assembly method based on multi-bolt hole spatial constraints of the present invention is that, before performing step S6, the following step is included: Construct a 3D mapping model of the installed tunnel segments and the tunnel segments to be assembled; The obtained optimal pose transformation matrix The tube segments to be assembled are applied to the 3D mapping model to generate the virtual pose of the tube segments after assembly. The pose of the virtual assembled tunnel segments is detected to verify the optimal pose transformation matrix. Does it meet the requirements? When performing step S6, the optimal pose transformation matrix that meets the requirements will be obtained. Send to the segment assembly machine.

[0006] A further improvement of the automatic assembly method for tunnel segments based on multi-bolt hole spatial constraints of the present invention is that the detection method for detecting the virtual pose of the segments to be assembled includes A and B; The detection method A is to detect whether there is model penetration between the virtual assembled tunnel segment to be assembled and the installed tunnel segment and the shield tail structure of the tunnel boring machine; The detection method B is: to detect whether the virtual installation path of the bolts, generated based on the center and direction vector of the bolt holes on the pipe segment to be assembled after virtual assembly, interferes with the already installed pipe segment; The verification of the optimal pose transformation matrix The verification method for whether the requirements are met is as follows: if the results of all detection methods are negative, then the verification is considered to meet the requirements; if the result of any detection method is positive, then the verification is considered to fail to meet the requirements.

[0007] A further improvement of the automatic shield tunnel segment assembly method based on multi-bolt hole spatial constraints of the present invention is that the detection method further includes: Calculate the residual distance between the center positions of each pair of bolt holes. ,and The calculation formula is: ; Calculate the residual angle between the direction vectors of each pair of bolt holes. ,and The calculation formula is: ; Compare the calculated residual distance value of each pair of bolt holes with the set position tolerance threshold, and compare the residual included angle value with the set angle tolerance threshold; Determine whether the residual distance values ​​of all bolt holes are greater than the set position tolerance threshold and whether the residual included angle values ​​are greater than the set angle tolerance threshold.

[0008] A further improvement of the automatic shield tunnel segment assembly method based on multi-bolt hole spatial constraints in this invention lies in the fact that, during the verification of the optimal pose transformation matrix... If the required steps are not followed, an alarm will be triggered to prompt the operator to check.

[0009] A further improvement of the automatic assembly method for tunnel segments based on multi-bolt hole spatial constraints of the present invention is that the installed segments include the segments of the previous ring and the adjacent segments of the same ring.

[0010] A further improvement of the automatic shield tunnel segment assembly method based on multi-bolt hole spatial constraints of the present invention is that, when performing step S2, it further includes the step of: The point cloud of the inner wall of the bolt hole is approximated as a spatial circular plane by using a cylindrical or toroidal fitting algorithm. The three-dimensional coordinates O of the center of the circular plane on the outer surface of the corresponding tube segment are calculated using the least squares method. Calculate the direction vector V of the circular plane based on the circular plane.

[0011] A further improvement of the automatic assembly method for tunnel segments based on multi-bolt hole spatial constraints in this invention is that the nonlinear optimization method is either the Levenberg-Marquardt algorithm or the Gauss-Newton method. Compared with the prior art, the beneficial effects of the present invention are as follows: This invention utilizes the inherent, high-precision machined standard bolt holes of the pipe segments as positioning features, without the need for any additional artificial markers. It is low-cost, highly adaptable, and makes full use of the complete geometric information of the bolt holes. The solution accuracy and stability far exceed those of traditional point cloud registration methods. Attached Figure Description

[0012] Figure 1 A schematic diagram of the assembly of the tube segments to be assembled according to the present invention is shown.

[0013] In the diagram: 1. The previous ring segment; 2. Adjacent segments in the same ring; 3. Segment to be assembled; 4. Bolt holes. Detailed Implementation

[0014] To address the problems of traditional tunnel segment assembly, which relies on manual labor, has low precision, insufficient constraints, and cannot simultaneously ensure the alignment of bolt hole centers and directions, this invention provides an automatic tunnel segment assembly method based on multi-bolt hole spatial constraints. The following detailed description, in conjunction with accompanying drawings, illustrates this automatic tunnel segment assembly method based on multi-bolt hole spatial constraints.

