Indirect positioning method for shield tunneling machine cutter fastening bolt based on point cloud data analysis

Abstract

The invention relates to the technical field of metal part disassembly and assembly, in particular to a point cloud data analysis-based indirect positioning method for a shield tunneling machine cutter fastening bolt, which comprises the following steps of: acquiring point cloud data of a preset feature plane in a shield tunneling machine cutter box; fitting the obtained point cloud data to obtain a fitting plane of the feature plane, and projecting the point cloud data to a fitting plane group to obtain a quadrilateral boundary of the point cloud projected to the fitting plane; calculating the geometric center of a fitting plane based on the obtained quadrangle boundary; and acquiring a fixed position relationship between the bolt and the calculation geometric center and a coordinate system conversion relationship, and calculating a bolt coordinate. Therefore, the defects of traditional manual operation are overcome, and the tool changing efficiency and safety are greatly improved.

Description

Technical Field

[0001] The invention relates to the technical field of metal parts disassembly and assembly, and in particular to an indirect positioning method for a shield machine cutter fastening bolt based on point cloud data analysis. Background Art

[0002] As an important equipment for tunnel excavation, shield machines are mainly used in the construction of subways, highways, railway tunnels and water diversion tunnels of hydropower stations. Their working environment is complex and the conditions are harsh. Wear and damage of the rollers on the cutterhead of the shield machine are inevitable, so they need to be replaced regularly. The traditional tool changing process usually relies on manual operation. Workers need to operate inside the narrow and dark cutterhead, which is labor-intensive, inefficient and has safety hazards. In order to improve the efficiency of tool changing and ensure the safety of operations, automated tool changing robots are gradually being used in the tool changing process of shield machines. The tool changing robots use mechanical arms and visual systems to realize the automatic disassembly and installation of the rollers, which not only improves the efficiency of tool changing, but also significantly reduces the labor intensity and operation risks of workers.

[0003] During the automated tool change process, accurate positioning of the fastening bolts that lock the hob inside the tool box is a prerequisite for the robot to automatically change the hob. As an advanced visual sensor, the binocular 3D camera can simultaneously capture the image of the target object through two cameras, and use the parallax principle to calculate the three-dimensional depth information of the object, thereby generating high-precision three-dimensional point cloud data. However, due to the small space on the back of the shield machine cutterhead, the small end area of ​​the bolts, and the limited field of view and working distance of the 3D camera, direct positioning of the bolts will result in incomplete point cloud data and inaccurate positioning. Summary of the invention

[0004] The purpose of the present invention is to provide an indirect positioning method for the tool fastening bolts of a shield machine based on point cloud data analysis, which aims to overcome the shortcomings of traditional manual operation and greatly improve the efficiency and safety of tool change. High-precision point cloud data is obtained by using a binocular 3D camera, and precise bolt positioning is achieved in a complex working environment through plane fitting and coordinate transformation.

[0005] To achieve the above object, the present invention provides an indirect positioning method for a shield machine cutter fastening bolt based on point cloud data analysis, comprising obtaining point cloud data of a preset characteristic plane inside a shield machine cutter box;

[0006] Fitting the acquired point cloud data to obtain a fitting plane of the feature plane, projecting the point cloud data onto the fitting plane group to obtain a quadrilateral boundary of the point cloud projected onto the fitting plane;

[0007] Calculate the geometric center of the fitting plane based on the acquired quadrilateral boundary;

[0008] Obtain the fixed position relationship between the bolt and the calculated geometric center and the coordinate system transformation relationship to calculate the bolt coordinates;

[0009] The tool changing robot removes the bolts and changes the tool based on the bolt coordinates, and after the tool change is completed, the tool is bolted again based on the bolt coordinates.

[0010] The specific steps of obtaining the point cloud data of the characteristic plane preset inside the shield machine cutter box include:

[0011] Obtain the total point cloud data obtained by shooting the inside of the shield machine cutter box after the robot reaches the specified position;

[0012] Extracting feature point cloud data corresponding to the feature surface from the total point cloud data based on preset geometric shape rules;

[0013] The feature point cloud data is filtered to remove noise.

[0014] Among them, when the robot reaches the designated position and shoots the inside of the shield machine cutter box, a binocular 3D camera is used to obtain point cloud data.

[0015] The specific steps of filtering the feature point cloud data to remove noise include:

[0016] Get the set fixed search radius;

[0017] Check the number of neighboring points within the search radius of each point in the feature point cloud data. If the number of neighboring points is less than the set threshold, the point is considered to be a noise point and is removed.

