A method for investigating and restoring water seepage accidents caused by failure of bulkheads in underground powerhouses of hydropower stations

Abstract

The present invention discloses an investigation and restoration method for water seepage accidents caused by failure of bulkheads in underground powerhouses of hydropower stations, which belongs to the field of accident detection. The method efficiently collects three-dimensional information of important components such as bulkheads at the accident site under complex environmental conditions, completely splices the information through a feature fitting algorithm, quantitatively analyzes key parameters of the bulkheads, and accurately restores the failure process of the bulkheads. It solves the problems of incomplete three-dimensional point cloud data collection, low accuracy of feature extraction and registration of point cloud data, and lack of dynamic restoration capabilities in complex environments, and forms a quantitative and precise investigation system for bulkhead failure accidents from data collection, feature analysis to dynamic simulation, breaking through the limitations of traditional methods in integrity, accuracy and mechanism restoration, providing technical support for the safe operation of hydropower stations, and providing a methodological reference for investigating the causes of similar accidents.

Description

Technical Field

[0001] The present invention relates to the field of accident detection, and in particular to a method for investigating and restoring a water seepage accident caused by a failure of a bulkhead in an underground powerhouse of a hydropower station. Background Art

[0002] As a critical sealing structure in pressure piping systems, the failure of bulkheads can lead to catastrophic water leaks. Traditional accident investigation methods rely on manual surveys, two-dimensional measurements, and empirical inference, making it difficult to accurately reconstruct the mechanism of bulkhead failure. With the application of 3D scanning technology, some studies have attempted to analyze the cause of accidents through reverse modeling of point cloud data. However, due to complex operating conditions and technical bottlenecks, existing methods still have significant deficiencies in data integrity, model accuracy, and dynamic recovery capabilities. Specifically, they are manifested in the following aspects: First, data acquisition integrity is insufficient. Existing 3D scanning technologies often use fixed-resolution single-site scanning, which is susceptible to occlusion and reflection interference in the complex environment of hydropower station underground powerhouses, resulting in point cloud loss in critical areas (such as the bulkhead fracture surface). Furthermore, after the bulkhead detached body is separated from the remaining body, existing methods often ignore the independent scanning of the detached body and rely solely on the on-site remaining body data for reconstruction, resulting in distorted model geometric features and failing to reflect the true shape before failure. Second, the accuracy of point cloud processing and feature extraction is low. Traditional denoising algorithms have difficulty distinguishing between noise and subtle structural features, especially in critical areas such as bulkhead welds and bolt holes. Edge detection mostly relies on the normal vector variance method, but it lacks sensitivity to areas with sudden changes in curvature, resulting in large deviations in the fracture path of fracture line extraction. The least squares method is often used for surface center fitting, without considering the distribution deviation caused by asymmetric wear. In addition, mainstream registration technology (such as the standard ICP algorithm) is often used for point cloud data registration, but this technology only relies on minimizing geometric distances and does not incorporate the actual stress conditions of the blind head (such as water flow pressure distribution and bolt preload), resulting in insufficient mechanical rationality of the assembled model. Third, there is a lack of dynamic recovery capabilities. Existing studies mostly generate three-dimensional models based on static point clouds, and can only infer the failure process through cross-sectional analysis. There is a lack of dynamic simulation of the crack propagation path and the blind head detachment trajectory, and the extraction of key parameters (such as the initiation point and stress peak) has low reliability. Summary of the Invention

[0003] In view of the above-mentioned deficiencies in the prior art, the present invention provides a method for investigating and restoring water seepage accidents caused by failure of bulkheads in underground powerhouses of hydropower stations.

