Method, device and storage medium for processing a round tube intersecting line weld

Abstract

This application discloses a method, apparatus, and storage medium for processing intersecting weld seams of circular pipes. The method acquires three-dimensional point cloud data of a first and a second circular pipe using a pre-scanning mode. The acquired three-dimensional point cloud data is preprocessed, and the cylindrical surface equations of the first and second circular pipes are fitted using this data. Then, a projection method is used to project the bottom circle of the second circular pipe cylinder onto the axis vector of the first circular pipe to obtain the intersecting weld seam data. The intersecting weld seam processing method provided in this application solves the problem in the prior art where, when obtaining the intersecting weld seam trajectory by teaching key points on the intersecting weld seam, the accuracy of the generated weld seam trajectory is low due to human error, thus affecting the welding formation effect and welding quality. This method only requires setting the scanning range to quickly and accurately calculate the intersecting weld seam data and allows for corresponding welding process adjustments based on actual conditions to improve the welding formation effect.

Description

Technical Field

[0001] This application relates to the field of welding technology, and in particular to methods, apparatus and storage media for treating intersecting welds of circular pipes. Background Technology

[0002] Intersection welds of circular pipes are quite common in the welding field, appearing in various types of circular pipe connection scenarios such as spray water pipes, oil pipes, and natural gas pipes. Figure 1 For the intersection line formed between two cylindrical surfaces, the welding of the intersection line weld of circular pipes is currently divided into two main categories: manual welding and robotic welding. However, manual welding of the intersection line of circular pipes is prone to problems such as poor welding quality, uneven welding, and unsightly welds. Robotic welding of the intersection line weld of circular pipes mainly involves teaching a few key points, which not only fails to guarantee consistent accuracy but also has low production efficiency. For example, the Chinese patent application filed on January 25, 2011, with application number 201110029075.9, adopts a teaching method to use a robot to weld the intersection line weld. This application involves operating the robot to teach at least four key points on the intersection line weld, then constructing a local coordinate system, using calibration to complete the transformation from the robot base coordinate system to the workpiece coordinate system, and finally generating the actual weld trajectory through an intersection line weld trajectory interpolation algorithm to achieve welding.

[0003] This method can be used to weld intersecting welds, but the teaching process itself introduces human error. Furthermore, intersecting welds may have inconsistent bevel or gap sizes, leading to lower accuracy in the generated weld trajectory and consequently affecting the weld formation and quality. Additionally, identifying welds through teaching also reduces work efficiency and affects the continuity of the process. Summary of the Invention

[0004] Therefore, the method for processing the intersection weld of circular pipes provided in this application can quickly and accurately calculate the data of the intersection weld, and can be used in all scenarios of intersection welds formed by the intersection of two circular pipes.

[0005] To achieve the above objectives, the first aspect of this application adopts the following technical solution: a method for processing the weld seam at the intersection of circular pipes, which involves obtaining the three-dimensional point cloud data of the circular pipes based on a pre-scanning mode, fitting the cylindrical surface equations of the first and second circular pipes, and obtaining the intersection line data based on the fitted cylindrical surface equations, including the following steps:

[0006] Project the bottom circle of the second cylindrical tube onto the axis vector N of the first cylindrical tube. m The coordinates P of the two points where the bottom circle intersects the axis of the first circular tube are obtained. L0 (x L0 ,y L0 ,z L0 ) and PL1 (x L1 , y L1 , z L1 ), where the distance L between P L0 and P L1 is regarded as the length of the intersection curve;

[0007] At equal step lengths step between P L0 and P L1 , take the center point P L of the bottom circle of the first circular tube cylinder surface. The center point P L of the bottom circle of the first circular tube cylinder surface can be obtained by formula 7;

[0008] P L (x L , y L , z L ) = P L0 (x L0 , y L0 , z L0 ) + step * N m (n mx , n my , n mz ) (Formula 7)

[0009] where step ∈ [0, L];

