Deformation detection device and method for vertical metal tank

Through a three-dimensional laser scanner, the inner wall of the metal tank body is comprehensively scanned and data processing is solved, and the problems of cumbersome detection methods and insufficient accuracy in the existing technology are achieved, and the deformation of the metal tank body is quickly and accurately detected.

CN118913126APending Publication Date: 2024-11-08DALIAN INST OF METROLOGY INSPECTION & TESTING CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411039346.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-31
Publication Date
2024-11-08

AI Technical Summary

Technical Problem

The existing railway tanker tank body deformation detection methods are cumbersome and cannot obtain local and overall precise deformation details, and the accuracy and speed of data measurement are insufficient.

Method used

A three-dimensional laser scanner is used to comprehensively scan the inner wall of the vertical metal tank to generate three-dimensional coordinates and point cloud data. By processing and reconstructing the three-dimensional model, slice processing is used to obtain radial deviation values ​​and ellipticity data, and analyze the deformation of the tank.

Benefits of technology

It realizes rapid, complete and accurate detection of metal tank deformation, and improves the accuracy of detection data.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118913126A_ABST
    Figure CN118913126A_ABST
Patent Text Reader

Abstract

The invention discloses a deformation detection device and method for a vertical metal tank, and belongs to the technical field of metal tank deformation detection, and the device comprises a storage tank body, a sealing ring groove is formed in the storage tank body, a sealing assembly is arranged in the sealing ring groove, an outer floating roof is slidably connected to the inner side of the storage tank body, and a floating disc is arranged at the top of the outer floating roof. And a three-dimensional laser scanner is arranged at the top of the floating disc. According to the deformation detection method of the vertical metal tank, the three-dimensional laser scanner is used for comprehensively scanning the inner wall of the tank body of the storage tank, the obtained three-dimensional point cloud data is processed and then the three-dimensional model is reconstructed, the three-dimensional point cloud data is sliced, and the radial deviation value and the ovality data of the obtained tangent plane point cloud data are solved. Therefore, the deformation data of the storage tank body can be analyzed, the deformation part of the storage tank body can be quickly and completely detected, and the accuracy of the detection data is improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention belongs to the technical field of metal can deformation detection, and in particular relates to a deformation detection method for a vertical metal can. Background Art

[0002] Large storage tanks are one of the commonly used equipment in petrochemical and pipeline enterprises. They belong to the category of atmospheric pressure container boxes and thin-walled structures. After long-term use, the deformation of the storage tank is mainly due to the permanent deformation caused by the long-term accumulated elastic deformation and the flexible deformation caused by other reasons. At the same time, the strength of the tank cylinder changes during the welding process, causing the tank body to produce permanent plastic deformation. When the tank body and the base frame are assembled, the plastic deformation is affected by the clamping force of the tank belt, and the accumulation gradually increases, resulting in a certain deflection deformation of the tank body. The gravity of the liquid in the tank causes the tank body to produce downward elastic deformation. Each transportation will turn part of the elastic deformation into a small permanent plastic deformation, which will form a certain deflection deformation after accumulation. Therefore, the shape and structure of the storage tank needs to be inspected regularly.

[0003] After searching, it was found that a contactless detection device and method for deformation of a railway tank car tank body was disclosed in an invention patent with Chinese patent publication number CN117387509A. The contactless detection method for deformation of a railway tank car tank body in the invention patent is to use a pan-tilt head to drive a contactless distance measuring instrument to obtain depth measurement data of the entire radial cross-section of the tank body detection part for multiple times, mark the detected data values ​​with connecting lines in the established coordinate system to obtain a measured cross-sectional diagram, and then compare the measured cross-sectional diagram with the theoretical cross-sectional diagram of the tank body, so as to obtain the actual deformation data of the tank body detection part, and then obtain the actual deformation amount of the detection part.