[0004] In order to achieve the above-mentioned object of the invention, the technical solution adopted by the present invention is:

[0005] A method for investigating and restoring a water seepage accident caused by a failure of a bulkhead in an underground powerhouse of a hydropower station comprises the following steps:

[0006] S1. Collect 3D point cloud data of the remaining bulkhead at the accident site of the hydropower station's underground powerhouse through non-contact multi-station 3D scanning, and collect 3D point cloud data of the bulkhead detached body through non-contact multi-station 3D scanning in a stable indoor location;

[0007] S2, performing splicing and denoising preprocessing on the complex environment three-dimensional point cloud data of the bulkhead residual body and the three-dimensional point cloud data of the bulkhead detached body obtained in S1;

[0008] S3, calculating the minimum variance unbiased estimate of the surface center of the spliced 3D point cloud data of the bulkhead detached body and the 3D point cloud data of the bulkhead residual body, and using the vector method to extract the edge feature lines of the bulkhead residual body and the bulkhead detached body to generate an edge point cloud data set;

[0009] S4, forcibly superimposing the minimum variance unbiased estimate of the surface center of the three-dimensional point cloud data of the bulkhead detached body and the three-dimensional point cloud data of the bulkhead residual body calculated in S3, and calculating the optimal solution of the surface splicing rotation matrix based on the obtained edge point cloud data set;

[0010] S5. Perform the point cloud coordinate rotation transformation again to complete the complete point cloud splicing of the bulkhead body and obtain the complete point cloud features of the bulkhead body. Perform three-dimensional quantitative analysis based on the obtained complete point cloud features of the bulkhead body to restore the accident process.

[0011] Furthermore, the S1 specifically includes the following steps:

[0012] S11. Set the scanning route and arrange the scanning stations according to the visibility, stability, overlap conditions of the scanning site and the location and shape of the scanning object;

[0013] S12. Level the 3D scanning instrument on the designed site, evenly arrange the common targets for splicing, and select the resolution and accuracy scanning parameters to collect 360-degree 3D information of the site;

[0014] S13, resetting the scanning parameters, performing a local encrypted scan on the bulkhead residue, and collecting data for the next site after completion;

[0015] S14, collecting three-dimensional point cloud data of the bulkhead detached body through non-contact multi-station three-dimensional scanning;

[0016] Furthermore, the S2 specifically includes the following steps:

[0017] S21. De-noise the cloud data of each site according to distance range, reflectivity, isolated points, and curvature factors;

[0018] S22. Uniformly number the public targets in the cloud data of each site, and automatically identify and extract the center points of the public targets using a 3D point cloud target detection algorithm;

[0019] S23. Coordinate registration and transformation calculation are performed on the data of adjacent sites, and weighted fusion is performed on the overlapping areas, so as to obtain point cloud data of the spliced bulkhead residual body and the bulkhead detached body respectively.

[0020] Furthermore, the S3 specifically includes the following steps:

[0021] S31, respectively extracting the maximum numerical profile curves of the spliced bulkhead residual body and bulkhead detached body point cloud data;

[0022] S32, let the maximum value profile curves of the point cloud data of the bulkhead residual body and the bulkhead detached body be on the same plane, and is the center of the circle, Fit the radius and calculate the optimization objective function of the square of its error;

[0023] S33, let the optimization objective function of the square of the error be the minimum value, and use the quadratic differential equation to find the minimum value of the center and radius of the plane S32, and obtain the minimum square error unbiased estimate of the surface center of the point cloud data of the bulkhead residual body and the bulkhead detached body;

[0024] S34, construct any sampling point A spherical domain with a radius of 2 times the point cloud resolution is centered, and vectors between a point and each point in the domain are constructed and normalized, and the sum of the vectors in the domain is calculated;

[0025] S35, judging the position of the point according to the magnitude of the sum of the calculated vectors in the domain, where a larger value indicates that the point is closer to the edge;

[0026] S36. Uniformly sample the point cloud data of the bulkhead residual body and the bulkhead detached body, repeat steps S34-S35, and filter according to the maximum value of the sum of the vectors in the domain to obtain an edge point cloud data set.

[0027] Furthermore, the optimization objective function of the square error in S32 is expressed as:

[0028]

[0029] Where, is the optimization objective function of the square error, It is the characteristic point of the surface of the point cloud data of the bulkhead residual body or bulkhead detached body. is the feature point index, is the total number of feature points.