[0010] Arbitrarily take a point P m on the cylinder surface of the first circular tube, obtain the coordinates of point P m , and substitute point P m into the cylinder surface equation of the second circular tube;

[0011] If P m also satisfies the cylinder surface equation of the second circular tube, then compare the distance d1 between P m and the centroid P bv of the valid three-dimensional point cloud data of the second circular tube with the distance d2 between P L and the centroid P bv of the valid three-dimensional point cloud data of the second circular tube. If d1 < d2, then P m is the true intersection curve data;

[0012] Preferably, the steps to obtain P m are as follows:

[0013] According to the vector cross product formula, cross multiply the axis vector N m of the first circular tube with the axis vector N b of the second circular tube to obtain the vector V a , and then cross multiply the obtained vector V a with the axis vector N mCross product yields vector V b :

[0014] V a (n ax ,n ay ,n az ) = N m (n mx ,n my ,n mz )×N b (n bx ,n by ,n bz ) (Formula 8)

[0015] V b (n bx ,n by ,n bz ) = N m (n mx ,n my ,n mz )×V a (n ax ,n ay ,n az ) (Formula 9)

[0016] Let P m (x m ,y m ,z m Let V be a point on the cylindrical surface of the first circular tube. a sum vector V b Consider them as two perpendicular vectors on the base circle of the first cylindrical surface. Therefore, point P... m The coordinates can be obtained using formula 10:

[0017] P m =P L +R m *(cosθ*V a +sinθ*V b ) (Formula 10)

[0018] Where θ∈[0, 2π], R m Let be the radius of the bottom circle of the first cylindrical tube.

[0019] Preferably, after acquiring the intersection line data, the intersection line data undergoes welding process processing, including:

[0020] Expand the intersection line data outward; sample the intersection line data at equal intervals; set the initial point of the equally sampled intersection line data on one side of the second circular tube.

[0021] Preferably, before obtaining the intersection line data, it is necessary to fit the cylindrical surface equations of the first and second circular tubes, specifically including the following steps:

[0022] S1. Scan the first circular tube and the second circular tube to obtain three-dimensional point cloud data of the surfaces of the first circular tube and the second circular tube;

[0023] S2. The three-dimensional point cloud data obtained in step S1 are preprocessed to remove outliers and downsample the three-dimensional point cloud data, including statistical filtering and voxel filtering.

[0024] S3. Fit the cylindrical surface equation of the first circular tube based on the preprocessed three-dimensional point cloud data in step S2.

[0025] S4. Filter out the three-dimensional point cloud data of the first circular tube surface that exists in the three-dimensional point cloud data of the second circular tube;

[0026] S5. Fit the equation of the cylindrical surface of the second circular tube based on the effective three-dimensional point cloud data of the second circular tube obtained in step S4, and output a point P on the axis of the second circular tube. b0 (x b0 ,y b0 ,z b0 ), axis vector N b (n bx ,n by ,n bz ) and the radius R of the bottom circle b .

[0027] Preferably, in step S3, the specific steps for fitting the equation of the first circular tube cylindrical surface are as follows:

[0028] S301. For each point in the three-dimensional point cloud data of the first circular tube, take several points in its neighborhood, fit a plane with the point and several points in its neighborhood, calculate the normal vector of the plane, and normalize the obtained plane normal vector. The obtained unit normal vector is the unit normal vector of the point.

