A method for detecting the height and width of circumferential welds in straight sections of pipelines

Abstract

The present invention relates to a method for detecting the height and width of a circumferential weld of a straight segment of a pipeline, comprising the following steps: obtaining point clouds of the pipeline weld and the straight segments before and after the weld; fitting small cylinders segment by segment with the weld design width as a step size; iteratively fusing adjacent cylindrical segments with similar radii to separate the straight segment and the weld rough positioning segment; calculating the distance from the weld rough positioning segment point cloud to the straight segment cylinder axis to obtain the precise positioning of the weld; calculating the difference between the maximum distance from the weld point cloud to the pipeline axis and the pipeline radius to obtain the weld height; projecting the weld point cloud onto the pipeline axis to obtain a new projected point cloud, calculating the maximum distance between the points in the projected point cloud to obtain the weld width. The present invention can detect the height and width of the circumferential weld of a straight segment pipeline weld, providing a new method for automated detection of the height and width of the circumferential weld of a straight segment pipeline.

Description

Technical Field

[0001] The present invention relates to automated detection, in particular to a method for detecting the height and width of a circumferential weld of a straight section of a pipeline, which is used to realize automated detection of the height and width of a circumferential weld of a straight section of a pipeline. Background Art

[0002] As the main transmission medium for gases and liquids, pipelines are widely used in the aerospace field due to their advantages such as light weight and high toughness. They are an important part of aircraft engine transmission systems.

[0003] Welding is the process of connecting and assembling different pipelines. To achieve high-quality welded pipelines and improve welding automation, weld quality inspection is crucial. Welding quality is primarily determined by several factors, including weld formation, weld microstructure, and weld mechanical properties. The weld's appearance, specifically its width and height, are crucial indicators of weld quality and are closely related to weld quality. A weld that is too wide indicates severe heat exposure. The increased grain size caused by thermal deformation in the weld will degrade the weld's mechanical properties, leading to weld deformation. A weld that is too narrow indicates poor fusion between the weld and the parent material, leading to stress concentration in the unfused area. This can cause surface defects such as cracks and undercuts, directly impacting the weld's load-bearing strength. Excessive weld reinforcement increases weld grinding, reducing welding efficiency. Excessive weld reinforcement reduces the weld's mechanical properties.

[0004] Due to the complex three-dimensional shape of actual pipelines, at this stage, general-sized enterprises still use traditional measuring tools combined with manual visual inspection to inspect the appearance quality of pipeline welds. Some institutions have conducted some relevant research on the automated identification and detection of pipeline welds. For example: 1) A digital positioning device and positioning method for pipeline welds, which are used to locate welds between sections in a pipeline. When locating a weld, the digital positioning device is placed at the weld to be located, and a built-in GPS device is turned on to record the position data of the weld. However, this method is not accurate enough and is costly. 2) A weld gap measurement device uses a taper measuring device to enter the weld gap in the melt pipeline for contact measurement. This contact measurement has a low degree of automation. 3) An improved automatic detection algorithm for pipeline welds determines the main direction of the pipeline by calculating the gradient-direction histogram of the pipeline image, enhances the weld features through HSV color space conversion, and realizes weld positioning through template matching. However, this method only realizes image positioning of the weld from a two-dimensional perspective and cannot provide three-dimensional measurement. 4) Structured light-based weld surface quality detection uses a random sampling consistency algorithm to fit a plane to segment the weld point cloud, uses a principal component analysis algorithm to obtain the main direction of the weld point cloud, obtains the weld contour line through point cloud slicing, and combines the slope to analyze the weld surface quality. This method cannot calculate the width and height of the weld. Summary of the Invention

[0005] In view of the above-mentioned defects in the prior art and the existing practical needs, the present invention provides a method for detecting the height and width of the circumferential weld of a straight segment of a pipeline, which collects images of a welded pipeline with a weld from multiple angles; obtains point clouds of the pipeline weld and the straight segments before and after the weld; fits small cylinders segment by segment with the weld design width as the step size; iteratively fuses adjacent cylindrical segments of the same radius to separate the straight segment and the weld coarse positioning segment; calculates the distance from the weld coarse positioning segment point cloud to the cylinder axis of the straight segment to obtain the precise positioning of the weld; calculates the difference between the maximum value of the distance from the weld point cloud to the pipeline axis and the pipeline radius to obtain the weld height; projects the weld point cloud onto the pipeline axis to obtain a new point cloud, calculates the maximum distance between the points of the new point cloud, and obtains the weld width.