[0018] The specific steps of fitting the acquired point cloud data to obtain a fitting plane of the feature plane, projecting the point cloud data onto the fitting plane group, and obtaining a quadrilateral boundary projected onto the fitting plane by the point cloud include:

[0019] Obtain feature point cloud data after noise removal;

[0020] The least square method is used to perform plane fitting on the feature point cloud data after noise removal to obtain the normal vector and plane parameters of the fitting plane;

[0021] Project all feature point cloud data onto the above fitting plane;

[0022] The projected point cloud data is subjected to geometric analysis to obtain a rectangular boundary.

[0023] The specific steps of using the least square method to perform plane fitting on the feature point cloud data after noise removal to obtain the normal vector and plane parameters of the fitting plane include:

[0024] Get the set fitting plane equation;

[0025] The error function is defined as the sum of the squares of the distances from all points to the fitting plane;

[0026] The fitting plane equation is differentiated and its derivative is set to 0 to obtain a system of linear equations. The least squares solution can be solved through singular value decomposition to obtain the normal vector and plane parameters of the fitting plane.

[0027] The specific steps of geometrically analyzing the projected point cloud data to obtain a rectangular boundary include:

[0028] Arrange the projected point cloud data in ascending order according to coordinates;

[0029] Use the convex hull algorithm to divide the projected point cloud data into upper and lower parts, solve the upper and lower convex hulls respectively, and finally merge them to get the minimum convex hull;

[0030] The rotating calcaneal method is used to find the minimum area rectangle containing the minimum convex hull and obtain the rectangular boundary.

[0031] The specific step of calculating the geometric center of the fitting plane based on the acquired quadrilateral boundary includes:

[0032] Determine the four vertices in the minimum quadrilateral based on the results of the rotating calcaneal method;

[0033] Calculate the average of the coordinates of the four vertices to get the geometric center.

[0034] The specific steps of obtaining the fixed position relationship between the bolt and the calculated geometric center and the coordinate system conversion relationship and calculating the bolt coordinates include:

[0035] Obtain the positional relationship between the hob fastening bolt and the geometric center in the cutter box model;

[0036] Convert the positional relationship to the coordinate system of the fitting plane to obtain the bolt center;

[0037] The coordinates of the bolt in the robot coordinate system can be calculated by using the conversion relationship between the robot coordinate system of the captured point cloud data and the coordinate system of the fitting plane.

[0038] The indirect positioning method of the shield machine tool fastening bolt based on point cloud data analysis of the present invention comprises: obtaining point cloud data of a preset characteristic plane inside the shield machine tool box; fitting the obtained point cloud data to obtain a fitting plane of the characteristic plane, projecting the point cloud data onto the fitting plane group, and obtaining a quadrilateral boundary of the point cloud projected onto the fitting plane; calculating the geometric center of the fitting plane based on the obtained quadrilateral boundary; obtaining the fixed position relationship and coordinate system conversion relationship between the bolt and the calculated geometric center and calculating the bolt coordinates. Thus, the point cloud data of the preset characteristic plane inside the tool box can be obtained by a 3D camera, and the three-dimensional coordinates of the fastening bolt can be indirectly obtained through a series of point cloud processing processes, thereby providing bolt coordinate information for the tool changing robot to perform automatic tool changing. The shortcomings of traditional manual operation are overcome, and the efficiency and safety of tool changing are greatly improved. High-precision point cloud data is obtained by using a binocular 3D camera, and precise positioning of bolts in complex working environments is achieved through plane fitting and coordinate conversion. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0040] Figure 1 It is a schematic diagram of the preset characteristic surface and calculated geometric center position inside the hob cutter box.

[0041] Figure 2 It is a schematic diagram of the depth position relationship between the tool box fastening bolts and the preset feature surface.

[0042] Figure 3 It is a flow chart of the indirect positioning method of the shield machine cutter fastening bolt based on point cloud data analysis according to the first embodiment of the present invention.

[0043] Figure 4 It is a flow chart of the first embodiment of the present invention for obtaining point cloud data of a preset characteristic plane inside a shield machine tool box.

[0044] Figure 5 It is a flowchart of filtering the feature point cloud data to remove noise according to the first embodiment of the present invention.