[0004] However, the above-mentioned non-contact detection device and method for the deformation of the tank body of a railway tank car adopts a method of multiple measurements using a non-contact distance measuring instrument, and its operation steps are relatively cumbersome. At the same time, this method still belongs to single-point measurement, that is, deformation data of a few observation points on a certain radial section are obtained to infer the overall deformation, and it is impossible to obtain accurate local and overall deformation details. Therefore, the accuracy and speed of data measurement need to be further improved. Therefore, a deformation detection method for a vertical metal tank is proposed. Summary of the invention

[0005] In order to solve the above-mentioned problems, the present invention proposes: a deformation detection device for a vertical metal tank, comprising a tank body, a sealing ring groove is opened inside the tank body, a sealing assembly is arranged inside the sealing ring groove, an external floating roof is slidably connected to the inner side of the tank body, a floating plate is arranged on the top of the external floating roof, and a three-dimensional laser scanner is arranged on the top of the floating plate.

[0006] A method for detecting deformation of a vertical metal can comprises the following steps:

[0007] S1. The liquid in the tank is discharged through the discharge pipe. The external floating roof and the floating plate descend with the liquid level. The inner wall of the tank is scanned by a three-dimensional laser scanner installed on the top of the floating plate.

[0008] S2, constructing a coordinate system toward the target deformation area according to the three-dimensional laser scanner, and obtaining the three-dimensional coordinates of the scanned storage tank body;

[0009] S3, generating three-dimensional point cloud data according to the acquired three-dimensional coordinates, and processing the acquired three-dimensional point cloud data;

[0010] S4, reconstructing a three-dimensional model of the storage tank body according to the processed three-dimensional point cloud data, and observing the deformation cloud map data of the inner wall of the storage tank body;

[0011] S5, performing uniform horizontal slicing on the processed three-dimensional point cloud data to obtain several sections of point cloud data of the tank body;

[0012] S6. Fit the center coordinates of each section according to the obtained section point cloud data, calculate the center radius of each section, obtain and analyze the radial deviation value and ellipticity data of the target deformation area, and obtain the deformation data of the tank body.

[0013] Furthermore, in step S1, the three-dimensional laser scanner includes a laser transmitter and a laser receiver, which transmit a detection laser to the tank body and receive a reflected laser signal. The three-dimensional laser scanner signal is connected to a computer microprocessor, and the computer microprocessor data is connected to a data storage device. The laser transmitter of the laser rangefinder is used to transmit a detection laser to the tank body and the deformed part to be detected. After being received by the laser receiver, the distance is measured by the distance measurement module, and then the scanning control module of the reflecting prism can measure the horizontal direction angle and the vertical direction angle of the scan.

[0014] Furthermore, in step S2, the three-dimensional laser scanner is provided with a distance measurement module and a scanning control module, the position of the three-dimensional laser scanner is set to the coordinate origin O, an xyz coordinate system is established according to the coordinate origin O, and a plurality of observation scanning points F are selected in the scanning area of ​​the inner wall of the storage tank. k (k = 0, 1, 2, ..., n), according to the pulse laser ranging algorithm of the distance measurement module, the distance from the laser beam of the 3D laser scanner to each observation scanning point F is calculated. k The distance value S k (k=0, 1, 2, ..., n).

[0015] Furthermore, in step S3, the scanning control module includes an angle counter for calculating the angle ɑ between the laser beam and the xOy plane and the angle θ between the projection of the laser beam on the xOy plane and the positive x axis, and calculating the angle of each observation scanning point F k The three-dimensional coordinate value d in the xyz coordinate system k (x k ,y k , z k )(k=0, 1, 2, ..., n), the calculation method is as follows:

[0016] For each observation scanning point F k The three-dimensional coordinate value d k (x k ,y k , z k ) is integrated to obtain the three-dimensional point cloud data of the tank body.

