Offshore pelagic pile foundation group pile measurement and superstructure installation method

Abstract

The invention provides an offshore open sea pile foundation group pile measurement and superstructure installation method, which comprises the following steps of: acquiring group pile point cloud data through a laser radar carried by an unmanned aerial vehicle, and acquiring a pile foundation vertex and a pile foundation top surface normal vector through a point cloud data processing algorithm; and designing a connecting device between the pile foundation and the upper structure according to the pile foundation vertex and the normal vector, and calculating absolute coordinates among the feature points of the connecting device according to the pile foundation vertex and the normal vector. The three-dimensional space position of the camera is calibrated and calculated in real time based on the coordinates of the feature points, the hoisting path of the upper structure is automatically decided according to the relative posture of the camera and the connecting device, the problems of pile group vertex positioning and top surface normal vector calculation are solved, and the design problem of the upper structure connecting device is solved. And the problem of relative attitude measurement between the pile foundation superstructure and the connecting device is also solved.

Description

Technical Field

[0001] The present invention relates to the field of offshore pile foundation construction, and in particular to a method for measuring an offshore deep-sea pile foundation group and installing a superstructure. Background Art

[0002] With the continuous development of marine engineering, the demand for offshore pile foundation projects is increasing. This is especially true for large-scale, high-density array pile group construction. The accuracy of pile foundation measurement directly impacts the quality and efficiency of superstructure installation. Measurement methods such as theodolites or total stations are only suitable for onshore or nearshore waters. For offshore pile foundation construction, theodolite or total station measurement platforms are difficult to set up and are susceptible to shaking due to factors such as waves and wind, resulting in reduced measurement accuracy. Therefore, it is necessary to conduct research on rapid, high-precision offshore pile position measurement technology to provide a new solution for the efficient installation of offshore pile foundation superstructures. Summary of the Invention

[0003] The purpose of the present invention is to solve the technical problems in the above background and propose an offshore deep-sea pile foundation group measurement and superstructure installation method, which includes the following steps:

[0004] S1. Construct pile group measurement system;

[0005] S2. Perform pile group measurement using a pile group measurement system;

[0006] S3. Install a connection device on the top of the pile foundation;

[0007] S4. Obtain the coordinates of the marker points of the connection device;

[0008] S5, real-time positioning of the camera space position;

[0009] S6. Locate the upper structure welding points;

[0010] S7. Perform posture measurement and deviation tracking on the pile foundation superstructure.

[0011] In a preferred solution, step S1 further includes the following steps:

[0012] S11. Use a drone-mounted laser radar to obtain pile group data and obtain the vertex of each pile foundation;

[0013] S12. Design a connecting device based on the vertex of each pile foundation, deploy the connecting device on the top of each pile foundation, and keep the top surface of the connecting device horizontal (the drone carries a laser radar to obtain point cloud data with its own coordinate system, and the plane in the coordinate system refers to the plane relative to the XY plane).

[0014] S13. Use a laser radar carried by a drone to obtain point cloud data after installation, identify feature points of the connection device based on a deep learning algorithm, and extract the three-dimensional coordinates of the feature points;

[0015] S14. Install cameras at several welding points of the upper structural components to be installed, with the cameras facing vertically downward to observe the connection devices. The camera positions are calculated in real time through the feature points of the connection devices to complete the posture measurement and motion trajectory tracking of the upper structure.

[0016] In a preferred embodiment, step S2 further comprises the following steps:

[0017] S21. Processing the pile group data based on a clustering algorithm to obtain individual pile foundation data;

[0018] S22, fitting the single pile foundation data based on the RANSAC algorithm to obtain the axis of the steel bar;

[0019] S23, projecting the pile foundation data onto the steel bar axis, extracting the highest point of the projected data and all points within the radius of the highest point, constructing a new point cloud set, extracting the point cloud data of all points in the cloud set before projection, performing circle fitting on the data in the point cloud data using the least squares method, and calculating the center point of the pile foundation;

[0020] S24. For each steel bar, obtain the normal vectors and vertex coordinates of all pile foundations according to steps S22 and S23.

[0021] In a preferred embodiment, step S3 includes the following steps:

[0022] S31. Design the height of the top of the pile foundation. The top structure height is determined according to the design elevation and the top of the pile foundation. Assume that the design height is (x 设 ,y 设 , z 设 ), the coordinates of the pile vertex (x, y, z), the thickness of the support plate is h 支 , then the production height of the top structure is:

[0023] h=z 设 -zh 支 ;

[0024] S32, the top structure of the connecting device is provided with an opening to facilitate the installation of the upper structure and the lower structure of the connecting device;

[0025] S33. Marking point design: Design five circle center marking points, one on the top surface and the other four in the middle of the four sides of the connecting device. Any two of these points should be 2 cm below the plane, and any two should be 4 cm below the plane. Number the five points, assuming the center point is 1#. In counterclockwise order, the remaining four points are 2#, 3#, 4#, and 5#.