[0029] S302. Randomly select N unit normal vectors from three-dimensional point cloud data, treat them as coordinate points, and fit a plane Planei to solve for the normal vector N of this plane. i (n ix ,n iy ,n iz ); where vector N i (n ix ,n iy ,n iz ) is the axis vector of the cylinder, and i represents the i-th iteration;

[0030] S303, Project all 3D point cloud data onto the fitting plane. i The base radius R of the fitted circle is obtained by fitting a circle Circei based on all the projection points. i and the center P i (x i0 ,y i0 ,z i0 ); where the radius R i Let P be the radius of the base circle of the cylinder. i (x i0 ,y i0 ,z i0 Let be a point on the axis of the cylinder, and let i represent the i-th iteration;

[0031] S304, Point P i (x i0 ,y i0 ,z i0 ), vector N i (n ix ,n iy ,n iz ) and radius R i Substituting into Formula 11, we obtain the equation of the cylindrical surface of the first circular tube:

[0032] (x-x0) 2 +(y-y0) 2 +(z-z0) 2 -[n x (x-x0)+n y (y-y0)+n z (z-z0)] 2 =R 2

[0033] (Formula 11)

[0034] S305. Calculate the error between each point in the 3D point cloud data and the fitted cylinder according to Formula 12. If error ErrorThresh (less than the set threshold) i If the point is an interior point, then it is considered an interior point.

[0035]

[0036] S306. Calculate the ratio (Ratio) of the number of interior points in the total number of points in the 3D point cloud data. i If Ratio i Greater than the set threshold RatioThresh i If the fit is successful, then point P is considered to be a successful fit. i (x i0,y i0 ,z i0 ), vector N i (n ix ,n iy ,n iz ) and radius R i Return to point P on the axis of the first circular tube. m0 (x m0 ,y m0 ,z m0 ), axis vector N m (n mx ,n my ,n mz ) and the radius R of the bottom circle m Otherwise, restart the iteration from step S302 until the maximum number of iterations is reached. The maximum number of iterations is the pre-set maximum number of cyclic fitting iterations. Before reaching the maximum number of iterations, if Ratio i Greater than the set threshold RatioThresh i If the fit is successful, exit the fitting program. If the maximum number of iterations is reached, Ratio... i Still less than the set threshold RatioThresh i If the result is negative, then the fitting is considered to have failed.

[0037] Preferably, in step S4, the error Error between each preprocessed second circular tube 3D point cloud data and the cylindrical surface of the first circular tube is calculated according to Formula 12. If the error Error is greater than the set error threshold Errorthresh, it is considered a valid second circular tube 3D point cloud data and is retained; otherwise, the point data is deleted from the dataset. The centroid P of all valid second circular tube 3D point cloud data is then calculated. bv (x bv ,y bv ,z bv ).

[0038] More preferably, the intersection line data can also be obtained using the following method:

[0039] A local coordinate system is established at the intersection of the axis of the first circular tube and the axis of the second circular tube;

[0040] Transform the cylindrical surface equations of the first and second circular tubes into this local coordinate system to obtain the cylindrical surface equations of the first and second circular tubes in the local coordinate system, respectively.

[0041] By simultaneously solving the equations of the first and second cylindrical surfaces of the circular tube in the local coordinate system, the equation of the intersection line is obtained, thus yielding the intersection line data.

[0042] In a second aspect, this application provides an apparatus for treating intersecting weld seams of circular pipes, including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the method described in the first aspect of this application.

[0043] In a third aspect, this application provides a computer-readable storage medium having a computer program / instructions stored thereon, which, when executed by a processor, implement the steps of the method described in the first aspect of this application.

[0044] Beneficial effects:

[0045] (1) The data of the intersection weld can be calculated quickly and accurately simply by setting the scanning range;

[0046] (2) It can make corresponding adjustments to the welding process according to the actual situation in order to improve the welding forming effect;

[0047] (3) It can be used in all scenarios of intersection welds formed by the intersection of two circular pipes. It can not only adapt to the intersection welds formed by the perpendicular intersection of two circular pipes, but also adapt to the situation of the two circular pipes being eccentrically intersecting.

[0048] (4) The calculated weld data has a small error, which can be controlled within 1mm, meeting the requirements for intersecting line welding. Attached Figure Description

[0049] Figure 1 This is a schematic diagram of the intersection line formed between two cylindrical surfaces;

[0050] Figure 2 This is a photograph of the intersection line of the spray pipes.