[0017] Furthermore, in step S3, processing the three-dimensional point cloud data includes the following steps:

[0018] S31. Perform stitching and registration processing on the three-dimensional point cloud data in the point cloud data processing software, and delete the point cloud data that exceeds the scanning setting range;

[0019] S32. De-noising and filtering are performed on the three-dimensional point cloud data after the splicing and registration processing based on the mean curvature flow algorithm. The algorithm is as follows:

[0020] Select a vertex H in the 3D point cloud data, estimate the normal vector and curvature of the vertex H based on the geometric information contained in the local neighborhood of the selected vertex H, and move the vertex H to the model surface on the normal vector. The vertex H in the mean curvature direction is estimated as follows:

[0021]

[0022] where k H is the mean curvature, e ij =X j -X i , β j ,ω j for e ij The opposite angle between two adjacent triangles, A is the angle around X i The area of ​​the triangle is obtained, and the vertex H is moved to the surface of the smooth surface at a speed of the average curvature of the local surface in the direction of the normal vector, and a denoising filter is performed to obtain the denoising filtered three-dimensional point cloud data.

[0023] Furthermore, in step S4, the 3D point cloud data after denoising and filtering is 3D reconstructed in the 3D modeling software, and the 3D reconstruction steps include point cloud coloring processing, polygon processing, contour line and surface patch processing, grid processing and NURBS surface processing, and the processed data file is output to obtain a 3D reconstructed model of the tank body, and the deformation cloud map of the 3D reconstructed model is observed.

[0024] Furthermore, in step S5, the three-dimensional point cloud data is sliced ​​to obtain a plurality of point cloud slices, each of which is spaced 0.6 m apart and has a thickness of 0.001 m. The center coordinates of each point cloud slice are fitted and the center diameter is calculated according to the least squares method.

[0025] Furthermore, in step S6, according to the center coordinates and center diameter of each point cloud slice, a fitting center and a fitting diameter are obtained by fitting, which are used as standard parameters, and the radial deviation value is calculated as follows:

[0026] σ=LS, where σ is the radial deviation value of the point cloud slice, L is the distance from the point cloud data of the remaining point cloud slices except the standard point cloud slice to the center of the circle, and S is the distance from the point cloud data of the standard point cloud slice to the center of the circle;

[0027] The ellipticity is calculated as follows:

[0028] Where η is the lateral ellipticity of the point cloud slice, L is the average diameter of the point cloud slice, and L max is the maximum value of radial deviation, L min is the minimum value of the radial deviation. The deformation data of the inner wall of the tank is analyzed according to the radial deviation value σ and the lateral ovality η.

[0029] The beneficial effects of the present invention are as follows: the present invention solves the problem that, during the deformation detection of some metal cans, only deformation data of a few observation points on a certain radial cross section can be obtained to infer the overall deformation, but accurate local and overall deformation details cannot be obtained, and the accuracy of the detection data needs to be further improved.

[0030] The deformation detection method of a vertical metal tank uses a 3D laser scanner to comprehensively scan the inner wall of the tank body, reconstructs a 3D model after processing the acquired 3D point cloud data, slices the 3D point cloud data, obtains radial deviation values ​​and ellipticity data from the section point cloud data, thereby analyzing the deformation data of the tank body, and can quickly and completely detect the deformed parts of the tank body, thereby improving the accuracy of the detection data. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 It is a schematic diagram of a storage tank body deformation detection assembly of the present invention;

[0032] Figure 2 A schematic diagram of a detection coordinate system of a three-dimensional laser scanner of the present invention;

[0033] Figure 3 The present invention is a flow chart of the method for detecting deformation of a storage tank body.