[0026] S34, support plate, the support plate is located at the bottom of the connecting device, and a certain number of screw holes are set on the support plate. The connecting device is installed to all pile heads through the screw holes of the support plate.

[0027] In a preferred embodiment, step S4 includes the following steps:

[0028] Step S41: using a drone carrying a laser radar to collect point cloud data of the connection device;

[0029] Step S42: using the Pointnet++ deep learning algorithm to segment, identify and extract feature point data of the acquisition device;

[0030] Step S43: extracting the boundary of the point cloud data.

[0031] In the preferred solution, in step S43, the boundary extraction direction is as follows: traverse each point in the feature point cloud, search for several points of each point's nearest neighbors, and calculate the difference between the x, y, z coordinates of each point and the x, y, z values ​​of several nearest neighboring points;

[0032] When all x differences are positive or negative, the point is determined to be a boundary point; when all y differences are positive or negative, the point is determined to be a boundary point; when all z differences are positive or negative, the point is determined to be a boundary point; otherwise, it is determined to be a non-boundary point.

[0033] In a preferred embodiment, step S5 includes the following steps:

[0034] S51, deploying cameras, respectively deploying two monocular cameras at specific positions near the welding points of the superstructure, with the optical centers of the cameras being λ centimeters away from the welding points;

[0035] S52, positioning the pixel coordinates of the marker point, using a camera to capture an image of the marker point coordinates, and using a Hough transform algorithm to identify the pixel coordinates of the center of the marker point on the image;

[0036] S53, based on the known pixel coordinates of the marker point and the three-dimensional coordinates of the marker point space, use the PNP algorithm to solve the spatial positions of the two cameras in real time, which are recorded as C1 (x1, y1, z1), C2 (x2, y2, z2) and the camera normal vector n 相 .

[0037] In a preferred embodiment, step S6 includes the following steps:

[0038] S61, construct vector C1C2 with the coordinates of the two cameras, and then calculate the vector perpendicular to C1C2 and the normal vector n 相 Vector n 焊 ;

[0039] S62, calculate the welding point position, and take the welding point W1 (x1, y1, z1) where the camera is located w 、y w 、z w ) as an example, the solution formula is:

[0040] W1=C1-λ·n 焊 ;

[0041] Where λ is the distance between the optical center of the camera and the welding point.

[0042] In a preferred solution, step S7 includes the following steps:

[0043] S71. Calculate overall position deviation: Calculate the deviations in the X, Y, and Z directions between the center points of the two welding points and the center points of the two connecting devices;

[0044] S72. The calculation formula for the overall posture deviation is:

[0045]

[0046] Where θ is the camera axis vector n 焊 Angle with the Z axis, n z is the vector along the Z axis;

[0047] The above steps complete the posture measurement and deviation tracking.

[0048] Compared with the prior art, the present invention has the following beneficial effects:

[0049] (1) A pile foundation vertex measurement algorithm based on point cloud data processing algorithm was proposed, which realized the one-time, large-scale and rapid measurement of offshore pile groups.

[0050] (2) The connection structure designed by the present invention has feature points that do not belong to the same three-dimensional space. The three-dimensional space points can calibrate the position of the camera in real time. Since the camera is installed in the upper structure, the device can be used in conjunction with the visual measurement algorithm to realize the posture measurement of the upper structure relative to the connection device, which has the function of assisting visual measurement.

[0051] (3) The connection device proposed in the present invention has the function of assisting in the measurement of the posture of the upper structure. The premise for the connection structure to function is that the connection device is converted to the coordinate system where the pile foundation is located. For this reason, the present invention proposes a method for converting the coordinates of the connection device marker points, and on this basis, according to the principle of photogrammetry, the connection device marker points are used to calculate the lifting posture of the upper structure in real time, thereby realizing real-time tracking of the lifting path. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 This is a schematic diagram of camera deployment.

[0053] Figure 2 It is a schematic diagram of the pile foundation vertex and top surface normal vector.

[0054] Figure 3 It is a side view of the top structure of the connecting device.

[0055] Figure 4 It is a top view of the connecting device.

[0056] Figure 5 This is the relationship diagram between welding points and camera positions. DETAILED DESCRIPTION

[0057] like Figures 1 to 5 As shown, a method for measuring an offshore pile group and installing a superstructure mainly includes the following steps:

[0058] Step S1: Measurement system construction.