[0015] See Figure 1 As shown, an automatic assembly method for tunnel lining segments based on spatial constraints of multiple bolt holes includes the following steps: S1. Collect (using a 3D laser scanner or depth camera) the inner wall point cloud of the bolt holes 4 that correspond one-to-one between the segments to be assembled and the installed segments and are used for splicing, and sort each pair of bolt holes 4 collected. S2. The collected point cloud of the inner wall of the bolt holes 4 is processed to extract the features of each bolt hole 4, thereby obtaining the features of the bolt holes 4 on the segment 3 to be assembled and the features of the bolt holes 4 on the installed segments. The features include the three-dimensional coordinates O of the center and the direction vector (unit normal vector) V perpendicular to the plane of the bolt hole 4 port. The features of the bolt holes 4 on the segment 3 to be assembled are as follows: The bolt holes 4 on the installed segments are characterized as follows: ; S3. Define the pose transformation matrix as T, where T includes the rotation matrix R and the translation vector t. The centers of each pair of bolt holes 4 coincide. If the direction vectors of each pair of bolt holes 4 are parallel and in the same direction, then The total error function, consisting of the center position error and the direction vector error, is defined as follows: The The formula is: ; Where α and β are the weighting coefficients for the center position error and the direction vector error, respectively, used to balance the contributions of the two constraints. They can be manually adjusted according to the measurement accuracy, and it is recommended that they be 0.5 respectively. i is the serial number, i=1,2……m+n, which means the i-th pair, and m+n is a natural number greater than 2. S4. Decompose the three-dimensional coordinates of the centers of at least three pairs of bolt holes 4 in the collected bolt holes 4 using SVD, calculate the initial values ​​of a rotation matrix R and a translation vector t, and substitute these initial values ​​of R and t into... In this study, a nonlinear optimization method is used to minimize the total error function with rotation matrix R and translation vector t as variables. (The total error is minimized by continuously changing the values ​​of R and t and substituting them into the calculation.) S5. The total error function obtained from at least two consecutive iterations When the value is less than the set threshold or the maximum number of iterations (20 recommended) is reached, the algorithm converges and outputs the optimal pose transformation matrix. : ; S6. Obtain the optimal pose transformation matrix The data is sent to the kinematic model of the segment assembly machine. After the target motion quantities of each joint of the assembly machine are analyzed (the assembly machine is existing technology, and the operations performed by its assembly system based on the target motion quantities of each joint will not be described in detail here), the automatic segment assembly operation is performed.

[0016] By directly utilizing the inherent, high-precision machined standard bolt holes 4 of the tube segment as positioning features, without the need for any additional artificial markers, the method is low-cost, highly adaptable, and fully utilizes the complete geometric information of the bolt holes 4. The solution accuracy and stability far exceed those of traditional point cloud registration methods.

[0017] The procedure before step S6 includes the following steps: Construct a 3D mapping model of the installed tunnel segments and the tunnel segments to be assembled; The obtained optimal pose transformation matrix The virtual pose of the assembled tube segment 3 is generated by applying it to the 3D mapping model. The virtual pose of the assembled segment 3 is detected to verify the optimal pose transformation matrix. Does it meet the requirements? When performing step S6, the optimal pose transformation matrix that meets the requirements will be obtained. Send to the segment assembly machine.

[0018] The detection method for detecting the virtual assembly pose of the tube segment 3 to be assembled includes A and B; The detection method A is to detect whether there is model penetration between the virtual assembled segment 3 to be assembled and the already installed segments and the shield tail structure of the tunnel boring machine; The detection method B is as follows: detect whether the virtual installation path of the bolt, generated based on the center and direction vector of the bolt hole 4 on the virtually assembled pipe segment 3, interferes with the already installed pipe segment (and other structures); the purpose of this detection is to ensure that the bolt can be smoothly inserted. The verification of the optimal pose transformation matrix The verification method for whether the requirements are met is as follows: if the results of all detection methods are negative, then the verification is considered to meet the requirements; if the result of any detection method is positive, then the verification is considered to fail to meet the requirements.