[0030] Furthermore, the S4 specifically includes the following steps:

[0031] S41. Assume that the center coordinates of the surface of the bulkhead residual body are , the center coordinates of the surface of the bulkhead detached body are , calculate the translation vector as , apply translation transformation to all point cloud coordinates of the bulkhead detached body to force the surface center of the bulkhead residual body to coincide with the surface center of the bulkhead detached body, eliminating the global position deviation;

[0032] S42, for each point in the point cloud data of the edge of the bulkhead detached body after translation, searching for the point with the closest Euclidean distance in the bulkhead residual body to form a point pair set, calculating the centroid of the point pair set, then constructing a covariance matrix and performing singular value decomposition on the covariance matrix to obtain an initial solution of the surface splicing rotation matrix;

[0033] S43, repeating steps S41-S42 until the registration error converges or the maximum number of iterations is reached;

[0034] S44. Using the obtained initial solution of the surface splicing rotation matrix as the starting point, nonlinear optimization of the splicing objective function PT function is performed. When the change of the splicing objective function PT function is less than a threshold or reaches the maximum number of iterations, the optimal solution of the splicing rotation matrix is obtained.

[0035] Furthermore, the stitching objective function PT in S44 is expressed as:

[0036]

[0037] Where, is the initial solution of the surface splicing rotation matrix, is the traditional ICP registration error, is the surface curvature consistency constraint, is the mechanical equilibrium constraint term, , , is the weight coefficient, which is calibrated according to specific experiments.

[0038] Furthermore, the S5 specifically includes the following steps:

[0039] S51, applying a rotation matrix to the translated point cloud of the bulkhead detached body, and further merging the obtained rotated data with the point cloud of the bulkhead residual body to obtain three-dimensional point cloud data of the complete bulkhead body;

[0040] S52. Mark potential fracture areas using a curvature mutation detection algorithm, convert the 3D point cloud data of the complete bulkhead into a NURBS surface or voxel model, and reconstruct the 3D geometric structure of the bulkhead;

[0041] S53. Set the accident condition boundary, perform transient dynamic simulation, and simulate the failure process of the bulkhead.

[0042] The present invention has the following beneficial effects:

[0043] ① In terms of data collection and modeling, multi-site dynamic parameter adjustment and local encrypted scanning technology are used, combined with reflectivity and curvature multimodal denoising algorithms to overcome the problem of occlusion and interference in the complex environment of underground factories.

[0044] ② Using vector edge detection and minimum variance unbiased surface fitting, the method reduces the error in fracture path extraction and improves the accuracy of circle center positioning. During the model assembly process, the method innovatively incorporates mechanical balance constraints and curvature consistency into point cloud registration, constructing a multi-objective optimization function (PT), overcoming the limitations of traditional ICP algorithms that rely solely on geometric alignment.

[0045] ③ In response to the difficulty of restoring the dynamic process of the accident, the present invention extracts parameters such as fracture lines, weld lines, and detonation points, and combines virtual reality technology to simulate the crack propagation path, the trajectory of the blind head falling off, and the water seepage process, thus achieving a leap from "two-dimensional empirical judgment" to "three-dimensional data-driven analysis."

[0046] This invention has significant innovations in three aspects: method architecture (3D scanning + quantitative restoration), algorithm improvement (edge extraction, stitching optimization) and application scenario (hydropower station bulkhead failure investigation), which significantly improves the accuracy and efficiency of hydropower station bulkhead failure accident investigation. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 The present invention is a flow chart of the method for investigating and restoring a water seepage accident caused by a failure of a bulkhead in an underground powerhouse of a hydropower station. DETAILED DESCRIPTION

[0048] The specific embodiments of the present invention are described below to facilitate understanding of the present invention by those skilled in the art. However, it should be clear that the present invention is not limited to the scope of the specific embodiments. For those skilled in the art, as long as various changes are within the spirit and scope of the present invention as defined and determined by the appended claims, these changes are obvious, and all inventions and creations utilizing the concepts of the present invention are protected.