[0034] In the figure: 1. Storage tank body; 2. Sealing ring groove; 3. External floating roof; 4. Floating plate; 5. 3D laser scanner. DETAILED DESCRIPTION

[0035] In order to make the technical means and objectives of the present invention easy to understand, the present invention is further described below in combination with specific embodiments. A deformation detection device for a vertical metal can, such as Figure 1-Figure 2 As shown, it includes a storage tank body 1, a sealing ring groove 2 is opened inside the storage tank body 1, an external floating roof 3 is slidably connected to the inner side of the storage tank body 1, a floating plate 4 is fixedly installed on the top of the external floating roof 3, and a three-dimensional laser scanner 5 is fixedly installed on the top of the floating plate 4.

[0036] A method for detecting deformation of a vertical metal can, such as Figure 3 As shown, the implementation steps include the following:

[0037] S1. The liquid in the tank body 1 is discharged through the discharge pipe, and the external floating roof 3 and the floating plate 4 drop with the liquid level, and the inner wall of the tank body 1 is scanned by a three-dimensional laser scanner 5 installed on the top of the floating plate 4;

[0038] S2. Construct a coordinate system according to the three-dimensional laser scanner 5 to the target deformation area to obtain the three-dimensional coordinates of the scanned tank body 1;

[0039] S3. Generate three-dimensional point cloud data according to the acquired three-dimensional coordinates, and process the acquired three-dimensional point cloud data;

[0040] S4. Reconstruct a three-dimensional model of the tank body 1 according to the processed three-dimensional point cloud data, and observe the deformation cloud map data of the inner wall of the tank body 1;

[0041] S5. Perform uniform horizontal slicing on the processed three-dimensional point cloud data to obtain several cross-sectional point cloud data of the tank body 1;

[0042] S6. Fit the center coordinates of each section according to the acquired section point cloud data, calculate the center radius of each section, obtain and analyze the radial deviation value and ellipticity data of the target deformation area, and obtain the deformation data of the tank body 1.

[0043] Furthermore, a sealing assembly is provided inside the sealing ring groove 2, and the three-dimensional laser scanner 5 includes a laser transmitter and a laser receiver, which are used to transmit detection laser to the tank body 1 and receive reflected laser signals. The three-dimensional laser scanner 5 signal is connected to a computer microprocessor, and the computer microprocessor data is connected to a data storage device.

[0044] Specifically, the three-dimensional laser scanner 5 is composed of a laser rangefinder and a reflective prism structure. The laser transmitter of the laser rangefinder emits a detection laser to the tank body 1 and the deformed part to be detected. After being received by the laser receiver, the distance is measured by the distance measurement module, and then the scanning control module of the reflective prism can measure the horizontal and vertical angles of the scan.

[0045] Furthermore, a distance measurement module and a scanning control module are provided inside the three-dimensional laser scanner 5. The position of the three-dimensional laser scanner 5 is set as the coordinate origin O, an xyz coordinate system is established according to the coordinate origin O, and a plurality of observation scanning points F are selected in the scanning area of ​​the inner wall of the storage tank body 1. k (k=0, 1, 2, ..., n), calculate the distance from the laser beam of the three-dimensional laser scanner 5 to each observation scanning point F according to the pulse laser ranging algorithm of the distance measurement module k The distance value S k (k=0, 1, 2, ..., n).

[0046] Specifically, the laser rangefinder continuously emits laser and receives the reflected signal of the measured object to measure the distance between the target and the laser rangefinder. The reflective prism is usually composed of a set of horizontal reflectors and a set of vertical reflectors, which are used to guide the laser and make the laser rangefinder rotate at the same angular speed in the horizontal and vertical directions. In this way, the distance between the three-dimensional laser scanner 5 and each observation scanning point F is measured. k The distance value S k When , the horizontal measurement angle θ and the vertical measurement angle ɑ can be obtained.

[0047] Furthermore, the scanning control module includes an angle counter for calculating the angle ɑ between the laser beam and the xOy plane and the angle θ between the projection of the laser beam on the xOy plane and the positive x axis, and calculating the angle of each observation scanning point F k The three-dimensional coordinate value d in the xyz coordinate system k (x k ,y k , z k )(k=0, 1, 2, ..., n), the calculation method is as follows:

[0048] For each observation scanning point F k The three-dimensional coordinate value d k (x k ,y k , zk ) is integrated to obtain the three-dimensional point cloud data of the tank body 1.