[0051] Figure 3 This is a flowchart of the method for treating intersecting weld seams as described in the embodiments of this application;

[0052] Figure 4 This is a schematic diagram of the three-dimensional point cloud data of the surface of the first circular tube as described in the embodiments of this application;

[0053] Figure 5 This is a schematic diagram of the three-dimensional point cloud data of the surface of the second circular tube as described in the embodiments of this application;

[0054] Figure 6 This is a schematic diagram of the filtered and effective three-dimensional point cloud data of the second circular tube as described in the embodiments of this application;

[0055] Figure 7 This is a schematic diagram of the intersection line data between the first circular tube and the second circular tube as described in the embodiments of this application;

[0056] Figure 8This is a schematic diagram showing the fitting results of the cylindrical surfaces of the first and second circular tubes described in the embodiments of this application;

[0057] Figure 9 This is a schematic diagram of the vector information and coordinate point information of the fitting process described in the embodiments of this application;

[0058] Figure 10 This application provides a schematic diagram of the final intersection line data after welding process, as described in the embodiments of this application. Detailed Implementation

[0059] The technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, it should be noted that in the description of the embodiments of this invention, the terms "first" and "second" are only used to distinguish descriptions and should not be construed as indicating or implying their degree of importance.

[0060] Example 1

[0061] The method for processing intersecting weld seams of circular pipes provided in this application acquires three-dimensional point cloud data of the circular pipe based on a pre-scanning mode. This embodiment uses the processing of intersecting weld seams of spray pipes as an example for detailed explanation. Figure 2 This is a physical diagram of the intersection line of the sprinkler pipes. Pipe 1 in the diagram is the first pipe, and pipe 2 is the second pipe. The treatment method is as follows: Figure 3 As shown, it includes the following steps:

[0062] S1. Use a 3D line structured light camera to scan the surfaces of the first and second circular tubes respectively, and obtain 3D point cloud data of the surface portions of the first and second circular tubes. Figure 4 and Figure 5 These are three-dimensional point cloud data images of the surfaces of the first and second circular tubes, obtained through scanning.

[0063] In this embodiment, acquiring the three-dimensional point cloud data of the circular tube using a 3D line structured light camera is only one preferred method. These points, which represent the positions in three-dimensional space, are usually obtained through laser scanning technology (such as LiDAR) or structured light scanning technology (line structured light camera or surface structured light camera). Therefore, this method is not limited to using a 3D line structured light camera. It is only necessary to be able to acquire the three-dimensional point cloud data of the surfaces of the first and second circular tubes. The 3D line structured light camera is just an example.

[0064] S2. The three-dimensional point cloud data obtained in step S1 are preprocessed to remove outliers and downsample the three-dimensional point cloud data, including statistical filtering and voxel filtering.

[0065] Because direct scanning of an object yields discrete 3D point cloud data, this raw data contains noise and redundancy. Therefore, preprocessing is necessary to reduce data uncertainty. First, statistical filtering algorithms are used to filter the acquired 3D point cloud data to remove outliers. The specific steps are as follows:

[0066] S201. Set the number of neighboring points K for calculating the average distance of each data point, that is, only calculate the parameter of the K nearest neighboring points of each data point.

[0067] S202. Calculate the mean distance MeanDist and standard deviation σ from each data point to all K-neighborhood points;

[0068] S203. Set the distance threshold DistThresh, calculated as shown in Formula 13, where k1 and k2 are coefficients:

[0069] DisThresh=k1*MeanDist+k2*σ (Formula 13)

[0070] S204. Compare the average distance MeanDist and the distance threshold DistThresh for each data point. If the condition MeanDist≥DistThresh is met, the data point is considered an outlier and is deleted from the dataset, thus completing the statistical filtering process.

[0071] Then, the 3D point cloud data is filtered using a voxel filtering algorithm to reduce the amount of point cloud data while preserving the shape features of the point cloud. First, the point cloud data is divided into several cubic units called voxels. Then, the average coordinates of all data points in each voxel are used as the coordinates of the center point of that voxel, thereby realizing the filtering of the point cloud data.

