Pipeline inspection path planning method

Abstract

The present invention discloses a pipeline detection path planning method, comprising: obtaining original three-dimensional point cloud data of the pipeline shape and preprocessing the three-dimensional point cloud data; finding the boundary points of the pipeline and establishing a workpiece coordinate system based on the preprocessed three-dimensional point cloud data; obtaining the pipeline section position and search direction, and calculating various section parameters; search exit judgment; section correction judgment; correcting the section position and search direction; calculating section parameters and moving in the search direction with an appropriate step distance to find a new section; and establishing path node postures. The present invention can improve the adaptability of pipeline detection and significantly reduce the workload of relevant personnel for pipeline detection.

Description

Technical Field

[0001] The present invention relates to the technical field of industrial robot intelligent flaw detection, and in particular to a pipeline detection path planning method. Background Art

[0002] Pipeline defect detection, using ultrasonic detection technology, captures workpiece defect images through scanning with an ultrasonic flaw detector and analyzes and evaluates pipeline quality. This is a critical step in pipeline manufacturing and processing, and it represents a high-precision, high-efficiency flaw detection technology. With the transformation and upgrading of the manufacturing industry, there is a strong demand for automated, flexible, and intelligent pipeline defect detection.

[0003] In pipeline defect detection, manual inspection or robots carrying instruments through preset paths are usually used to inspect the pipeline. The former cannot meet the efficiency of large-scale workpiece inspection, while the latter can effectively improve inspection efficiency, but the preset path is difficult to meet the diverse workpieces and is costly.

[0004] With the development and application of laser scanners and depth cameras in measurement and visual sensing, actual workpieces can be discretized into 3D point cloud data, providing new insights into intelligent path planning for robotic flaw detection. Therefore, intelligent path planning for robotic pipeline inspections based on 3D point cloud data, for pipe workpieces within acceptable dimensions, is key to improving the automation, flexibility, and intelligence of robotic pipeline defect detection. Summary of the Invention

[0005] In order to solve the problems existing in the prior art, the purpose of the present invention is to provide a pipeline detection path planning method, which can improve the adaptability of pipeline detection and significantly reduce the workload of relevant personnel for pipeline detection.

[0006] To achieve the above object, the present invention adopts a technical solution: a pipeline detection path planning method, comprising the following steps:

[0007] a. Obtaining original three-dimensional point cloud data of the pipeline shape and preprocessing the three-dimensional point cloud data;

[0008] b. Find the boundary points of the pipeline and establish the workpiece coordinate system based on the pre-processed 3D point cloud data;

[0009] c. Obtain the pipe section location and search direction, and calculate various section parameters;

[0010] d. Search exit determination;

[0011] e. Determination of cross-section correction;

[0012] f. Correct the section position and search direction;

[0013] g. Calculate the cross-section parameters and move in the search direction with appropriate steps to find a new cross-section;

[0014] h. Establish path node poses.

[0015] As a further improvement of the present invention, in step a, original three-dimensional point cloud data of the pipeline shape is obtained by a 3D laser scanner.

[0016] As a further improvement of the present invention, the step b specifically includes the following steps:

[0017] b1. Find a boundary point O in the 3D point cloud data: by finding the point cloud centroid, find the point farthest from the point cloud centroid as the boundary point O;

[0018] b2. Based on the points found in step b1, take the boundary point O as the coordinate origin, establish a translation matrix with the coordinate data of point O, and establish a workpiece coordinate system;

[0019] b3. Convert the 3D point cloud data from the scanner coordinate system to the workpiece coordinate system.

[0020] As a further improvement of the present invention, the step c specifically includes the following steps:

[0021] c1. Cut the pipe plane with the specified plane and project the points within a certain range near the cross section onto the cross section;

[0022] c2. Find three non-collinear points P1, P2, and P3 on the cross section: the point P0 (x0 y0 z0) closest to the origin O of the workpiece coordinate system on the cross section, with point P0 as the base point, find the point P1 on the cross section that is farthest from point P0, project P1 onto the z = 0 plane to obtain a new point P1 (x1 y1 0), use Euclidean distance to find the point P2 (x1 y2 z2) farthest from P1 on the cross section, project point P2 onto the z = 0 plane to obtain a new point P2 (x1 y2 0), use P1 and P2 as the major axis of the ellipse, take the midpoint as the center O1, and we know O1 (x1 (y1 + y2) / 2 0); find the point P3 (x1 y3 z3) closest to O1 in the y direction on the cross section, move P3 to point O1 in the z-axis direction to obtain the third point P3 (x1 (y1 + y2) / 2 z3);

[0023] c3. Based on the points obtained in step c1, construct plane X1 with |O1P1|=r1 and |O1P3|=r2 as the major and minor semi-axes respectively, construct vector S1 with O1 and P3, construct vector S2 with O1 and P1, and calculate the normal vector n1 of X1.