[0049] A method for investigating and restoring water seepage accidents caused by failure of bulkheads in underground powerhouses of hydropower stations, such as Figure 1 As shown, the following steps are included:

[0050] S1. Collect 3D point cloud data of the remaining bulkhead (R) at the accident site of the hydropower station's underground powerhouse through non-contact multi-station 3D scanning, and collect 3D point cloud data of the detached bulkhead (L) through non-contact multi-station 3D scanning in a stable indoor location.

[0051] In this embodiment, non-contact multi-site 3D scanning is used to collect 3D point cloud data of the deadhead residue (R) and complex environment at the accident site of the underground powerhouse of the hydropower station.

[0052] S11. Set the scanning route and arrange the scanning stations according to the visibility, stability, overlap conditions of the scanning site and the location and shape of the scanning object;

[0053] S12. Level the 3D scanning instrument on the designed site, evenly arrange the common targets for splicing, and select the resolution and accuracy scanning parameters to collect 360-degree 3D information of the site.

[0054] S13, resetting the scanning parameters, performing a local encrypted scan on the bulkhead residue, and collecting data for the next site after completion;

[0055] S14, collecting three-dimensional point cloud data of the bulkhead detached body (L) through non-contact multi-station three-dimensional scanning;

[0056] S2, splicing and preprocessing the complex environment three-dimensional point cloud data of the bulkhead residual body and the three-dimensional point cloud data of the bulkhead detached body obtained in S1;

[0057] In this embodiment, the following steps are specifically included:

[0058] S21. De-noise the cloud data of each site according to distance range, reflectivity, isolated points, and curvature factors;

[0059] S22. Uniformly number the public targets in the cloud data of each site, and automatically identify and extract the center points of the public targets using a 3D point cloud target detection algorithm;

[0060] S23. Coordinate registration and transformation calculation are performed on the data of adjacent sites, and weighted fusion is performed on the overlapping areas to obtain the point cloud data of the spliced bulkhead residual body (R) and the bulkhead detached body (L).

[0061] S3, calculating the minimum variance unbiased estimate of the surface center of the spliced 3D point cloud data of the bulkhead detached body and the 3D point cloud data of the bulkhead residual body, and using the vector method to extract the edge feature lines of the bulkhead residual body and the bulkhead detached body to generate an edge point cloud data set;

[0062] In this embodiment, the following steps are specifically included:

[0063] S31, extract the maximum numerical profile curve (Q R , Q L );

[0064] S32, let the maximum value profile curves of the point cloud data of the bulkhead residual body and the bulkhead detached body be on the same plane, and is the center of the circle, Fit the radius and calculate the optimization objective function of the square of its error;

[0065] Let QR , Q L In the same plane, the center of the plane is and radius The optimization objective function of the square error of the fitting is:

[0066] ,

[0067] in( , ) is Q R or Q L Feature points on the curve.

[0068] S33, let the optimization objective function of the square of the error be the minimum value, and use the quadratic differential equation to find the minimum value of the center and radius of the S32 plane to obtain the minimum variance unbiased estimate of the surface center of the point cloud data of the bulkhead residual body and the bulkhead detached body. 、 ;

[0069] The vector method is used to intelligently extract R and L edge feature lines and generate edge point cloud datasets.

[0070] S34, construct any sampling point The collection of points in a spherical area with a radius of 2 times the point cloud resolution as the center , k is the number of points in the field, construct the point and the vectors of each point in its domain , and perform vector normalization, then calculate the sum of the vectors in the domain ;

[0071] S35, according to The value is judged intelligently, the larger the value, the more Beyond the edge;

[0072] S36, uniformly sample the point cloud data of the bulkhead residual body and the bulkhead detached body, repeat steps S34-S35, and calculate the sum of the vectors in the domain. The maximum value of is screened to obtain the edge point cloud dataset.