[0072] S3. Based on the preprocessed 3D point cloud data from step S2, fit the cylindrical surface equation of the first circular tube, specifically including:

[0073] S301. For each point in the three-dimensional point cloud data of the first circular tube, take several points in its neighborhood, fit a plane with the point and several points in its neighborhood, calculate the normal vector of the plane, and normalize the obtained plane normal vector. The obtained unit normal vector is the unit normal vector of the point.

[0074] S302. Randomly select N unit normal vectors from three-dimensional point cloud data, treat them as coordinate points, and fit a plane Planei to solve for the normal vector N of this plane. i (n ix ,n iy ,n iz ); where vector Ni (n ix ,n iy ,n iz ) is the axis vector of the cylinder, and i represents the i-th iteration;

[0075] S303, Project all 3D point cloud data onto the fitting plane. i The base radius R of the fitted circle is obtained by fitting a circle Circei based on all the projection points. i and the center P i (x i0 ,y i0 ,z i0 ); where the radius R i Let P be the radius of the base circle of the cylinder. i (x i0 ,y i0 ,z i0 Let be a point on the axis of the cylinder, and let i represent the i-th iteration;

[0076] S304, Point P i (x i0 ,y i0 ,z i0 ), vector N i (n ix ,n iy ,n iz ) and radius R i Substituting into formula 14, we obtain the equation of the cylindrical surface of the first circular tube:

[0077] (x-x0) 2 +(y-y0) 2 +(z-z0) 2 -[n x (x-x0)+n y (y-y0)+n z (z-z0)] 2 =R 2

[0078] (Formula 14)

[0079] S305. Calculate the error between each point in the 3D point cloud data and the fitted cylinder according to Formula 15. If error ErrorThresh (less than the set threshold) i If the point is an interior point, then it is considered an interior point.

[0080]

[0081] S306. Calculate the ratio (Ratio) of the number of interior points in the total number of points in the 3D point cloud data.i If Ratio i Greater than the set threshold RatioThresh i If the fit is successful, then point P is considered to be a successful fit. i (x i0 ,y i0 ,z i0 ), vector N i (n ix ,n iy ,n iz ) and radius R i Return to point P on the axis of the first circular tube. m0 (x m0 ,y m0 ,z m0 ), axis vector N m (n mx ,n my ,n mz ) and the radius R of the bottom circle m Otherwise, restart the iteration from step S302 until the maximum number of iterations is reached. The maximum number of iterations is the pre-set maximum number of cyclic fitting iterations. Before reaching the maximum number of iterations, if Ratio i Greater than the set threshold RatioThresh i If the fit is successful, the fitting process is considered successful and the fitting program is exited; if the maximum number of iterations is reached, Ratio is adjusted. i Still less than the set threshold RatioThresh i If the result is negative, then the fitting is considered to have failed.

[0082] S4. Filter out the three-dimensional point cloud data of the first circular tube surface that exists in the three-dimensional point cloud data of the second circular tube;

[0083] like Figure 5 As shown, the 3D point cloud data of the second circular tube may contain some 3D point cloud data of the surface of the first circular tube. This data will affect the fitting accuracy and stability of the cylindrical surface of the second circular tube. Therefore, it is necessary to filter out the first circular tube data in the second circular tube data first. According to Formula 15, the error Error between each preprocessed 3D point cloud data of the second circular tube and the cylindrical surface of the first circular tube is calculated. If the error Error is greater than the set error threshold Errorthresh, it is considered to be valid 3D point cloud data of the second circular tube and is retained. Otherwise, the point data is deleted from the dataset. The filtered valid 3D point cloud data of the second circular tube is as follows: Figure 6 As shown; and calculate the centroid P of all valid second circular tube 3D point cloud data. bv (x bv ,y bv ,z bv ).

