Foundation scouring repair effect evaluation method based on TIN model

Abstract

The invention relates to an offshore pile foundation construction acceptance technology, and provides an acceptance method for solving the problems that the effective amount of riprap, solidified soil or other similar repair construction is low in proportion and unclear. Three-dimensional point cloud data before and after construction is obtained by combining a multi-beam sounding system with LiDAR, after preprocessing such as denoising and downsampling, a triangular network is constructed by adopting a Bowyer-Watson algorithm, the difference value between each vertex and the design elevation is calculated, scour / siltation areas are classified, and the volume is counted by utilizing a triangular network method to obtain the total project amount. According to the method, the positioning precision reaches the centimeter level, the triangulation network method is suitable for complex terrains, the total project amount and the effective throwing and filling amount can be accurately estimated, the construction quality can be supervised, stone waste is avoided, the transportation and construction cost is saved, and a visual report can be generated.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of offshore wind turbine construction, and in particular to a foundation scour repair effect evaluation method based on a TIN model. BACKGROUND

[0002] The offshore wind turbine is affected by the too changeable and complex natural environment conditions near the coast, some offshore wind power sites are located in a sea area with strong tidal power and large flow velocity, and thus the tidal current scouring capacity and sand carrying capacity are relatively large, and the continuous marine load can easily cause local scouring problems of the seabed around the pile foundation,

[0003] For the scouring problem of the wind turbine, the main solution at present is to use precise riprapping or solidified soil filling at the scouring part of the wind turbine pile foundation to make the surrounding mud surface elevation of the wind turbine reach the design elevation or a relatively safe elevation. After the construction is completed, the amount of riprapping is settled, and this method is difficult to control the construction quality, for example, too large or too small riprapping area will affect the riprapping effect, and the actual effective amount in the total riprapping amount may not be high. SUMMARY

[0004] The technical problem to be solved by the application is that the actual effective proportion in the total riprapping amount is not high and the effective proportion is not clear after the construction.

[0005] To solve the above technical problems, the technical scheme adopted by the application is: a foundation scour repair effect evaluation method based on a TIN model, comprising:

[0006] Step one, data acquisition: use a multi-beam sounding system to obtain three-dimensional point cloud data with a grid of 0.25 before and after the construction, and the data density can be improved by means of LiDAR;

[0007] Step two, data preprocessing: remove outliers by statistical outlier removal or radius filtering, reduce data complexity by random sampling or curvature reservation algorithm, manually review points with a delta h exceeding a set threshold, and smooth the grid by means of Laplacian filtering;

[0008] Step three, triangular net construction: project the processed point set to the horizontal plane, generate a triangular grid by iteratively inserting points through the Bowyer-Watson algorithm, ensure that the circumcircle of any triangle does not contain other points, and use constrained Delaunay triangulation for steep change areas such as the edge of the pile foundation, so that each triangle vertex has an elevation attribute;

[0009] Step four, elevation difference calculation: calculate the difference between each triangular grid vertex and the design elevation H to obtain the elevation change value;

[0010] Step five, regional classification: according to the change value of the three vertices of the triangle, if ≥2 points are erosion, the whole is classified as an erosion area; if ≥2 points are deposition, the whole is classified as a deposition area; if the number of erosion / deposition is equal, the area weight interpolation processing is carried out;

[0011] Step six, volume and total amount statistics: first, calculate the horizontal projection area of the triangle Then calculate the average elevation change Δh through triangle interpolation avg = 1 / 3 × (Δh1+Δh2+Δh3), and then obtain a single volume V=S·Δh avg , respectively, the amount of erosion and deposition before and after the work, and calculate the total amount of engineering=(work before erosion amount-work after erosion amount)+(work after deposition amount-work before deposition amount).

[0012] Preferably, when LiDAR is used to improve data density, the scanning resolution is not less than 0.1 meters.

[0013] Preferably, the set threshold is 1.5-2 times the measurement accuracy of the multi-beam sounding system.

[0014] Preferably, the iteration number of the Laplacian filter is 3-5 times.