[0045] Figure 6 This is a flowchart of the first embodiment of the present invention, which is to fit the point cloud data to obtain a fitting plane of the feature plane, project the point cloud data onto the fitting plane group, and obtain the quadrilateral boundary projected onto the point cloud on the fitting plane.

[0046] Figure 7 It is a flowchart of setting the fitting plane equation according to the first embodiment of the present invention.

[0047] Figure 8 This is a flow chart of performing geometric analysis on projected point cloud data to obtain a rectangular boundary according to the first embodiment of the present invention.

[0048] Fig. 9 It is a flow chart of calculating the geometric center of the fitting plane based on the quadrilateral boundary according to the first embodiment of the present invention.

[0049] Fig.10 It is a flow chart of calculating the bolt coordinates according to the fixed position relationship between the bolt and the calculated geometric center and the coordinate system conversion relationship of the first embodiment of the present invention.

[0050] Fig.11 This is a flow chart of the tool changing robot of the first embodiment of the present invention removing bolts and changing tools based on the bolt coordinates, and re-bolting the tool based on the bolt coordinates after the tool change is completed.

[0051] Fig.12 It is an overall flow chart of the indirect positioning method of the hob box fastening bolts based on a 3D camera of the present invention. DETAILED DESCRIPTION

[0052] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention, and should not be construed as limiting the present invention.

[0053] First embodiment

[0054] See also Figures 1 to 12 The present invention provides an indirect positioning method for a shield machine cutter fastening bolt based on point cloud data analysis, comprising:

[0055] S101 obtains point cloud data of a preset characteristic plane inside a shield machine cutter box;

[0056] The specific steps include:

[0057] S201 obtains total point cloud data obtained by photographing the inside of the shield machine cutter box after the robot reaches the designated position;

[0058] The binocular 3D camera is installed on the end effector of the tool changing robot in the form of an eye on the hand and moves with the robot. When the tool changing robot needs to be positioned, the camera is triggered to collect data and obtain the point cloud data in the tool box. Since the point cloud data corresponding to the small area of ​​the bolt end face is incomplete, the following points are extracted from the collected point cloud: Figure 1 The preset feature plane point cloud data corresponding to the oblique line area in the tool box. At this time, the point cloud coordinates are established based on the left camera coordinate system {A} of the binocular 3D camera.

[0059] When the robot reaches the designated position and photographs the interior of the cutter box of the shield machine, a binocular 3D camera is used to obtain point cloud data.

[0060] S202 extracts feature point cloud data corresponding to the feature surface from the total point cloud data based on a preset geometric shape rule;

[0061] S203: filtering the feature point cloud data to remove noise.

[0062] Use the radius filtering method to process the point cloud data. There are usually outliers in the extracted point cloud, which will affect the subsequent calculations, so the data must be preprocessed first. Use the radius filtering method to filter out outliers. The specific steps include:

[0063] S301 obtains a set fixed search radius;

[0064] S302 checks the number of neighboring points within the search radius of each point in the feature point cloud data. If the number of neighboring points is less than a set threshold, the point is considered to be a noise point and is removed.

[0065] Assume that the point cloud data is P = {p1, p2, ..., p n}, where each point P i The coordinates of (x i ,y i ,z i ), for each point P i , calculate all points P in its neighborhood j satisfy:

[0066] d ij =√(x i -x j ) 2 +(y i -y j ) 2 +(z i -z j ) 2 ≤r

[0067] Among them, d ij Point P i and P j The Euclidean distance between them, r is the search radius.

[0068] S102: fitting the acquired point cloud data to obtain a fitting plane of the feature plane, projecting the point cloud data onto the fitting plane group to obtain a quadrilateral boundary projected onto the fitting plane;

[0069] The specific steps include:

[0070] S401 obtains feature point cloud data after noise removal;

[0071] S402 uses the least square method to perform plane fitting on the feature point cloud data after noise removal to obtain the normal vector and plane parameters of the fitting plane;

[0072] The plane is fitted to the above point cloud using the least squares method, and the parameters of the plane equation are solved by minimizing the sum of the squares of the vertical distances from the points to the fitting plane. The specific steps include:

[0073] S501 obtains a set fitting plane equation;

[0074] First, set the plane equation to ax+by+cz+d=0, where (a, b, c) is the plane normal vector and d is the distance from the plane to the origin.