[0084] S5. Fit the equation of the cylindrical surface of the second circular tube based on the effective three-dimensional point cloud data of the second circular tube obtained in step S4, and output a point P on the axis of the second circular tube. b0 (x b0 ,y b0 ,z b0 ), axis vector N b (n bx ,n by ,n bz ) and the radius R of the bottom circle b .

[0085] S6. Obtain the intersection line data between the first circular tube and the second circular tube;

[0086] Its intersection line data diagram is as follows Figure 7 As shown, obtaining the intersection line data specifically includes the following steps:

[0087] S601, the cylindrical surface results of the fitted first and second circular tubes are as follows: Figure 8 As shown, the bottom circle of the second cylindrical tube is projected onto the axis vector N of the first cylindrical tube. m Above, such as Figure 9 As shown, the coordinates P of the two points where the bottom circle intersects the axis of the first circular tube are obtained. L0 (x L0 ,y L0 ,z L0 ) and P L1 (x L1 ,y L1 ,z L1 ), where P L0 With P L1 The distance L between them is considered as the length of the intersection line;

[0088] S602, in P L0 To P L1 The step size is equal, and the center point P of the bottom circle of the first cylindrical tube is taken. L The center point P of the bottom circle of the first cylindrical tube L It can be obtained from formula 16;

[0089] P L (x L ,y L ,z L ) = P L0 (x L0 ,y L0 ,z L0 )+step*N m (n mx ,n my ,n mz ) (Formula 16)

[0090] Where step∈[0,L], P L The value of is determined by the coefficient step, and the value of step is between [0, L]. Therefore, point P in the figure... L This is just one possible point;

[0091] S603. Take any point P on the cylindrical surface of the first circular tube. m Get point P m Coordinates, and point P m Substitute the equation of the cylindrical surface of the second circular tube;

[0092] Specifically, according to the vector cross product formula, the first circular tube axis vector N is... m Vector N of the second circular tube axis b Performing the cross product yields the vector V. a and the resulting vector V a Then, with the first circular tube axis vector N m Cross product yields vector V b :

[0093] V a (n ax ,n ay ,n az ) = N m (n mx ,n my ,n mz )×N b (n bx ,n by ,n bz ) (Formula 17)

[0094] V b (n bx ,n by ,n bz ) = N m (n mx ,n my ,n mz )×V a (n ax ,n ay ,n az ) (Formula 18)

[0095] Let P m (x m ,y m ,z m Let V be a point on the cylindrical surface of the first circular tube. a sum vector V b Consider them as two perpendicular vectors on the bottom circle of the first cylindrical surface. Therefore, point P can be obtained according to formula 19. m coordinate:

[0096] P m = P L + R m *(cosθ * V a + sinθ * V b ) (Equation 19)

[0097] where θ ∈ [0, 2π], R m is the radius of the bottom circle of the first circular tube cylinder surface;

[0098] S604. If P m also satisfies the cylindrical surface equation of the second circular tube, then compare the distance d1 between P m and the centroid P bv of the valid three-dimensional point cloud data of the second circular tube, and the distance d2 between P L and the centroid P bv of the valid three-dimensional point cloud data of the second circular tube. Because there are two intersection lines, upper and lower, when the first circular tube intersects with the second circular tube, but in this embodiment, the required intersection line is the upper intersection line, that is, the intersection line closer to the centroid of the point cloud data of the second circular tube. Therefore, if the condition d1 < d2 is satisfied, then P m is the true intersection line data.