[0024] As a further improvement of the present invention, in step d, whether the cross-sectional data can complete the condition of constructing an elliptical cross-section is used to determine whether the pipeline detection path is completed. When there are fewer data points on the cross-section than can fit an ellipse, it is considered that the condition is met and the search is exited.

[0025] As a further improvement of the present invention, in step e, whether to correct the cross section is determined based on whether the major and minor semi-axes r1 and r2 obtained in step c3 satisfy |r1-r2|<δ. When δ is small enough, it is considered that the judgment condition is met and the cross section correction is exited.

[0026] As a further improvement of the present invention, the step f specifically includes the following steps:

[0027] f1. Based on the circle center position O1 and the cross-section semi-minor axis r2 obtained in step c, find the point P4 in the 3D point cloud data in the workpiece coordinate system that is closest to O1 and different from P3, and replace P4 with a point on the circle with O1 as the center and r2 as the radius;

[0028] f2, construct section X2 with points O1, P3, and P4, construct vector S3 with points O1 and P4; calculate the normal vector n2 of section X2;

[0029] f3. Calculate the angle θ between plane X1 and section X2 using n1 and n2;

[0030] f4. Establish the rotation matrix R Z (θ), with the workpiece coordinate system axis as the rotation axis, the base plane of the pipe cross section obtained by the workpiece coordinate system is rotated to a position parallel to the cross section X2. Since the workpiece coordinate system is rotated, the subsequent search direction remains unchanged.

[0031] As a further improvement of the present invention, the step g is specifically as follows:

[0032] The least squares method is used to fit the cross-section points to an ellipse, and the coordinates of the cross-section center and the major and minor axis diameters are obtained. The minor axis diameter of the ellipse is used to replace the circle radius, and the new cross-section is found by moving in the search direction with an appropriate step distance to obtain the next detection position.

[0033] As a further improvement of the present invention, the step h is specifically as follows:

[0034] Based on the workpiece coordinate system, a series of pipeline center coordinates are obtained. j (x j y j z j ), j∈(0, N), where N is the number of nodes, and an increment z_h is set in its z-axis direction j , z_h j=R, where R is the radius of the pipeline circular section, and the node position data point of the pipeline detection path is obtained. j (x j y j z j +z_h j ), j∈(0, N), for the node point j ,have:

[0035] ① The path node coordinate x direction is to be established, with α j Represents the x-direction of the node coordinate system, which is:

[0036]

[0037] ② The path node coordinate z direction is established, with γ j Represents the z direction of the node coordinate system, which is:

[0038] γ j =(0 0 1)

[0039] ③ The path node coordinate y direction is established, with β j Represents the y direction of the node coordinate system, which is:

[0040] β j =γ j ×α j

[0041] ④ Known β j , γ j Perpendicular to each other, re-establish the path node coordinate x direction α j ,have

[0042] α j =β j ×γ j

[0043] With α j , β j , γ j Construct path node poses.

[0044] The beneficial effects of the present invention are:

[0045] The present invention can improve the adaptability of pipeline detection and significantly reduce the workload of relevant staff for pipeline detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 is a schematic diagram of a flow chart of an embodiment of the present invention;

[0047] Figure 2 Schematic diagram of the pre-processed three-dimensional point cloud data of the arc-shaped pipeline workpiece to be inspected in an embodiment of the present invention;

[0048] Figure 3 Schematic diagram of converting the three-dimensional point cloud data of the arc-shaped pipe workpiece to be inspected into a three-dimensional point cloud in the workpiece coordinate system in an embodiment of the present invention;

[0049] Figure 4 3D point cloud of a cross section and a schematic diagram of the X1 cross section point of a curved pipe workpiece in a coordinate system according to an embodiment of the present invention;

[0050] Figure 5 Schematic diagram of the position of the X2 section point constructed in the modified section and search direction in an embodiment of the present invention;

[0051] Figure 6 Schematic diagram of the position of the cross section vector in the correction cross section and the search direction in an embodiment of the present invention;