[0075] S502 defines the error function as the sum of squares of distances from all points to the fitting plane;

[0076] Define the error function E as the sum of the squares of the distances from all points to the plane:

[0077]

[0078] S503 derives the fitting plane equation and sets its derivative to 0 to obtain a linear equation system, which can be solved by singular value decomposition to obtain a least square solution to obtain the normal vector and plane parameters of the fitting plane.

[0079] In order to minimize the error, by taking the derivative of E and setting its derivative to 0, we get the linear equation system Ap=0. The least squares solution can be obtained through singular value decomposition to obtain the fitting plane parameters.

[0080] in, p=(a,b,c,d).

[0081] S403 projects all feature point cloud data onto the fitting plane;

[0082] After the best fitting plane is determined through the above process, the orthogonal projection method is used to project each point onto the plane in a direction perpendicular to the fitting plane. While projecting, a new coordinate system {B} is established with the first point projected onto the fitting plane as the origin, the fitting plane as the xoy plane, and the normal vector of the fitting plane as the z-axis direction. The projected point cloud data will all be on the xoy plane, and their z coordinates are all 0, which is convenient for the subsequent extraction of two-dimensional geometric features. The specific process includes: for each point cloud data point (x i ,y i ,z i ), calculate its vertical projection point to the fitting plane, and calculate its coordinates in the new coordinate system as (x′ i ,y′ i ,0).

[0083] The normal vector n = (a, b, c) of the fitted plane is used as the z′ axis of the new coordinate system. A vector on the plane is selected and normalized to obtain u. After the cross product operation v = w × u, v is obtained. Thus, the transformation matrix from the original coordinates to the new coordinates is established:

[0084]

[0085] Assume that the coordinates of point p in the original coordinate system {A} are (x, y, z), and the coordinates in the newly established coordinate system {B} are (x′, y′, z′), and the transformation relationship is p′=T(pp′0).

[0086] Wherein, p′0 is the coordinate of the origin of the newly created coordinate system {B} in the original coordinate system {A}.

[0087] S404 performs geometric analysis on the projected point cloud data to obtain a rectangular boundary.

[0088] The specific steps include:

[0089] S601 arranges the projected point cloud data in ascending order of coordinates;

[0090] The projected point cloud data {B} is all on the xoy plane in the coordinate system, and the z coordinates of all points are 0. First, the minimum convex hull containing all projected points is calculated, and all point sets are arranged from small to large according to the x coordinates. If the x coordinates are the same, they are sorted by the y coordinates.

[0091] S602 uses a convex hull algorithm to divide the projected point cloud data into upper and lower parts, respectively solves the upper and lower convex hulls, and finally merges them to obtain the minimum convex hull;

[0092] Using the divide-and-conquer idea of ​​Andrew's convex hull algorithm, the point set is divided into upper and lower parts, the upper and lower convex hulls are solved respectively, and finally merged to obtain the minimum convex hull.

[0093] S603 uses the rotating caliper method to find the minimum area rectangle containing the minimum convex hull, and obtains the rectangular boundary.

[0094] The rotating calcaneal method is used to find the minimum area rectangle containing the convex hull. The basic idea is: starting from one side of the convex hull, rotate the rectangle so that the side of the rectangle is parallel to one side of the convex hull. At each rotation position, calculate the area of ​​the rectangle and record the minimum area and the corresponding rectangle. After comparison, the minimum target rectangle is finally obtained.

[0095] After the minimum rectangular boundary surrounding the projected point cloud is determined, its geometric center is calculated. The geometric center refers to the geometric center point of the fitted rectangle, which is calculated by the average coordinates of the rectangle vertices. The specific process is to determine the four vertices in the minimum quadrilateral according to the results of the rotating caliper method, calculate the average x and y coordinates of the four vertices, and obtain the coordinates of the geometric center P = (x_c, y_c, 0). The calculation of the geometric center provides an accurate reference point for subsequent bolt positioning.

[0096] S103 calculates the geometric center of the fitting plane based on the acquired quadrilateral boundary;

[0097] The specific steps include:

[0098] S701 determines four vertices in the minimum quadrilateral according to the result of the rotating calculus method;

[0099] The coordinates of the four vertices are obtained in the above manner.

[0100] S702 calculates the average values ​​of the coordinates of the four vertices respectively to obtain the geometric center.