[0099] S7. After obtaining the intersection line data, perform welding process treatment on the intersection line data, including:

[0100] Considering that there may be a little gap in the intersection line formed between the two circular tubes, the intersection line data can be expanded according to the actual situation;

[0101] Considering the problem of excessive robot welding posture for the intersection line, sample the intersection line data at equal intervals, and the sampling interval can be set according to the actual situation;

[0102] To prevent problems caused by large changes in scanning and welding postures, set the initial point of the equally spaced sampled intersection line data on one side of the second circular tube;

[0103] Moreover, considering the aesthetics and sealing of welding, the initial point and the end point should not only be set to coincide, but also overlap a part. For example, if arranged in a clockwise direction from the initial point to the end point, then the actual desired end point should be at a certain distance in the clockwise direction from the initial point position. This distance is called the overlap length, and the length can be set according to the actual situation. After the welding process treatment, the intersection line data as shown in Figure 10 can be obtained.

[0104] Embodiment 2

[0105] The intersection line data can also be obtained using the following method:

[0106] A local coordinate system is established at the intersection of the axis of the first circular tube and the axis of the second circular tube;

[0107] Transform the cylindrical surface equations of the first and second circular tubes into this local coordinate system to obtain the cylindrical surface equations of the first and second circular tubes in the local coordinate system, respectively.

[0108] By simultaneously solving the equations of the first and second cylindrical surfaces of the circular tube in the local coordinate system, the equation of the intersection line is obtained, thus yielding the intersection line data.

[0109] Example 3

[0110] Another embodiment of this application proposes a device for treating intersecting welds of circular pipes. This device is used to implement the above-described method for treating intersecting welds of circular pipes. The device includes a memory, a processor, and a computer program stored in the memory. The processor executes the computer program to implement the steps of any of the methods described in Embodiment 1 of this application.

[0111] The memory is used to store computer programs, and the processor is used to process the computer programs, so that the device can implement the above-mentioned method for processing the intersecting weld seam of the circular tube to obtain the intersecting line data and subsequently achieve precise welding.

[0112] Example 4

[0113] Another embodiment of this application provides a computer-readable storage medium having a computer program / instructions stored thereon, which, when executed by a processor, implements the steps of any of the methods described in Embodiment 1 of this application.

[0114] The embodiments described in this application are merely some, not all, of the embodiments described herein. All other embodiments obtained by those skilled in the art based on the embodiments described herein without inventive effort are within the scope of protection of this application.

Claims

1. A method for treating the weld seam at the intersection of circular pipes, comprising: acquiring three-dimensional point cloud data of the circular pipes using a pre-scanning mode; fitting the cylindrical surface equations of a first and a second circular pipe; and obtaining the intersection line data based on the fitted cylindrical surface equations; characterized in that... Includes the following steps: Project the bottom circle of the second cylindrical tube onto the axis vector of the first cylindrical tube. N m The coordinates of the two points where the bottom circle intersects the axis of the first circular tube are obtained. and ,in and Distance between L Considered as the length of the intersection line; exist arrive equal step size step Take the center point of the bottom circle of the first cylindrical tube. The center point of the bottom circle of the first cylindrical tube It can be obtained from Formula 1: Official 1 in, step [ 0,L ]; Take any point on the cylindrical surface of the first circular tube. P m Acquisition Points P m Coordinates, and the point P m Substitute the equation of the cylindrical surface of the second circular tube; like P m It also satisfies the equation of the cylindrical surface of the second circular tube, then compare P m To the centroid of the effective second circular tube 3D point cloud data P bv Distance between d 1 and P L To the centroid of the effective second circular tube 3D point cloud data P bv Distance between d 2, if d 1< d 2, then P m For true intersection line data; P m The steps to obtain it are as follows: According to the cross product formula, the vector of the first circular tube axis is... N m Vector of the second circular tube axis N b Perform the cross product to obtain a vector. V a and the resulting vector V a Then, the vector of the first circular tube axis. N m Cross product yields a vector. V b : Official 2 Official 3 set up Let be a point on the cylindrical surface of the first circular tube, and let the vector be... V a sum vector V b Consider them as two perpendicular vectors on the base circle of the first cylindrical surface; therefore, point... P m The coordinates can be obtained using formula 4: Official 4 in, , R m Let be the radius of the bottom circle of the first cylindrical tube.