[0059] Step S11: Obtaining pile foundation point cloud data: Use a drone-mounted laser radar to obtain pile group data and the vertex of each pile foundation.

[0060] Step S12: Connecting device deployment: Design the connecting device based on the vertex and deploy it on top of each pile foundation, with the top surface of the connecting device kept horizontal.

[0061] Step S13: Acquisition of point cloud data of the connection device. The laser radar carried by the drone is used to acquire the point cloud data after the installation is completed. The feature points of the connection device are identified based on the deep learning algorithm and the three-dimensional coordinates of the feature points are extracted.

[0062] Step S14: Camera deployment. Cameras are installed at the two welding points of the upper component to be installed, with the cameras facing vertically downward to observe the connection device. The camera position is calculated in real time based on the feature points of the connection device to complete the upper structure posture measurement and motion trajectory tracking.

[0063] Step S2: pile group measurement.

[0064] Step S21: Processing the pile group data based on a clustering algorithm to obtain individual pile foundation data;

[0065] Step S22: fitting the single pile foundation data based on the RANSAC algorithm to obtain the axis of the steel bar;

[0066] Step S23: Project the pile foundation data onto the steel bar axis, extract the highest point of the projected data, and all points within a 5 mm radius of the highest point, construct a new point cloud set R, extract the point cloud data of all points in R before projection, record it as Q, use the least squares method to perform circle fitting on the data in Q, and calculate the center point (x, y, z) of the pile foundation;

[0067] Step S24: For each steel bar, obtain the vertices of all pile foundations according to S22 and S23.

[0068] Step S3: Design of the connection device, which includes a top structure, a support plate, and a marking point;

[0069] Step S31: Design the top structure height. The top structure height is determined based on the design elevation and the top of the pile foundation. Assume that the design height is (x 设 ,y 设 , z 设 ), the coordinates of the pile vertex (x, y, z), the thickness of the support plate is h 支 , then the production height of the top structure is:

[0070] h=z 设 -zh 支

[0071] Step S32: The top structure of the connecting device is designed with an opening to facilitate installation of the upper structure and the lower structure of the connecting device;

[0072] Step S33: Marking point design: design 5 circle center marking points, one point is located on the top surface, and the other four points are located in the middle of the four sides of the connecting device. Any two points are 2 cm below the plane, and any two points are 4 cm below the plane. The 5 points are numbered, assuming that the center point is 1#, and in counterclockwise order, the other four points are 2#, 3#, 4#, and 5#;

[0073] Step S34: a support plate, the support plate is located at the bottom of the connecting device, a certain number of screw holes are set on the support plate, and the connecting device is installed to all pile heads through the screw holes of the support plate.

[0074] Step S4: Acquisition of coordinates of connection device marker points

[0075] Step S41: using a drone carrying a laser radar to collect point cloud data of the connection device;

[0076] Step S42: using the Pointnet++ deep learning algorithm to segment, identify and extract feature point data of the acquisition device;

[0077] Step S43: Extract the boundary of the point cloud data. The boundary extraction direction is designed as follows:

[0078] Traverse each point in the feature point cloud and search for the 5 nearest neighboring points of each point. The difference between the x, y, and z coordinates of each point and the x, y, and z values ​​of the 5 nearest neighboring points is calculated. In the following three cases, the point is determined to be a boundary point:

[0079] If all x differences are positive or negative, the point is considered a boundary point;

[0080] If all y differences are positive or negative, the point is determined to be a boundary point;

[0081] If all z differences are positive or negative, the point is determined to be a boundary point;

[0082] The rest of the cases are considered as non-boundary points.

[0083] Step S44: Fit the boundary point cloud data of each marker point based on the least squares method to extract the three-dimensional coordinates of the center of the marker point.

[0084] Step S5: Real-time positioning of the camera space position

[0085] Step S51: Deploy cameras. Two monocular cameras are respectively deployed near the welding points of the upper structure, with the optical center of the camera λ cm away from the welding points.

[0086] Step S52: Locate the pixel coordinates of the marker point, use a camera to capture the marker point coordinate image, and use the Hough transform algorithm to recognize the pixel coordinates of the center of the marker point on the image;

[0087] Step S53: Based on the known pixel coordinates of the marker points and the three-dimensional coordinates of the marker points, the PNP algorithm is used to solve the spatial positions of the two cameras in real time, which are recorded as C1 (x1, y1, z1), C2 (x2, y2, z2) and the camera's normal vector n phase.

[0088] Step S6: Positioning the upper structure welding points.