[0101] The structural form of the hob cutter box has been determined, so the relative position between the center of the fastening bolt and the fitted geometric center is determined and can be obtained by measuring in the model, such as Figure 1 and 2 The figure shows the relative position relationship between the calculated geometric center and the center of the bolt head end face. The geometric center is calculated based on the coordinate system {B}. On the xoy plane, a distance l1 is extended in the direction passing through the geometric center point P and parallel to the longer side of the rectangle. On this basis, a distance l2 is extended in the direction parallel to the shorter side of the fitted rectangle. Since the distance between the bolt head end face and the internal characteristic surface of the tool box is l3, the coordinates of the bolt end center in the coordinate system {B} can be derived:

[0102] P {B} =(x_target,y_target,z_target)

[0103] x_target=x_c+l1·d long +l2·d short

[0104] y_target=y_c+l1·d long +l2·d short

[0105] z_target=l3

[0106] Among them, d long is the direction vector of the longer side of the fitted rectangle, d short is the direction vector of the shorter side of the fitted rectangle.

[0107] S104 obtains the fixed position relationship and coordinate system conversion relationship between the bolt and the calculated geometric center and calculates the bolt coordinates;

[0108] The specific steps include:

[0109] S801 obtains the position relationship between the hob fastening bolt and the geometric center in the cutter box model;

[0110] S802 converts the positional relationship into the coordinate system of the fitting plane to obtain the bolt center;

[0111] S803 uses the conversion relationship between the robot coordinate system of the captured point cloud data and the coordinate system of the fitting plane to calculate the bolt coordinates in the robot coordinate system.

[0112] As can be seen from the above, the transformation matrix T from the original coordinate system {A} to the established coordinate system {B} has been determined, and the coordinates of the end center of the fastening bolt in the coordinate system {B} have been calculated.

[0113] P {B} =(x_target, y_target, z_target), and thus the coordinate P of the bolt in the original coordinate system {A} can be calculated according to the conversion relationship between the coordinate systems. {A} , the specific conversion relationship is P {A} =P {B} T -1 This achieves the indirect calculation of the bolt coordinates through the larger feature surface inside the tool box corresponding to more point clouds, avoiding the problem of inaccurate point cloud calculation directly extracted from the bolt area.

[0114] S105 The tool changing robot removes the bolts and changes the tool based on the bolt coordinates, and after the tool change is completed, bolts the tool again based on the bolt coordinates.

[0115] The specific steps include:

[0116] S901 According to the detected bolt coordinates, the robot adjusts the position of its end effector to dock with the bolt;

[0117] In this step, the S105 tool changing robot first identifies the specific position coordinates of each bolt on the current tool through its built-in sensor or external positioning system. Then, the robot control system will command the end effector (usually a robotic arm with a wrench or other tool) to move to the first bolt that needs to be operated and ensure that the head of the wrench or other tool is accurately engaged with the bolt.

[0118] S902 rotates the end effector to loosen the bolt;

[0119] When the end effector is properly docked with the bolt, the robot will start the drive device on the end effector to generate appropriate rotational torque to start loosening the bolt. This process needs to ensure that the bolt is completely out of its original fixed state to facilitate the next step of disassembly.

[0120] S903 When all bolts are loosened, the robot takes out the old tool from the tool box;

[0121] After all the bolts have been loosened, the robot will carefully remove the old tool that is no longer in use from the tool box (or tool holder) according to the preset program. This step may involve the coordinated movement of multiple degrees of freedom of the robot arm to ensure that the tool is removed safely and without error.

[0122] The S904 robot places the new tool into the tool box and aligns the bolt holes of the tool and the tool box;

[0123] After removing the old tool, the robot will automatically or under the operator's instruction select the appropriate new tool and install it in the tool box. During this process, the robot must ensure that the bolt holes on the new tool are accurately aligned with the corresponding holes on the tool box to facilitate subsequent fixing operations.

[0124] The S905 robot again uses the end effector to tighten the bolt according to the predetermined torque value and bolt coordinates.

[0125] Finally, the robot will repeat the action similar to step S901, but this time to tighten the bolts. Through the end effector, the robot will apply the correct torque value to tighten each bolt, ensuring that the new tool is firmly installed on the tool box and that all connection points meet the process requirements.

[0126] What is disclosed above is only a preferred embodiment of the present invention, and it certainly cannot be used to limit the scope of rights of the present invention. Ordinary technicians in this field can understand that all or part of the processes of the above embodiment and equivalent changes made according to the claims of the present invention still fall within the scope of the invention.