2. The method for treating the intersecting weld of a circular pipe according to claim 1, characterized in that: After obtaining the intersection line data, the intersection line data is processed using welding techniques, including: Expand the intersection line data outward; sample the intersection line data at equal intervals; set the initial point of the equally sampled intersection line data on one side of the second circular tube.

3. The method for treating the intersecting weld of a circular pipe according to claim 1, characterized in that, Before obtaining the intersection line data, it is necessary to fit the cylindrical surface equations of the first and second circular tubes, which includes the following steps: S1. Scan the first circular tube and the second circular tube to obtain three-dimensional point cloud data of the surfaces of the first circular tube and the second circular tube; S2. The three-dimensional point cloud data obtained in step S1 are preprocessed to remove outliers and downsample the three-dimensional point cloud data, including statistical filtering and voxel filtering. S3. Fit the cylindrical surface equation of the first circular tube based on the preprocessed three-dimensional point cloud data in step S2. S4. Filter out the three-dimensional point cloud data of the first circular tube surface that exists in the three-dimensional point cloud data of the second circular tube; S5. Fit the equation of the cylindrical surface of the second circular tube based on the effective three-dimensional point cloud data of the second circular tube obtained in step S4, and output a point on the axis of the second circular tube. , axis vector and the radius of the bottom circle .

4. The method for treating the intersecting weld of a circular pipe according to claim 3, characterized in that: In step S3, the specific steps for fitting the equation of the first cylindrical surface of the tube are as follows: S301. For each point in the three-dimensional point cloud data of the first circular tube, take several points in its neighborhood, fit a plane with the point and several points in its neighborhood, calculate the normal vector of the plane, and normalize the obtained plane normal vector. The obtained unit normal vector is the unit normal vector of the point. S302, Random selection N The unit normal vector of a 3D point cloud data is treated as a coordinate point and fitted to a plane. Plane i Find the normal vector of the plane. ; where vector The axis vector of the cylinder. i Indicates the first i The next iteration; S303, Project all 3D point cloud data onto the fitting plane. Plane i Above, and fit a circle based on all projection points. Circle i Find the fitted circle Circle i base radius and the center Among them, radius Let the radius of the base circle of the cylinder be point. Let be a point on the axis of the cylinder. i Indicates the first i The next iteration; S304, Point ,vector and radius Substituting into Formula 5, we obtain the equation of the cylindrical surface of the first circular tube: Official 5 S305. Calculate the error between each point in the 3D point cloud data and the fitted cylinder according to Formula 6. If the error Less than the set threshold ErrorThresh i If the point is an interior point, then it is considered an interior point. Official 6 S306. Calculate the proportion of interior points in the total number of interior points in the entire 3D point cloud data. Ratio i ,like Ratio i Greater than the set threshold RatioThresh i If the fit is successful, the points are set to... ,vector and radius Return to a point on the axis of the first circular tube respectively. , axis vector and the radius of the bottom circle Otherwise, start iterating again from step S302 until the maximum number of iterations is reached.

5. The method for treating the intersecting weld of a circular pipe according to claim 4, characterized in that: In step S4, the error between the preprocessed 3D point cloud data of each second circular tube and the cylindrical surface of the first circular tube is calculated according to formula 6. Error If the error Error Greater than the set error threshold Errorthresh If a point is found to be valid, it is considered valid 3D point cloud data for the second circular tube and is retained; otherwise, the point data is deleted from the dataset. The centroid of all valid 3D point cloud data for the second circular tube is then calculated. .

6. A device for treating intersecting weld seams of circular pipes, comprising a memory, a processor, and a computer program stored in the memory, characterized in that: The processor executes the computer program to implement the steps of the method according to any one of claims 1-5.

7. A computer-readable storage medium having a computer program / instructions stored thereon, characterized in that: When the computer program / instructions are executed by the processor, they implement the steps of the method described in any one of claims 1-5.

Images