[0089] Step S61: construct vector C1C2 using the coordinates of the two cameras, and then calculate vector nweld perpendicular to C1C2 and n-phase;

[0090] Step S62: Calculate the welding point position. Taking the welding point W1 (xw, yw, zw) where the C1 (x1, y1, z1) camera is located as an example, the solution formula is as follows:

[0091] W1=C1-λ·n weld;

[0092] In one embodiment, λ is 5.

[0093] Step S6: Superstructure posture measurement and deviation tracking.

[0094] Step S61: Calculate the overall position deviation, calculating the deviations in the X, Y, and Z directions between the center points of the two welding points and the center points of the two connecting devices;

[0095] Step S62: The overall posture deviation is calculated as the angle between the camera axis vector and the Z axis. The Z axis vector is nz. The calculation formula is:

[0096]

[0097] In summary, all position deviations are tracked.

[0098] Preferably, the calculation of the pile foundation vertex and normal vector in step S2 is performed by fitting the pile foundation axis using an improved RANSAC algorithm, specifically including:

[0099] PCA initialization: Calculate the point cloud covariance matrix C:

[0100]

[0101] where p i is the point cloud coordinate, is the mean point, and N is the number of point clouds involved in the calculation.

[0102] RANSAC objective function:

[0103] d(p i ,n,c)=||(p i -c)×n||;

[0104] Among them, d is the vertical distance from the point cloud to the axis, p i is the coordinate of the point cloud to be measured, n is the unit vector (normal vector) in the axis direction, and c is any reference point on the axis

[0105] By least squares circle fitting (pile foundation vertex), specifically:

[0106] Optimization goal:

[0107]

[0108] Among them, (x c ,y c , z c ) represents the three-dimensional coordinates of the center of the fitting circle, r is the radius of the fitting circle, (x i ,y i , z i ) represents the coordinates of the point cloud involved in fitting, inliers: the valid point set filtered by residuals.

[0109] Residual threshold: remove|d i -r|>2σ (σ is the fitting standard deviation)

[0110] Preferably, step S53 further includes:

[0111] PNP pose solution, the formula is:

[0112] u j =K[R|t]P j ;

[0113] Among them, u j is the pixel coordinate of the marker point in the image, K is the camera internal parameter matrix, R is the rotation matrix, t is the translation vector, P j The world coordinate system 3D coordinate of the marker point;

[0114] Fusion with IMU:

[0115]

[0116] Preferably, step S62 further includes:

[0117] Multi-camera cross-validation, the formula is:

[0118]

[0119] Where W1 is the world coordinate of the superstructure welding point, C1 is the optical center coordinate of camera 1, C2 is the optical center coordinate of camera 2, λ is the design distance from the camera optical center to the welding point, and n 焊 Represents the weld plane normal vector.

[0120] Preferably, during the superstructure installation phase, the influence of wave motion on the docking accuracy of the pile foundation and superstructure is offset by wave coupling correction. The wave dynamic position correction formula is specifically:

[0121]

[0122] Wave force coupling formula:

[0123] F wave =ρgAcos(ωt+φ k )d wave ;

[0124] Wave phase calculation formula:

[0125]

[0126] in, is the corrected pile position, P k is the original design position, A is the wave amplitude, ω is the wave angular frequency ω=2π / T, φ k is the phase shift, d wave is the unit vector of the wave propagation direction, ρ is the density of seawater, g is the acceleration of gravity, and λ is the wave wavelength.

[0127] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A method for measuring offshore pile groups and installing superstructures, characterized by: The following steps are involved: S1. Construct pile group measurement system; S2. Perform pile group measurement using a pile group measurement system; S3. Install a connection device on the top of the pile foundation; S4. Obtain the coordinates of the marker points of the connection device; S5, real-time positioning of the camera space position; S6. Locate the upper structure welding points; S7. Perform posture measurement and deviation tracking on the pile foundation superstructure.

2. The offshore deep-sea pile foundation group measurement and superstructure installation method according to claim 1 is characterized by: Step S1 further includes the following steps: S11. Use a drone-mounted laser radar to obtain pile group data and obtain the vertex of each pile foundation; S12. Design a connecting device based on the top of each pile foundation, and deploy the connecting device on the top of each pile foundation, with the top surface of the connecting device kept horizontal; S13. Use a laser radar carried by a drone to obtain point cloud data after installation, identify feature points of the connection device based on a deep learning algorithm, and extract the three-dimensional coordinates of the feature points; S14. Install cameras at several welding points of the upper structural components to be installed, with the cameras facing vertically downward to observe the connection devices. The camera positions are calculated in real time through the feature points of the connection devices to complete the posture measurement and motion trajectory tracking of the upper structure.