[0015] Preferably, in the process of inserting points in Bowyer-Watson algorithm iteration, when the distance between adjacent points is less than 0.05 meters, the insertion is stopped.

[0016] Preferably, in the constrained Delaunay triangulation, the error allowed range of the constraint edge is ±0.03 meters.

[0017] Preferably, when the area weight interpolation processing is carried out, the final determination result is calculated according to the area proportion of different regions in the triangle.

[0018] Preferably, when calculating the horizontal projection area of the triangle, the result is kept to three decimal places.

[0019] Preferably, when the error rate of the total amount of engineering calculation result exceeds 5%, all the steps of the method are re-executed.

[0020] Preferably, after the calculation is completed, a visual report is generated, including three-dimensional model display and data chart analysis.

[0021] The present application provides a kind of based on TIN model's basic erosion repair effect evaluation method, with following beneficial effects.

[0022] 1. The multibeam echo sounding system employs real-time dynamic differential positioning (RTK) technology, achieving positioning accuracy at the centimeter or even millimeter level. It can accurately measure elevation changes before and after construction. The earthwork volume calculation using the triangulation method is highly accurate, produces realistic models, and is suitable for various complex terrains, including areas with significant undulations. It can handle earthwork volume calculations of various shapes and sizes, helping to confirm the amount of fill dumped by the construction party. Using Surfer software, not only can elevation changes before and after construction be calculated, but also vivid renderings can be created.

[0023] 2. It can accurately estimate the total amount of construction work, thereby supervising the actual amount of work done by the construction party; it can calculate the effective amount of filling within the filling area, thereby supervising the construction party's control of construction quality, effectively avoiding waste of stone materials, and saving transportation and construction costs. Attached Figure Description

[0024] The present invention will be further described below with reference to the accompanying drawings and embodiments:

[0025] Figure 1 This is a flowchart for calculating point cloud data using the triangular mesh method.

[0026] Figure 2 This is a Surfer diagram of the pile foundation before construction.

[0027] Figure 3 This is a Surfer diagram of the pile foundation after construction. Detailed Implementation

[0028] like Figures 1-3 As shown in the figure. This invention proposes a method for evaluating the effect of foundation scour repair based on the TIN model. A multi-beam system is used to measure the elevation of the wind turbine before and after construction. After data processing, cloud data with a grid size of 0.25m is obtained before and after construction. Real-time dynamic differential positioning (RTK) technology is used, and differential data transmission between the base station and the rover station eliminates or reduces errors in satellite signal propagation, achieving centimeter-level or even millimeter-level positioning accuracy. The obtained cloud data is "cut" to obtain pre- and post-construction data for the filling area. The triangulation method is used to calculate the pre- and post-construction scour and sedimentation volumes relative to the design elevation. By comparing the scour and sedimentation volumes before and after construction, the total filling volume of the filling area is obtained. This verifies the actual volume of construction. The cloud data before and after construction are mapped and calculated using Surfer software to determine the construction effect.

[0029] Summary of the invention and specific implementation plan:

[0030] The purpose of this invention is to provide a method for evaluating the effectiveness of basic scour repair based on the TIN model. To achieve the above objective, this invention provides the following technical solution:

[0031] 1. Use the multi-beam sounding system to carry out pre-construction measurement on the wind turbine to be constructed to obtain three-dimensional point cloud data with a grid of 0.25 m before construction;

[0032] 2. Use the multi-beam sounding system to carry out post-construction measurement on the wind turbine to be constructed to obtain three-dimensional point cloud data with a grid of 0.25 m after construction;

[0033] 3. Construct a triangular mesh model (see the calculation flowchart Figure 1 )

[0034] 3.1 Point cloud data processing:

[0035] 3.1.1 Input: seabed topography point cloud data (three-dimensional coordinates: x i ,y i ,h i )

[0036] 3.1.2 Noise filtering: eliminate outliers by statistical outlier rejection (SOR) or radius filtering.

[0037] 3.1.3 Data simplification: use random sampling or curvature preservation algorithm to reduce computational complexity.