[0049] Specifically, the pulse laser ranging algorithm of the distance measurement module is used to calculate the distance from the laser beam of the three-dimensional laser scanner 5 to each observation scanning point F. k The distance value S k Then, according to the angle ɑ between the laser beam and the xOy plane and the angle θ between the projection of the laser beam on the xOy plane and the positive x axis, the angle F of each observation scanning point is calculated. k The three-dimensional coordinate value d in the xyz coordinate system k .

[0050] Furthermore, the three-dimensional point cloud data processing includes the following steps:

[0051] L1. Perform stitching and registration processing on the 3D point cloud data in the point cloud data processing software, and delete the point cloud data that exceeds the scanning setting range;

[0052] L2. Based on the mean curvature flow algorithm, the 3D point cloud data after splicing and registration is subjected to denoising and filtering. The algorithm is as follows:

[0053] Select a vertex H in the 3D point cloud data, estimate the normal vector and curvature of the vertex H based on the geometric information contained in the local neighborhood of the selected vertex H, and move the vertex H to the model surface on the normal vector. The vertex H in the mean curvature direction is estimated as follows:

[0054]

[0055] where k H is the mean curvature, e ij =X j -X i , β j ,ω j for e ij The opposite angle between two adjacent triangles, A is the angle around X i The area of ​​the triangle is obtained, and the vertex H is moved to the surface of the smooth surface at a speed of the average curvature of the local surface in the direction of the normal vector, and a denoising filter is performed to obtain the denoising filtered three-dimensional point cloud data.

[0056] Specifically, Leica Cyclone 6.0 software is used for point cloud stitching. Since the amount of data acquired by the scanner is very large, some of the point cloud data does not work for 3D reconstruction and monitoring analysis. If all the data is used, the error range will be enlarged. At the same time, the laser beam may have a discrete effect during propagation, and the laser receiver will receive many messy light beams. These messy light beams will cause many noise points. The noise points will cause a large amount of redundant data in the measurement results, affecting the accuracy and efficiency of monitoring and analysis. Therefore, it is necessary to perform registration and denoising filtering on each point cloud data collected.

[0057] Furthermore, the 3D point cloud data after denoising and filtering is reconstructed in the 3D modeling software. The steps of 3D reconstruction include point cloud coloring processing, polygon processing, contour line and surface patch processing, grid processing and NURBS surface processing. The processed data file is output to obtain the 3D reconstructed model of the tank body 1, and the deformation cloud map of the 3D reconstructed model is observed.

[0058] Specifically, the point cloud is colored to make it more intuitive, the non-connected items and isolated points outside the tank are removed and then packaged, a flow type is created for the packaged polygon data, and then the holes are filled, the protrusions or redundant parts of the polygons are removed, the surface is polished smooth, the tank model is relaxed, the intersecting areas are repaired, the irregular triangle data are removed, the boundaries of various places are edited, holes are created or fitted, and other operations are performed to obtain a regular polygonal model;

[0059] First, contour processing is performed, including detecting contours, editing contours, detecting curvatures, moving curvature lines, subdividing or extending contours, editing or extending, upgrading or constraining, relaxing contours, and automatically fitting surfaces, etc. Then, surface patch processing is performed, including constructing surface patches, relaxing surface patches, editing surface patches, moving surface patches, moving panels, compressing surface patch layers, repairing surface patches, and drawing surface patch layout diagrams, etc. Then, grid processing is performed, including constructing grids, specifying sharp-angle contours, etc. Then, NURBS surface processing is completed, including fitting surfaces, merging surfaces, deleting surfaces, 3D comparison, etc. Finally, an ideal NURBS surface is obtained, and the file is output, thereby obtaining a three-dimensional reconstructed model of the tank body 1.