[0052] Figure 7 Schematic diagram of the center position of the pipe cross section based on the workpiece coordinate system in an embodiment of the present invention;

[0053] Figure 8 Schematic diagram of the node positions of the pipeline detection path based on the workpiece coordinate system in an embodiment of the present invention;

[0054] Figure 9 Schematic diagram of the pipeline detection path node pose based on the workpiece coordinate system in an embodiment of the present invention. DETAILED DESCRIPTION

[0055] The embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0056] Example

[0057] This example uses an arc-shaped pipe as the workpiece to be inspected:

[0058] like Figure 1 As shown, a pipeline inspection path autonomous planning method includes the following steps:

[0059] a. Obtain the original 3D point cloud data of the pipeline workpiece shape and pre-process the original 3D point cloud data;

[0060] b. Find the boundary points of the pipeline and establish the workpiece coordinate system;

[0061] c. Obtain the pipe section location and search direction, and calculate various section parameters;

[0062] d. Search exit determination;

[0063] e. Determination of cross-section correction;

[0064] f. Correct the section position and search direction;

[0065] g. Calculate the cross-section parameters and move in the search direction with appropriate steps to find a new cross-section;

[0066] h. Establishment of path node pose.

[0067] In step a, the 3D point cloud data of the pipeline is obtained using a 3D laser scanner. Because the point cloud data obtained by a laser scanner in one scan is too large and contains some noise points and points outside the pipeline, the 3D point cloud data of the arc-shaped pipeline to be inspected in this embodiment has been pre-processed. The schematic diagram of the 3D point cloud of the arc-shaped pipeline to be inspected after pre-processing is shown in FIG. Figure 2 shown.

[0068] In step b, the specific method for finding the pipeline boundary point and establishing the workpiece coordinate system is as follows:

[0069] b1. Find a boundary point O in the 3D point cloud data. By obtaining the point cloud centroid, find the point farthest from the point cloud centroid as the boundary point O.

[0070] b2. Based on the point found in b1, take O as the coordinate origin, establish a translation matrix with the coordinate data of point O, establish a rotation matrix with the relationship between the scanner coordinate system and the worktable coordinate system, and establish a workpiece coordinate system with the coordinate axis z perpendicular to the worktable surface;

[0071] b3. Convert the 3D point cloud data from the scanner coordinate system to the workpiece coordinate system. The converted 3D point cloud diagram of the arc-shaped pipe to be inspected is as follows: Figure 3 shown.

[0072] like Figure 4 、 6 As shown, in step c, the position and search direction of the pipeline section are obtained, and the parameters of the section are calculated, which specifically includes:

[0073] c1. Cut the pipeline plane by the base plane of the workpiece coordinate system, and project the points within a certain range near the cross section onto the cross section.

[0074] c2. Find three non-collinear points P1, P2, and P3 on the cross section: the point P0 (x0 y0 z0) closest to the origin O of the workpiece coordinate system on the cross section, with point P0 as the base point, find the point P1 on the interface that is farthest from point P0, project P1 onto the z=0 plane to get a new point P1 (x1 y1 0), when the cross section is x=0, point P0 coincides with point P1, use Euclidean distance to find the point P2 (x1 y2 z2) farthest from P1 on the cross section, project point P2 onto the z=0 plane to get a new point P2 (x1 y2 0), take P1 and P2 as the major axis of the ellipse, take the midpoint as the center O1, and we know O1 (x1(y1+y2) / 2 0); find the point P3 (x1 y3 z3), move P3 to point O1 in the z-axis direction to obtain the third point P3(x1(y1+y2) / 2 z3);

[0075] c3. Based on the points obtained in c1, with |O1P1|=r1 and |O1P3|=r2 as the major and minor semiaxes respectively, construct plane X1 with O1, P1, and P3, construct vector S1 with O1 and P3, and construct vector S2 with O1 and P1, and obtain the normal vector n1 of plane X1, n1=S1×S2.

[0076] The search exit determination in step d refers to determining whether the pipeline detection path is complete based on whether the cross-section data can complete the construction of the elliptical cross-section. The search is exited when the number of data points on the cross-section is less than that required to fit the ellipse.

[0077] The section correction judgment in step e refers to judging whether the section needs to be corrected and whether the section correction is completed. The judgment condition is whether the major and minor semi-axes r1 and r2 obtained from c2 satisfy |r1-r2|<δ. When δ is small enough, it is considered that the judgment condition is met and the section correction is exited.