[0038] 3.2 Delaunay triangulation

[0039] 3.2.1 Definition: ensure that the circumcircle of any triangle does not contain other points, avoiding narrow triangular facets.

[0040] 3.2.3 Algorithm implementation steps:

[0041] Project the point set to the horizontal plane (xOy coordinate system).

[0042] Generate a triangular mesh based on the Bowyer-Watson algorithm iterative insertion of points.

[0043] 3.2.4 Output: TIN model composed of triangular facets, each triangular vertex containing elevation attribute.

[0044] 3.3 Calculate the difference between the design elevation

[0045] Extract the design elevation: according to the design file or drawing, find the design elevation H, which is generally a global unified elevation (such as the Yellow Sea elevation datum) or a relative value in the local coordinate system of the pile foundation;

[0046] Calculate the elevation difference: for each vertex in the triangular mesh, calculate the difference between its elevation and the design elevation H to obtain the elevation change value.

[0047] Calculation formula:

[0048] Δh i =h iH(i = 1, 2, …, n)

[0049] Symbol convention:

[0050] Δh i <0: Scour area (elevation decreases)

[0051] Δh i >0: Deposition area (elevation increases)

[0052] 3.4 Triangular patch classification

[0053] Decision rule: Prioritize classification according to vertex attribute proportion:

[0054] If ≥2 points in the triangle vertex are scour, the whole is classified as a scour area;

[0055] If ≥2 points in the vertex are deposition, it is classified as a deposition area.

[0056] Special case: If the number of vertex scour / deposition is equal, it is processed by area weight interpolation.

[0057] 3.5 Statistics of scour and deposition

[0058] Classification: According to the elevation change value, the points are divided into two categories: scour (elevation decreases) and deposition (elevation increases);

[0059] Volume calculation: Use the triangular net method to calculate the area of each small triangle, and combine the elevation change value to calculate the volume of the scour and deposition area.

[0060] Result summary: Add the scour and deposition amounts of each small triangle to get the pre-construction scour amount a and deposition amount b relative to the design elevation H, and the post-construction scour amount c and deposition amount d.

[0061] 3.5.1 Single triangular volume calculation

[0062] Triangular projected area (horizontal plane):

[0063] where, and are the plane vectors of the two sides of the triangle

[0064] Average elevation change (triangular interpolation): Δh avg = 1 / 3 × (Δh1 + Δh2 + Δh3)

[0065] Single volume unit: V = S·Δh avg

[0066] 3.5.2 Classification and result summary

[0067] Total scour (pre-construction):

[0068] Total amount of siltation (before construction):

[0069] Post-construction amount calculation: Repeat the above process, use post-construction point cloud data to regenerate TIN and compare, get post-construction scouring amount c and siltation amount d.

[0070] Total amount of engineering = (a-c) + (d-b).

[0071] 4. Error control and optimization strategy

[0072] 4.1 Precision improvement method

[0073] Encryption point cloud resolution: improve data density through multi-beam sonar or LiDAR.

[0074] Triangular net optimization: use constrained Delaunay triangulation for steep area (such as pile foundation edge).

[0075] 4.2 Abnormal processing

[0076] Remove the elevation mutation point: manually review the points with Δh> threshold value, and exclude measurement error.

[0077] Grid smoothing: use Laplacian filter to eliminate local oscillation.

[0078] 5. Example: the design mud surface elevation of a certain fan pile foundation is-13.55m. Before and after construction, it is scanned.

[0079] Repair engineering quantity calculation

[0080] Earthwork calculation uses triangular net TIN model calculation, the blue part on the map is the excavation part (not including pile foundation), and the red and black part is the throwing and filling part.

[0081] Table 1 wind turbine repair engineering quantity calculation (m 3 )

[0082]

[0083] 6. Use Surfer software to calculate elevation change and mapping of point cloud data before and after construction to determine construction effect.