[0060] Furthermore, the three-dimensional point cloud data is sliced ​​to obtain several point cloud slices, each of which is 0.6 m apart and 0.001 m thick. The center coordinates of each point cloud slice are fitted according to the least squares method and the center diameter is calculated.

[0061] Specifically, several point cloud slices include standard point cloud slices and point cloud slices of deformed parts. The least squares method is used to fit the center coordinates of each point cloud slice and calculate the center diameter to obtain the corresponding fitting center and fitting diameter, which are used as standard parameters.

[0062] Furthermore, according to the center coordinates and center diameter of each point cloud slice, the fitting center and fitting diameter are obtained by fitting, which are used as standard parameters, and the radial deviation value is calculated as follows:

[0063] σ=LS, where σ is the radial deviation value of the point cloud slice, L is the distance from the point cloud data of the remaining point cloud slices except the standard point cloud slice to the center of the circle, and S is the distance from the point cloud data of the standard point cloud slice to the center of the circle;

[0064] The ellipticity is calculated as follows:

[0065] Where η is the lateral ellipticity of the point cloud slice, L is the average diameter of the point cloud slice, and L max is the maximum value of radial deviation, L min is the minimum value of the radial deviation. The deformation data of the inner wall of the tank body 1 is analyzed according to the radial deviation value σ and the lateral ovality η.

[0066] Specifically, the radial deviation value and ovality data are obtained for analysis. If σ is a positive value, the inner wall detection portion of the tank body 1 bulges outward and presents an outward convex trend. If σ is a negative value, the inner wall detection portion of the tank body 1 is sunken inward and presents an inward concave trend.

[0067] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical solutions and concepts of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A deformation detection device for a vertical metal can, characterized in that: The invention comprises a tank body, wherein a sealing ring groove is provided inside the tank body, a sealing assembly is arranged inside the sealing ring groove, an external floating roof is slidably connected to the inner side of the tank body, a floating plate is arranged on the top of the external floating roof, and a three-dimensional laser scanner is arranged on the top of the floating plate.

2. A method for detecting deformation of a vertical metal can, characterized in that: The steps include: S1. The liquid in the tank is discharged through the discharge pipe. The external floating roof and the floating plate descend with the liquid level. The inner wall of the tank is scanned by a three-dimensional laser scanner installed on the top of the floating plate. S2, constructing a coordinate system toward the target deformation area according to the three-dimensional laser scanner, and obtaining the three-dimensional coordinates of the scanned storage tank body; S3, generating three-dimensional point cloud data according to the acquired three-dimensional coordinates, and processing the acquired three-dimensional point cloud data; S4, reconstructing a three-dimensional model of the storage tank body according to the processed three-dimensional point cloud data, and observing the deformation cloud map data of the inner wall of the storage tank body; S5, performing uniform horizontal slicing on the processed three-dimensional point cloud data to obtain several sections of point cloud data of the tank body; S6. Fit the center coordinates of each section according to the obtained section point cloud data, calculate the center radius of each section, obtain and analyze the radial deviation value and ellipticity data of the target deformation area, and obtain the deformation data of the tank body.

3. The method for detecting deformation of a vertical metal can according to claim 2, characterized in that: In step S1, the three-dimensional laser scanner includes a laser transmitter and a laser receiver, which transmit a detection laser to the tank body and receive a reflected laser signal. The three-dimensional laser scanner signal is connected to a computer microprocessor, and the computer microprocessor data is connected to a data storage device. The laser transmitter of the laser rangefinder is used to transmit a detection laser to the tank body and the deformed part to be detected. After being received by the laser receiver, the distance is measured by the distance measurement module, and then the scanning control module of the reflecting prism can measure the horizontal direction angle and the vertical direction angle of the scan.