Claims

1. Indirect positioning method of shield machine tool fastening bolts based on point cloud data analysis, It is characterized in that Including: obtaining point cloud data of a preset characteristic plane inside the shield machine cutter box; Fitting the acquired point cloud data to obtain a fitting plane of the feature plane, projecting the point cloud data onto the fitting plane group to obtain a quadrilateral boundary of the point cloud projected onto the fitting plane; Calculate the geometric center of the fitting plane based on the acquired quadrilateral boundary; Obtain the fixed position relationship and coordinate system transformation relationship between the bolt and the calculated geometric center and calculate the bolt coordinates.

2. The indirect positioning method for the shield machine tool fastening bolt based on point cloud data analysis according to claim 1 is characterized in that: The specific steps of obtaining the point cloud data of the characteristic plane preset inside the shield machine cutter box include: Obtain the total point cloud data obtained by shooting the inside of the shield machine cutter box after the robot reaches the specified position; Extracting feature point cloud data corresponding to the feature surface from the total point cloud data based on preset geometric shape rules; The feature point cloud data is filtered to remove noise.

3. The indirect positioning method for the tool fastening bolts of a shield machine based on point cloud data analysis according to claim 2, characterized in that: When the robot reaches the designated position and photographs the interior of the shield machine cutter box, a binocular 3D camera is used to obtain point cloud data.

4. The indirect positioning method for the tool fastening bolts of a shield machine based on point cloud data analysis according to claim 2, characterized in that: The specific steps of filtering the feature point cloud data to remove noise include: Get the set fixed search radius; Check the number of neighboring points within the search radius of each point in the feature point cloud data. If the number of neighboring points is less than the set threshold, the point is considered to be a noise point and is removed.

5. The indirect positioning method for the shield machine tool fastening bolt based on point cloud data analysis according to claim 1, characterized in that: The specific steps of fitting the acquired point cloud data to obtain a fitting plane of the feature plane, projecting the point cloud data onto the fitting plane group, and obtaining a quadrilateral boundary projected onto the fitting plane by the point cloud include: Obtain feature point cloud data after noise removal; The least square method is used to perform plane fitting on the feature point cloud data after noise removal to obtain the normal vector and plane parameters of the fitting plane; Project all feature point cloud data onto the fitting plane; The projected point cloud data is subjected to geometric analysis to obtain a rectangular boundary.

6. The indirect positioning method for the shield machine tool fastening bolt based on point cloud data analysis according to claim 5, characterized in that: The specific steps of using the least square method to perform plane fitting on the feature point cloud data after noise removal to obtain the normal vector and plane parameters of the fitting plane include: Get the set fitting plane equation; The error function is defined as the sum of the squares of the distances from all points to the fitting plane; The fitting plane equation is differentiated and its derivative is set to 0 to obtain a system of linear equations. The least squares solution can be solved through singular value decomposition to obtain the normal vector and plane parameters of the fitting plane.

7. The indirect positioning method for the tool fastening bolts of a shield machine based on point cloud data analysis according to claim 5, characterized in that: The specific steps of performing geometric analysis on the projected point cloud data to obtain a rectangular boundary include: Arrange the projected point cloud data in ascending order according to coordinates; Use the convex hull algorithm to divide the projected point cloud data into upper and lower parts, solve the upper and lower convex hulls respectively, and finally merge them to get the minimum convex hull; The rotating calcaneal method is used to find the minimum area rectangle containing the minimum convex hull and obtain the rectangular boundary.

8. The indirect positioning method for the tool fastening bolts of a shield machine based on point cloud data analysis according to claim 1, characterized in that: The specific steps of calculating the geometric center of the fitting plane based on the acquired quadrilateral boundary include: Determine the four vertices in the minimum quadrilateral based on the results of the rotating calcaneal method; Calculate the average of the coordinates of the four vertices to get the geometric center.

9. The indirect positioning method for the tool fastening bolts of a shield machine based on point cloud data analysis according to claim 1, characterized in that: The specific steps of obtaining the fixed position relationship between the bolt and the calculated geometric center and the coordinate system conversion relationship and calculating the bolt coordinates include: Obtain the positional relationship between the hob fastening bolt and the geometric center in the cutter box model; Convert the positional relationship to the coordinate system of the fitting plane to obtain the bolt center; The coordinates of the bolt in the robot coordinate system can be calculated by using the conversion relationship between the robot coordinate system of the captured point cloud data and the coordinate system of the fitting plane.

Images