[0078] like Figure 5 、 6 As shown, the correction of the cross-section position and the search direction in step f specifically includes:

[0079] f1. Based on the center position O1 and the minor semi-axis r2 of the cross section obtained in step c, find the point P4 in the point cloud data in the workpiece coordinate system that is closest to O1 and different from P3. Point P4 is replaced by a point on the circle with O1 as the center and r2 as the radius.

[0080] f2. Construct plane X2 from points O1, P3, and P4, and construct vector S3 from points O1 and P4; calculate the normal vector n2 of plane X2, where n2 = S1 × S3;

[0081] f3, calculate the angle θ between plane X1 and plane X2 using n1 and n2,

[0082] f4. Create a rotation matrix With the workpiece coordinate system z-axis as the rotation axis, the base plane of the pipe section obtained by the workpiece coordinate system is rotated to a position parallel to the X2 section, and the pipe section is re-established. Since the workpiece coordinate system is rotated, the subsequent search direction remains unchanged;

[0083] Calculating the section parameters in step g and moving the appropriate step distance in the search direction to find a new section means using the least squares method to fit the section points to an ellipse after the current section has been determined to be a standard circular section, obtaining the section center coordinates and the major and minor axis diameters, replacing the circle radius with the minor semi-axis diameter of the ellipse, and moving the appropriate step distance in the search direction to find a new section to obtain the next detection position.

[0084] In step h, the path node pose is established, and a series of pipeline center coordinates are obtained based on the workpiece coordinate system after the above process is completed. j (x j y j z j ), j∈(0, N), where N is the number of nodes, such as Figure 7 As shown; set an increment z_h in its z-axis direction j (z_h j =R), where R is the radius of the pipeline circular section, and the node position data point of the pipeline detection path is obtained. j (x j y j z j +z_h j ), j∈(0, N), the pipeline detection path node position diagram is as follows Figure 8 shown.

[0085] For the node point j ,have:

[0086] ① The x-direction of the path node coordinate system is established, with α j Represents the x-direction of the node coordinate system, which is:

[0087]

[0088] ② The path node coordinate system is established in the z direction, with γ j Represents the z direction of the node coordinate system, which is:

[0089] γ j =(0 0 1)

[0090] ③ The path node coordinate system is established in the y direction, with β j Represents the y direction of the node coordinate system, which is:

[0091] β j =γj ×α j

[0092] ④ Known β j , γ j Perpendicular to each other, correct the path node coordinate system x direction α j ,have:

[0093] α j =β j ×γ j

[0094] With α j , β j , γ j Construct the path node posture, the pipeline detection path node posture diagram is as follows Figure 9 shown.

[0095] At this point, the pipeline detection path pose data based on the workpiece coordinate system has been completed. Subsequently, the data will be converted from the workpiece coordinate system back to the scanner coordinate system to obtain the pipeline detection path in the scanner coordinate system. According to the relationship between the scanner coordinate system and the tool coordinate system, the detection path pose data in the scanner coordinate system can be converted to the robot tool coordinate system equipped with the ultrasonic flaw detector probe.

[0096] The above-described embodiments merely represent specific implementations of the present invention. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that a person skilled in the art would be able to make numerous variations and improvements without departing from the spirit of the present invention, and all such variations and improvements fall within the scope of protection of the present invention.

Claims

1. A pipeline inspection path planning method, characterized in that: The following steps are involved: a. Obtaining original three-dimensional point cloud data of the pipeline shape and preprocessing the three-dimensional point cloud data; b. Find the boundary points of the pipeline and establish the workpiece coordinate system based on the pre-processed 3D point cloud data; c. Obtain the pipe section location and search direction, and calculate various section parameters; d. Search exit determination; e. Determination of cross-section correction; f. Correct the section position and search direction; g. Calculate the cross-section parameters and move in the search direction with appropriate steps to find a new cross-section; h. Establish path node poses; The step h is specifically as follows: Based on the workpiece coordinate system, a series of pipeline center coordinates are obtained. j (x j y j z j ), j∈(0, N), where N is the number of nodes, and an increment z_h is set in its z-axis direction j , z_h j =R, where R is the radius of the pipeline circular section, and the node position data point of the pipeline detection path is obtained. j (x j y j z j +z_h j ), j∈(0, N), for the node point j ,have: ① The path node coordinate x direction is to be established, with α j Represents the x-direction of the node coordinate system, which is: ② The path node coordinate z direction is established, with γ j Represents the z direction of the node coordinate system, which is: c j =(0 0 1) ③ The path node coordinate y direction is established, with β j Represents the y direction of the node coordinate system, which is: b j =c j ×a j ④ Known β j , γ j Perpendicular to each other, re-establish the path node coordinate x direction α j ,have a j =b j ×c j With α j , β j , γ j Construct path node poses.