4. The method for detecting deformation of a vertical metal can according to claim 2, characterized in that: In step S2, the three-dimensional laser scanner is provided with a distance measurement module and a scanning control module, the position of the three-dimensional laser scanner is set to the coordinate origin O, an xyz coordinate system is established according to the coordinate origin O, and a plurality of observation scanning points F are selected in the scanning area of ​​the inner wall of the tank. k (k = 0, 1, 2, ..., n), according to the pulse laser ranging algorithm of the distance measurement module, the distance from the laser beam of the 3D laser scanner to each observation scanning point F is calculated. k The distance value S k (k=0, 1, 2, ..., n).

5. The method for detecting deformation of a vertical metal can according to claim 2, characterized in that: In step S3, the scanning control module includes an angle counter for calculating the angle ɑ between the laser beam and the xOy plane and the angle θ between the projection of the laser beam on the xOy plane and the positive x axis, and calculating the angle of each observation scanning point F k The three-dimensional coordinate value d in the xyz coordinate system k (x k ,y k , z k )(k=0, 1, 2, ..., n), the calculation method is as follows: For each observation scanning point F k The three-dimensional coordinate value d k (x k ,y k , z k ) is integrated to obtain the three-dimensional point cloud data of the tank body.

6. The method for detecting deformation of a vertical metal can according to claim 2, characterized in that: In step S3, processing the three-dimensional point cloud data includes the following steps: S31. Perform stitching and registration processing on the three-dimensional point cloud data in the point cloud data processing software, and delete the point cloud data that exceeds the scanning setting range; S32. De-noising and filtering are performed on the three-dimensional point cloud data after the splicing and registration processing based on the mean curvature flow algorithm. The algorithm is as follows: Select a vertex H in the 3D point cloud data, estimate the normal vector and curvature of the vertex H based on the geometric information contained in the local neighborhood of the selected vertex H, and move the vertex H to the model surface on the normal vector. The vertex H in the mean curvature direction is estimated as follows: where k H is the mean curvature, e ij =X j -X i , β j ,ω j for e ij The opposite angle between two adjacent triangles, A is the angle around X i The area of ​​the triangle is obtained, and the vertex H is moved to the surface of the smooth surface at a speed of the average curvature of the local surface in the direction of the normal vector, and a denoising filter is performed to obtain the denoising filtered three-dimensional point cloud data.

7. The method for detecting deformation of a vertical metal can according to claim 2, characterized in that: In step S4, the 3D point cloud data after denoising and filtering is 3D reconstructed in the 3D modeling software, and the steps of 3D reconstruction include point cloud coloring processing, polygon processing, contour line and surface patch processing, grid processing and NURBS surface processing. The processed data file is output to obtain a 3D reconstructed model of the tank body, and the deformation cloud map of the 3D reconstructed model is observed.

8. The method for detecting deformation of a vertical metal can according to claim 2, characterized in that: In step S5, the three-dimensional point cloud data is sliced ​​to obtain a plurality of point cloud slices, each of which is spaced 0.6 m apart and 0.001 m thick. The center coordinates of each point cloud slice are fitted and the center diameter is calculated using the least squares method.

9. The method for detecting deformation of a vertical metal can according to claim 2, characterized in that: In step S6, according to the center coordinates and center diameter of each point cloud slice, a fitting center and a fitting diameter are obtained by fitting, which are used as standard parameters, and the radial deviation value is calculated as follows: σ=LS, where σ is the radial deviation value of the point cloud slice, L is the distance from the point cloud data of the remaining point cloud slices except the standard point cloud slice to the center of the circle, and S is the distance from the point cloud data of the standard point cloud slice to the center of the circle; The ellipticity is calculated as follows: Where η is the lateral ellipticity of the point cloud slice, L is the average diameter of the point cloud slice, and L max is the maximum value of radial deviation, L min is the minimum value of the radial deviation. The deformation data of the inner wall of the tank is analyzed according to the radial deviation value σ and the lateral ovality η.