[0073] S4, forcing the minimum variance unbiased estimate of the surface center calculated in S3 to coincide, and calculating the optimal solution of the surface splicing rotation matrix based on the obtained edge point cloud data set;

[0074] This embodiment specifically includes the following steps:

[0075] S41. Assume that the center coordinates of the surface of the bulkhead residual body are , the center coordinates of the surface of the bulkhead detached body are , calculate the translation vector as , then apply translation transformation to all point cloud coordinates of the bulkhead dropout (DL) , force the surface center of L to coincide with the surface center OR of R to eliminate the global position deviation;

[0076] S42, for each point in the point cloud data of the edge of the bulkhead detached body after translation, searching for the point with the closest Euclidean distance in the bulkhead residual body to form a point pair set, calculating the centroid of the point pair set, then constructing a covariance matrix and performing singular decomposition on the covariance matrix to obtain an initial solution of the surface splicing rotation matrix;

[0077] Based on the ICP registration of bulkhead detached body and bulkhead residual body,

[0078] ① The edge point cloud dataset of the bulkhead detached body after translation Each point in , search for the point with the closest Euclidean distance in DR , forming a point pair set .

[0079] ②Calculate the centroid of the two groups of points , , then construct the covariance matrix , perform singular value decomposition (SVD) on H: , get the initial solution of the surface splicing rotation matrix .

[0080] in is the transpose of the matrix, is a left singular vector matrix whose column vectors are The feature vector represents the main direction of the target point cloud (DR). is a right singular vector matrix whose column vectors are The feature vector of the point cloud to be registered main direction. is a diagonal matrix containing The singular values of represent the alignment strength of the main directions of the two sets of point clouds.

[0081] S43, repeating steps S41-S42 until the registration error converges or the maximum number of iterations is reached;

[0082] S44. Using the obtained initial solution of the surface splicing rotation matrix as a starting point, nonlinear optimization of the PT function is performed. When the change of the splicing objective function PT function is less than a threshold or reaches the maximum number of iterations, the optimal solution of the splicing rotation matrix is obtained.

[0083] Since the ICP algorithm only relies on geometric distance matching, it will ignore the continuity of the bulkhead surface, mechanical constraints, etc., and optimize the PT function.

[0084]

[0085] Where, is the initial solution of the surface splicing rotation matrix, is the traditional ICP registration error, is the mechanical equilibrium constraint term, , , is the weight coefficient, which is calibrated according to specific experiments. is the surface curvature consistency constraint, and calculates the curvature difference between R and L after splicing. It is a mechanical equilibrium constraint item. Based on the stress model before the failure of the bulkhead (such as water flow pressure, bolt preload, etc.), the rationality of the stress after assembly is verified. , , is the weight coefficient, which is calibrated according to specific experiments.

[0086] To obtain As the starting point, the Levenberg-Marquardt algorithm is used to perform nonlinear optimization on PT(T). When the change is less than the threshold (such as 1e-5) or the maximum number of iterations is reached, the optimal solution of the surface splicing rotation matrix is obtained ( ).

[0087] S5. Perform the point cloud coordinate rotation transformation again to complete the complete point cloud splicing of the bulkhead body and obtain the complete point cloud features of the bulkhead body. Perform three-dimensional quantitative analysis based on the obtained complete point cloud features of the bulkhead body to restore the accident process.

[0088] In this embodiment, the following steps are specifically included:

[0089] S51, applying a rotation matrix to the translated point cloud of the bulkhead detached body, and further merging the obtained rotated data with the point cloud of the bulkhead residual body to obtain three-dimensional point cloud data of the complete bulkhead body;

[0090] Point cloud of the detached body after translation Applying a rotation matrix

[0091]

[0092] Will Merge with the residual body point cloud DR to obtain the complete bulkhead body (W) three-dimensional point cloud data.

[0093] S52. Mark potential fracture areas using a curvature mutation detection algorithm, convert the 3D point cloud data of the complete bulkhead into a NURBS surface or voxel model, and reconstruct the 3D geometric structure of the bulkhead;

[0094] Potential fracture areas are marked using a sudden change curvature detection algorithm, and the fracture path is manually corrected to extract parameters such as fracture line length, strike angle, and fracture roughness. Based on point cloud reflectivity differences, weld areas are identified, weld width and penetration are measured, and near the fracture endpoints, local stress concentration analysis is used to determine the energy release source and the detonation point. Discontinuous areas in the assembled point cloud are extracted, and the location of the break is verified by combining the bolt hole coordinates. The break point is then determined, and the point cloud W is converted into a NURBS surface or voxel model to reconstruct the 3D geometry of the bulkhead.