[0084] The Riegl VQ-820-GLiDAR device is used to scan the construction area, with a resolution of 0.08 meters. This device uses laser pulse ranging combined with GPS positioning to collect high-density point cloud data within a 50-meter range around the pile foundation. Compared to multi-beam sounding systems, LiDAR can still obtain point cloud data with a spacing of ≤0.1 meters in a wave disturbance environment, allowing the minimum triangle side length in triangulation to be controlled within 0.2 meters, significantly improving the identification accuracy of the scour area boundary.

[0085] When the measurement accuracy of a multi-beam sounding system (such as the Kongsberg EM710) is 0.15 meters, the threshold is set to 0.25 meters (1.67 times the accuracy). During data preprocessing, points with Δh > 0.25 meters are manually reviewed: by comparing adjacent point cloud data and sonar images, abnormal values caused by seabed rock reflection or device jitter are excluded. In a certain project, 12% of outliers were removed through review, reducing the height deviation of the subsequent triangulation model from ±0.3 meters to ±0.1 meters.

[0086] When processing data from a strong tidal current scour area, five iterations of Laplacian filtering are used. The specific parameters are set as follows: filter coefficient λ = 0.5, and the point cloud coordinates are updated to the weighted average of the neighborhood points in each iteration. After processing, the local height oscillation (fluctuation amplitude ±0.4 meters) caused by sea current disturbance is smoothed to within ±0.1 meters, ensuring that the steep features (such as scour pit walls) around the pile foundation are retained during triangulation, while eliminating noise interference.

[0087] When constructing the triangulation around the wind turbine pile foundation, the initial point cloud spacing is 0.25 meters. Bowyer-Watson algorithm is used for iterative point insertion, and the process is automatically terminated when the distance between the new point and adjacent points is less than 0.03 meters. In the final generated triangulation, 90% of the triangle side lengths are between 0.1-0.3 meters, meeting the Delaunay criteria (circumscribed circle does not contain other points), and avoiding narrow triangles (minimum internal angle > 30°), with a volume calculation error controlled within 3%.

[0088] When processing the steep area at the junction of the pile foundation and the seabed, constrained Delaunay triangulation is used: the pile foundation edge line is used as the constraint edge, and the error allowed range is ±0.03 meters. This is achieved using the scipy.spatial library in Python. For a pile foundation with a diameter of 8 meters, the generated triangulation accurately captures the vertical drop (about 2.5 meters) at the pile foundation-seabed interface, reducing the volume calculation deviation in this area from 15% to 4% compared to unconstrained triangulation.

[0089] When two of the three vertices of a triangle are scouring (Δh = -0.5m, -0.3m), and one is silting (Δh = +0.2m), the areas of the three sub-regions corresponding to the three sides are 0.8㎡, 1.2㎡, and 1.0㎡ respectively. According to the area weight, the scouring proportion is (0.8+1.2) / (0.8+1.2+1.0) = 66.7%, so it is classified as a scouring area. Compared with the simple point determination, this method improves the classification accuracy by 20% in the transition zone of the terrain.

[0090] When calculating the projected area of a triangle (vertex coordinates A(0, 0, 0), B(1.234, 0.567, 0), C(0.345, 1.567, 0)), the vector cross product is used:

[0091] AB → =(1.234, 0.567, 0), AC → =(0.345, 1.567, 0),

[0092] Cross product length = |1.234 x 1.567 - 0.567 x 0.345| = 1.836,

[0093] Area S = 1 / 2 x 1.836 = 0.918㎡ (rounded to three decimal places).

[0094] According to the CAD software verification, the error of this calculation and the actual projected area is less than 0.001㎡.

[0095] The total amount of riprap in a certain preliminary calculation of a project is 1250m 3 , compared with the 1320m 3 recorded by the construction party, the error rate = |1250-1320| / 1320 = 5.3% > 5%, triggering the re-calculation mechanism. Through inspection, it is found that the constraint section parameter setting of the pile foundation edge in step three is incorrect. After adjusting the constraint edge error to ±0.03 meters, the total amount of re-calculation is 1305m 3 , the error rate is reduced to 1.1%, meeting the acceptance requirements.