[0019] The detection method further includes: Calculate the residual distance between the four center positions of each pair of bolt holes. ,and The calculation formula is: Calculate the residual angle between the four-direction vectors of each pair of bolt holes. ,and The calculation formula is: ; Compare the calculated residual distance value of each pair of bolt holes 4 with the set position tolerance threshold (1 mm recommended), and compare the residual included angle value with the set angle tolerance threshold (0.021° recommended); Determine whether the residual distance values ​​of all bolt holes 4 are greater than the set position tolerance threshold and whether the residual included angle values ​​are greater than the set angle tolerance threshold.

[0021] By adopting the above design, virtual assembly simulation, model penetration detection, installation path detection, residual distance detection, and residual angle detection are uniquely integrated into the pose calculation process, forming a closed loop of calculation-simulation-verification. This effectively avoids the direct execution of invalid or dangerous poses caused by algorithmic peculiarities, data noise, or complex working conditions, greatly improving the system's safety and first-time assembly success rate, and is a key guarantee for achieving fully automated assembly.

[0022] Among them, in performing the verification of the optimal pose transformation matrix If the required steps are not followed, an alarm will be triggered to prompt the operator to check.

[0023] The installed segments include the previous ring segment 1 and the adjacent segment 2 in the same ring.

[0024] Further, in step S3, m is the number of bolt holes 4 on the segment to be assembled and bolt holes 4 on the segment 1 of the previous ring, and n is the number of bolt holes 4 on the segment to be assembled and bolt holes 4 on the adjacent segment 2 of the same ring. The execution of step S2 also includes the following steps: The point cloud of the inner wall of bolt hole 4 is approximated as a spatial circular plane by using a cylindrical or toroidal fitting algorithm. The three-dimensional coordinates O(x,y,z) of the center of the circular plane on the outer surface of the corresponding tube segment are calculated using the least squares method. Calculate the direction vector V of the circular plane based on this plane. n x , n y , n z ).

[0025] By adopting the above design, the direction vector is perpendicular to the port plane of bolt hole 4 and points to the outside of the tube segment (i.e., the insertion direction of the bolt), representing the axial direction of bolt hole 4.

[0026] The nonlinear optimization method is either the Levenberg-Marquardt algorithm or the Gauss-Newton method. It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises 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.

[0027] The present invention has been described in detail above with reference to the accompanying drawings and embodiments. Those skilled in the art can make various modifications to the present invention based on the above description. Therefore, certain details in the embodiments should not be construed as limiting the present invention, and the scope of protection of the present invention shall be defined by the appended claims.

Claims

1. An automatic assembly method for tunnel lining segments based on spatial constraints of multiple bolt holes, characterized in that, Includes the following steps: S1. Collect the inner wall point cloud of the bolt holes that correspond one-to-one between the pipe segments to be assembled and the installed pipe segments and are used for splicing, and sort each pair of bolt holes collected. S2. The collected point cloud of the inner wall of the bolt holes is processed to extract the features of each bolt hole, thereby obtaining the features of the bolt holes on the segments to be assembled and the bolt holes on the installed segments. The features include the three-dimensional coordinates O of the center and the direction vector V perpendicular to the plane of the bolt hole port. The features of the bolt holes on the segments to be assembled are as follows: The bolt holes on the installed segments are characterized as follows: ; S3. Define the pose transformation matrix as T, where T includes a rotation matrix R and a translation vector t. The centers of each pair of bolt holes coincide. If the direction vectors of each pair of bolt holes are parallel and in the same direction, then The total error function, consisting of the center position error and the direction vector error, is defined as follows: The The formula is: ; Where α and β are the weighting coefficients of the center position error and the direction vector error, respectively, i is the index, i=1,2……m+n, representing the i-th pair, and m+n is a natural number greater than 2; S4. Decompose the three-dimensional coordinates of the centers of at least three pairs of bolt holes in the collected bolt holes using SVD to calculate the initial values ​​of a rotation matrix R and a translation vector t. Substitute these initial values ​​of R and t into... In this study, a nonlinear optimization method is used to minimize the total error function with rotation matrix R and translation vector t as variables. ; S5. The total error function obtained from at least two consecutive iterations When the value is less than the set threshold or the maximum number of iterations is reached, the algorithm converges and outputs the optimal pose transformation matrix. ; S6. Obtain the optimal pose transformation matrix The data is sent to the kinematic model of the segment assembly machine. After the target motion quantities of each joint of the assembly machine are analyzed, the automatic segment assembly operation is performed.