[0095] S53. Set the accident condition boundary, perform transient dynamic simulation, and simulate the failure process of the bulkhead.

[0096] Set the accident condition boundaries (such as water flow pressure, temperature, etc.), perform instantaneous dynamic simulation, simulate the failure process of the bulkhead, and combine virtual reality technology to display the crack propagation, bulkhead detachment trajectory and water seepage process in time series.

[0097] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0098] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0099] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0100] Specific embodiments are used in the present invention to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core ideas. At the same time, for those skilled in the art, according to the ideas of the present invention, there may be changes in the specific implementation methods and application scopes. In summary, the contents of this specification should not be understood as limiting the present invention.

[0101] Those skilled in the art will appreciate that the embodiments described herein are intended to help readers understand the principles of the present invention, and it should be understood that the scope of protection of the present invention is not limited to such specific descriptions and embodiments. Those skilled in the art can make various other specific variations and combinations based on the technical teachings disclosed in the present invention without departing from the essence of the present invention, and such variations and combinations are still within the scope of protection of the present invention.

Claims

1. A method for investigating and restoring water seepage accidents caused by failure of bulkheads in underground powerhouses of hydropower stations, characterized in that: The steps include: S1. Collect 3D point cloud data of the remaining bulkhead at the accident site of the hydropower station's underground powerhouse through non-contact multi-station 3D scanning, and collect 3D point cloud data of the bulkhead detached body through non-contact multi-station 3D scanning in a stable indoor location; S2, performing splicing and denoising preprocessing on the complex environment three-dimensional point cloud data of the bulkhead residual body and the three-dimensional point cloud data of the bulkhead detached body obtained in S1; S3, calculating the minimum variance unbiased estimate of the surface center of the spliced 3D point cloud data of the bulkhead detached body and the 3D point cloud data of the bulkhead residual body, and using the vector method to extract the edge feature lines of the bulkhead residual body and the bulkhead detached body to generate an edge point cloud data set; S4, forcibly superimposing the minimum variance unbiased estimate of the surface center of the three-dimensional point cloud data of the bulkhead detached body and the three-dimensional point cloud data of the bulkhead residual body calculated in S3, and calculating the optimal solution of the surface splicing rotation matrix based on the obtained edge point cloud data set, specifically including the following steps: S41. Assume that the center coordinates of the surface of the bulkhead residual body are , the center coordinates of the surface of the bulkhead detached body are , calculate the translation vector as , apply translation transformation to all point cloud coordinates of the bulkhead detached body to force the surface center of the bulkhead residual body to coincide with the surface center of the bulkhead detached body, eliminating the global position deviation; S42, for each point in the point cloud data of the edge of the bulkhead detached body after translation, searching for the point with the closest Euclidean distance in the bulkhead residual body to form a point pair set, calculating the centroid of the point pair set, then constructing a covariance matrix and performing singular value decomposition on the covariance matrix to obtain an initial solution of the surface splicing rotation matrix; S43, repeating steps S41-S42 until the registration error converges or the maximum number of iterations is reached; S44, using the obtained initial solution of the surface splicing rotation matrix as the starting point, a nonlinear optimization of the splicing objective function PT function is performed. When the change of the splicing objective function PT function is less than a threshold or the maximum number of iterations is reached, the optimal solution of the splicing rotation matrix is obtained, wherein the splicing objective function PT is expressed as: Where, is the initial solution of the surface splicing rotation matrix, is the traditional ICP registration error, is the surface curvature consistency constraint, is the mechanical equilibrium constraint term, , , is the weight coefficient, which is calibrated according to specific experiments; S5. Perform the point cloud coordinate rotation transformation again to complete the complete point cloud splicing of the bulkhead body and obtain the complete point cloud features of the bulkhead body. Perform three-dimensional quantitative analysis based on the obtained complete point cloud features of the bulkhead body to restore the accident process.