[0096] Using Surfer software to render the pre-construction / post-construction triangular network model of a certain project, generate a three-dimensional view with elevation color spectrum, and generate a bar chart through Excel to compare the pre-construction scouring amount (a = 820m 3 ), post-construction scouring amount (c = 150m 3 ), pre-construction silting amount (b = 310m 3 ), post-construction silting amount (d = 1280m 3 ), total amount of the project = (820-150)+(1280-310) = 1640m 3Visual report helps the supervisor quickly locate the area of insufficient filling (such as the northeast side of the pile foundation erosion area is not completely covered) to guide the construction party to supplement the filling.

Claims

1. A method for evaluating the effectiveness of basic scour repair based on the TIN model, characterized in that, include: Step 1: Data Acquisition: Use a multibeam echo sounder to acquire 3D point cloud data with a grid of 0.25 before and after construction. LiDAR can be used to improve the data density. Step 2, Data Preprocessing: Outliers are removed by statistical outlier removal or radius filtering. Random sampling or curvature preservation algorithms are used to reduce data complexity. Points with Δh exceeding the set threshold are manually checked, and the grid is smoothed using Laplacian filtering. Step 3: Triangular mesh construction: Project the processed point set onto the horizontal plane, and generate a triangular mesh by iteratively inserting points using the Bowyer-Watson algorithm, ensuring that the circumcircle of any triangle does not contain other points. For steeply changing areas such as the edge of the pile foundation, constrained Delaunay triangulation is used to give each triangle vertex an elevation attribute. Step 4: Calculate the elevation difference: Based on the design elevation H, calculate the difference between each vertex of the triangular grid and H to obtain the elevation change value; Step 5, Area Classification: Classify areas based on the elevation changes of the three vertices of the triangle: If ≥2 points are scour areas, the entire area is classified as a scour zone; if ≥2 points are siltation areas, the entire area is classified as a siltation zone; if the number of scour / siltation points is equal, interpolate according to area weight. Step Six: Volume and Total Amount Calculation: First, calculate the projected area of ​​the horizontal plane of the triangle. Then, the average elevation change Δh is calculated using triangle interpolation. avg = 1 / 3 × (Δh1 + Δh2 + Δh3), thus yielding the single volume V = S·Δh avg The total amount of scouring and sedimentation before and after processing is calculated as follows: (Scrubbing amount before processing - Scrubbing amount after processing) + (Sedimentation amount after processing - Sedimentation amount before processing).

2. The method for evaluating the basic scour repair effect based on the TIN model according to claim 1, characterized in that, When using LiDAR to increase data density, the scanning resolution should be no less than 0.1 meters.

3. The method for evaluating the basic scour repair effect based on the TIN model according to claim 1, characterized in that, The set threshold is 1.5-2 times the measurement accuracy of the multibeam echo sounding system.

4. The method for evaluating the basic scour repair effect based on the TIN model according to claim 1, characterized in that, The Laplacian filter is iterated 3-5 times.

5. The method for evaluating the basic scour repair effect based on the TIN model according to claim 1, characterized in that, During the iterative insertion process of the Bowyer-Watson algorithm, insertion is stopped when the distance between adjacent points is less than 0.05 meters.

6. The method for evaluating the basic scour repair effect based on the TIN model according to claim 1, characterized in that, In the constrained Delaunay triangulation, the allowable error range for the constrained edges is ±0.03 meters.

7. The method for evaluating the basic scour repair effect based on the TIN model according to claim 1, characterized in that, When performing area-weighted interpolation, the final determination result is calculated based on the area proportion of different regions within the triangle.

8. The method for evaluating the basic scour repair effect based on the TIN model according to claim 1, characterized in that, When calculating the projected area of ​​a triangle on its horizontal plane, the result is rounded to three decimal places.

9. The method for evaluating the basic scour repair effect based on the TIN model according to claim 1, characterized in that, If the error rate of the total project calculation exceeds 5%, all steps of the method shall be re-executed.

10. A method for evaluating the basic scour repair effect based on a TIN model according to any one of claims 1-9, characterized in that, Once the calculations are complete, a visualization report is generated, which includes a 3D model display and data chart analysis.

Images