2. The automatic assembly method for tunnel segments based on multi-bolt hole spatial constraints as described in claim 1, characterized in that, The steps preceding step S6 are as follows: Construct a 3D mapping model of the installed tunnel segments and the tunnel segments to be assembled; The obtained optimal pose transformation matrix The tube segments to be assembled are applied to the 3D mapping model to generate the virtual pose of the tube segments after assembly. The pose of the virtual assembled tunnel segments is detected to verify the optimal pose transformation matrix. Does it meet the requirements? When performing step S6, the optimal pose transformation matrix that meets the requirements will be obtained. Send to the segment assembly machine.

3. The automatic assembly method for tunnel segments based on multi-bolt hole spatial constraints as described in claim 2, characterized in that, The detection method for detecting the pose of the virtual assembled segment to be assembled includes A and B; The detection method A is to detect whether there is model penetration between the virtual assembled tunnel segment to be assembled and the installed tunnel segment and the shield tail structure of the tunnel boring machine; The detection method B is: to detect whether the virtual installation path of the bolts, generated based on the center and direction vector of the bolt holes on the pipe segment to be assembled after virtual assembly, interferes with the already installed pipe segment; The verification of the optimal pose transformation matrix The verification method for whether the requirements are met is as follows: if the results of all detection methods are negative, then the verification is considered to meet the requirements; if the result of any detection method is positive, then the verification is considered to fail to meet the requirements.

4. The automatic assembly method for tunnel segments based on multi-bolt hole spatial constraints as described in claim 3, characterized in that, The detection method further includes: Calculate the residual distance between the center positions of each pair of bolt holes. ,and The calculation formula is: ; Calculate the residual angle between the direction vectors of each pair of bolt holes. ,and The calculation formula is: ; Compare the calculated residual distance value of each pair of bolt holes with the set position tolerance threshold, and compare the residual included angle value with the set angle tolerance threshold; Determine whether the residual distance values ​​of all bolt holes are greater than the set position tolerance threshold and whether the residual included angle values ​​are greater than the set angle tolerance threshold.

5. The automatic assembly method for tunnel segments based on multi-bolt hole spatial constraints as described in claim 4, characterized in that, Execute the verification of the optimal pose transformation matrix If the required steps are not followed, an alarm will be triggered to prompt the operator to check.

6. The automatic assembly method for tunnel segments based on multi-bolt hole spatial constraints as described in claim 1, characterized in that, The installed segments include the segments of the preceding ring and the adjacent segments of the same ring.

7. The automatic assembly method for tunnel segments based on multi-bolt hole spatial constraints as described in claim 1, characterized in that, When performing step S2, the following steps are also included: The point cloud of the inner wall of the bolt hole is approximated as a spatial circular plane by using a cylindrical or toroidal fitting algorithm. The three-dimensional coordinates O of the center of the circular plane on the outer surface of the corresponding tube segment are calculated using the least squares method. Calculate the direction vector V of the circular plane based on the circular plane.

8. The automatic assembly method for tunnel segments based on multi-bolt hole spatial constraints as described in claim 1, characterized in that, The nonlinear optimization method is either the Levenberg-Marquardt algorithm or the Gauss-Newton method.

Images