3. The offshore deep-sea pile foundation group measurement and superstructure installation method according to claim 1 is characterized by: Step S2 further includes the following steps: S21. Processing the pile group data based on a clustering algorithm to obtain individual pile foundation data; S22, fitting the single pile foundation data based on the RANSAC algorithm to obtain the axis of the steel bar; S23, projecting the pile foundation data onto the steel bar axis, extracting the highest point of the projected data and all points within the radius of the highest point, constructing a new point cloud set, extracting the point cloud data of all points in the cloud set before projection, performing circle fitting on the data in the point cloud data using the least squares method, and calculating the center point of the pile foundation; S24. For each steel bar, obtain the normal vectors and vertex coordinates of all pile foundations according to steps S22 and S23.

4. The offshore deep-sea pile foundation pile group measurement and superstructure installation method according to claim 1 is characterized by: Step S3 includes the following steps: S31. Design the height of the top of the pile foundation. The top structure height is determined according to the design elevation and the top of the pile foundation. Assume that the design height is (x 设 ,y 设 , z 设 ), the coordinates of the pile vertex (x, y, z), the thickness of the support plate is h 支 , then the production height of the top structure is: h=z 设 -z-h 支 ; S32, the top structure of the connecting device is provided with an opening to facilitate the installation of the upper structure and the lower structure of the connecting device; S33. Marking point design: Design five circle center marking points, one on the top surface and the other four in the middle of the four sides of the connecting device. Any two of these points should be 2 cm below the plane, and any two should be 4 cm below the plane. Number the five points, assuming the center point is 1#. In counterclockwise order, the remaining four points are 2#, 3#, 4#, and 5#. S34, support plate, the support plate is located at the bottom of the connecting device, and a certain number of screw holes are set on the support plate. The connecting device is installed to all pile heads through the screw holes of the support plate.

5. The offshore deep-sea pile foundation pile group measurement and superstructure installation method according to claim 1 is characterized by: Step S4 includes the following steps: Step S41: using a drone carrying a laser radar to collect point cloud data of the connection device; Step S42: using the Pointnet++ deep learning algorithm to segment, identify and extract feature point data of the acquisition device; Step S43: extracting the boundary of the point cloud data.

6. The offshore deep-sea pile foundation pile group measurement and superstructure installation method according to claim 5 is characterized by: In step S43, the boundary extraction method is as follows: traverse each point in the feature point cloud, search for several points of each point's nearest neighbors, and calculate the difference between the x, y, and z coordinates of each point and the x, y, and z values ​​of several neighboring points; When all x differences are positive or negative, the point is determined to be a boundary point; when all y differences are positive or negative, the point is determined to be a boundary point; when all z differences are positive or negative, the point is determined to be a boundary point; The rest of the cases are considered as non-boundary points.

7. The offshore deep-sea pile foundation pile group measurement and superstructure installation method according to claim 1 is characterized by: Step S5 includes the following steps: S51, deploying cameras, respectively deploying two monocular cameras at specific positions near the welding points of the superstructure, with the optical centers of the cameras being λ centimeters away from the welding points; S52, positioning the pixel coordinates of the marker point, using a camera to capture an image of the marker point coordinates, and using a Hough transform algorithm to identify the pixel coordinates of the center of the marker point on the image; S53, according to the known pixel coordinates of the marker point and the three-dimensional coordinates of the marker point space, use the PNP algorithm to solve the spatial positions of the two cameras in real time, which are recorded as C1 (x1, y1, z1), C2 (x2, y2, z2) and the camera normal vector n 相 .

8. The offshore deep-sea pile foundation group measurement and superstructure installation method according to claim 1 is characterized by: Step S6 includes the following steps: S61, construct vector C1C2 with the coordinates of the two cameras, and then calculate the vector perpendicular to C1C2 and the normal vector n 相 Vector n 焊 ; S62, calculate the welding point position, and take the welding point W1 (x1, y1, z1) where the camera is located w 、y w 、z w ) as an example, the solution formula is: W1=C1-λ·n 焊 ; Where λ is the distance between the optical center of the camera and the welding point.

9. The offshore deep-sea pile foundation pile group measurement and superstructure installation method according to claim 1 is characterized by: Step S7 includes the following steps: S71. Calculate overall position deviation: Calculate the deviations in the X, Y, and Z directions between the center points of the two welding points and the center points of the two connecting devices; S72. The calculation formula for the overall posture deviation is: Where θ is the camera axis vector n 焊 Angle with the Z axis, n z is the vector along the Z axis; The above steps complete the posture measurement and deviation tracking.

Images