2. The pipeline detection path planning method according to claim 1, characterized in that: In step a, the original three-dimensional point cloud data of the pipeline shape is obtained by a 3D laser scanner.

3. The pipeline detection path planning method according to claim 2, characterized in that: Described step b specifically comprises the following steps: b1. Find a boundary point O in the 3D point cloud data: by finding the point cloud centroid, find the point farthest from the point cloud centroid as the boundary point O; b2. Based on the points found in step b1, take the boundary point O as the coordinate origin, establish a translation matrix with the coordinate data of point O, and establish a workpiece coordinate system; b3. Convert the 3D point cloud data from the scanner coordinate system to the workpiece coordinate system.

4. The pipeline detection path planning method according to claim 3, characterized in that: The step c specifically comprises the following steps: c1. Cut the pipe plane by the specified plane and project the points within a certain range near the cross section onto the cross section; c2. Find three non-collinear points P1, P2, and P3 on the cross section: the point P0 (x0y0z0) closest to the origin O of the workpiece coordinate system on the cross section, with point P0 as the base point, find the point P1 on the cross section that is farthest from point P0, project P1 onto the z=0 plane to obtain a new point P1 (x1y10), use Euclidean distance to find the point P2 (x1y2z2) farthest from P1 on the cross section, project point P2 onto the z=0 plane to obtain a new point P2 (x1y20), use P1 and P2 as the major axis of the ellipse, take the midpoint as the center O1, and we know O1 (x1(y1+y2) / 20); find the point P3 (x1y3z3) closest to O1 in the y direction on the cross section, move P3 to point O1 in the z-axis direction to obtain the third point P3 (x1(y1+y2) / 2z3); c3. Based on the points obtained in step c1, construct plane X1 with |O1P1|=r1 and |O1P3|=r2 as the major and minor semi-axes respectively, construct vector S1 with O1 and P3, construct vector S2 with O1 and P1, and calculate the normal vector n1 of X1.

5. The pipeline detection path planning method according to claim 4, characterized in that: In step d, whether the cross-section data can complete the conditions for constructing an elliptical cross-section is used to determine whether the pipeline detection path is completed. When there are fewer data points on the cross-section than can fit an ellipse, the conditions are considered to be met and the search is exited.

6. The pipeline detection path planning method according to claim 5, characterized in that: The section correction judgment in step e refers to judging whether the section needs to be corrected and whether the section correction is completed. The judgment condition is whether the major and minor semi-axes r1 and r2 obtained from c3 satisfy |r1-r2|<δ. When δ is small enough, it is considered that the judgment condition is met and the section correction is exited.

7. The pipeline detection path planning method according to claim 4, 5 or 6, characterized in that: The step f specifically comprises the following steps: f1. Based on the circle center position O1 and the cross-section semi-minor axis r2 obtained in step c, find the point P4 in the 3D point cloud data in the workpiece coordinate system that is closest to O1 and different from P3, and replace P4 with a point on the circle with O1 as the center and r2 as the radius; f2, construct section X2 with points O1, P3, and P4, construct vector S3 with points O1 and P4; calculate the normal vector n2 of section X2; f3. Calculate the angle θ between plane X1 and section X2 using n1 and n2; f4. Establish the rotation matrix R Z (θ), with the workpiece coordinate system axis as the rotation axis, the base plane of the pipe cross section obtained by the workpiece coordinate system is rotated to a position parallel to the cross section X2. Since the workpiece coordinate system is rotated, the subsequent search direction remains unchanged.

8. The pipeline detection path planning method according to claim 7, characterized in that: The step g is specifically as follows: The least squares method is used to fit the cross-section points to an ellipse, and the coordinates of the cross-section center and the major and minor axis diameters are obtained. The minor axis diameter of the ellipse is used to replace the circle radius, and the new cross-section is found by moving in the search direction with an appropriate step distance to obtain the next detection position.

Images