2. The method for investigating and restoring a water seepage accident caused by a failure of a bulkhead in an underground powerhouse of a hydropower station according to claim 1, characterized in that: The S1 specifically includes the following steps: S11. Set the scanning route and arrange the scanning stations according to the visibility, stability, overlap conditions of the scanning site and the location and shape of the scanning object; S12. Level the 3D scanning instrument on the designed site, evenly arrange the common targets for splicing, and select the resolution and accuracy scanning parameters to collect 360-degree 3D information of the site; S13, resetting the scanning parameters, performing a local encrypted scan on the bulkhead residue, and collecting data for the next site after completion; S14. Collect three-dimensional point cloud data of the bulkhead detached body through non-contact multi-station three-dimensional scanning.

3. The method for investigating and restoring a water seepage accident caused by a failure of a bulkhead in an underground powerhouse of a hydropower station according to claim 1, characterized in that: The S2 specifically includes the following steps: S21. De-noise the cloud data of each site according to distance range, reflectivity, isolated points, and curvature factors; S22. Uniformly number the public targets in the cloud data of each site, and automatically identify and extract the center points of the public targets using a 3D point cloud target detection algorithm; S23. Coordinate registration and transformation calculation are performed on the data of adjacent sites, and weighted fusion is performed on the overlapping areas, so as to obtain point cloud data of the spliced bulkhead residual body and the bulkhead detached body respectively.

4. The method for investigating and restoring a water seepage accident caused by a failure of a bulkhead in an underground powerhouse of a hydropower station according to claim 1, characterized in that: The S3 specifically includes the following steps: S31, respectively extracting the maximum numerical profile curves of the spliced bulkhead residual body and bulkhead detached body point cloud data; S32, let the maximum value profile curves of the point cloud data of the bulkhead residual body and the bulkhead detached body be on the same plane, and is the center of the circle, Fit the radius and calculate the optimization objective function of the square of its error; S33, let the optimization objective function of the square of the error be the minimum value, and use the quadratic differential equation to find the minimum value of the center and radius of the plane S32, and obtain the minimum square error unbiased estimate of the surface center of the point cloud data of the bulkhead residual body and the bulkhead detached body; S34, construct any sampling point A spherical domain with a radius of 2 times the point cloud resolution is centered, and vectors between a point and each point in the domain are constructed and normalized, and the sum of the vectors in the domain is calculated; S35, judging the position of the point according to the magnitude of the sum of the calculated vectors in the domain, where a larger value indicates that the point is closer to the edge; S36. Uniformly sample the point cloud data of the bulkhead residual body and the bulkhead detached body, repeat steps S34-S35, and filter according to the maximum value of the sum of the vectors in the domain to obtain an edge point cloud data set.

5. The method for investigating and restoring water seepage accidents caused by failure of bulkheads in underground powerhouses of hydropower stations according to claim 4 is characterized in that: The optimization objective function of the square error in S32 is expressed as: Where, is the optimization objective function of the square error, It is the characteristic point of the surface of the point cloud data of the bulkhead residual body or bulkhead detached body. is the feature point index, is the total number of feature points.

6. The method for investigating and restoring a water seepage accident caused by a failure of a bulkhead in an underground powerhouse of a hydropower station according to claim 1, characterized in that: The S5 specifically includes the following steps: S51, applying a rotation matrix to the translated point cloud of the bulkhead detached body, and further merging the obtained rotated data with the point cloud of the bulkhead residual body to obtain three-dimensional point cloud data of the complete bulkhead body; S52. Mark potential fracture areas using a curvature mutation detection algorithm, convert the 3D point cloud data of the complete bulkhead into a NURBS surface or voxel model, and reconstruct the 3D geometric structure of the bulkhead; S53. Set the accident condition boundary, perform transient dynamic simulation, and simulate the failure process of the bulkhead.

Images