[0006] The technical solution adopted by the present invention to achieve the above-mentioned purpose is:

[0007] A method for detecting the height and width of a circumferential weld of a straight section of a pipeline is provided, which is used for detecting the weld position after welding two known straight sections of a pipeline. The method comprises the following steps:

[0008] Collect images of welded pipes with welds from multiple angles and fuse them into original point clouds; preliminarily estimate candidate areas of the weld circumferential point cloud based on weld design parameters;

[0009] Taking the weld design width as the step length, the candidate area of ​​the weld circumferential point cloud is axially segmented, and each small cylinder is fitted one by one to obtain the radius and axis of the cylinder;

[0010] Iteratively merge adjacent cylindrical segments to segment the weld rough positioning segment and the two straight pipe segments A and B on both sides of it;

[0011] The two straight pipelines are further fitted into a fused pipeline passing through the weld, and the distance from the weld rough positioning segment point cloud to the fused pipeline axis is calculated to obtain the weld precise point cloud P.

[0012] Calculate the distance from the precise point cloud of the weld to the axis of the fused pipeline, and take the difference between the maximum distance and the radius of the fused pipeline as the weld height; project the precise point cloud of the weld onto the axis of the fused pipeline to obtain the projected point cloud P1, and calculate the span of the projected point cloud on the axis as the weld width, so as to finally identify the weld on the welding pipeline.

[0013] The multi-angle acquisition of images of welded pipelines with welds is as follows: a plurality of positioning points are evenly arranged around the circumference of the welded pipeline, and multiple cameras are used to simultaneously capture weld images of corresponding angles at each positioning point, or a single camera is used to capture weld images of corresponding angles at each positioning point one by one; or a single camera is controlled to rotate with the welded pipeline centered on the axial direction so as to display different angles at a preset position, so as to capture weld images of multiple angles.

[0014] The fusion into the original point cloud is to fuse the local weld images collected from multiple angles, and filter out abnormal points and interfering noise points through filtering and denoising to obtain a weld pipeline point cloud with welds and circumferential integrity.

[0015] The weld seam design parameter is the weld seam design width d required by the welding process.

[0016] The preliminary estimation of the candidate area of ​​the weld point cloud based on the welding parameters includes: taking the weld design width d required by the process as a benchmark, defining several multiples N×d of the weld design width d as further identifying the axial span of the weld, and an integer N>9 is used to narrow the point cloud range to obtain the candidate area of ​​the weld circumferential point cloud.

[0017] The iterative fusion of adjacent cylindrical segments to segment the weld rough positioning segment and the two straight line segments A and B on both sides thereof includes:

[0018] If the radius difference and axis misalignment of adjacent cylindrical segments meet the error threshold, the point cloud data of adjacent cylindrical segments are fused and fitted to generate a new cylindrical segment, and the new radius and axis are obtained;

[0019] The above process is repeated iteratively to traverse all small cylinders, and finally three new cylinder segments are generated;

[0020] The cylindrical section with the largest radius and the largest angle between its axis and other axes is used as the rough positioning section of the weld, and the other two cylindrical sections are used as the straight pipelines A and B on both sides of the rough positioning section of the weld.

[0021] The two straight pipe segments A and B are further fitted into a fused cylindrical segment passing through the weld, and the distance from the weld rough positioning segment point cloud to the fused cylindrical segment axis is calculated to achieve precise weld positioning, including:

[0022] The point clouds of the cylindrical segments A and B before and after the weld rough positioning segment are fused, and the cylindrical equation is fitted to obtain the axis L and radius R of the fused cylindrical segment;

[0023] The distance h from any point cloud in the rough weld positioning segment to the axis of the fused cylindrical segment is calculated. If the difference between h and R is greater than the threshold, it is considered to be the weld point cloud P, and the weld is accurately positioned.

[0024] The span is the maximum value of the distance between any two points in the weld point cloud P and the projected point cloud P1.

[0025] The present invention has the following beneficial effects and advantages:

[0026] The method of the present invention provides a method for detecting the circumferential weld height and width of a straight segment pipeline. By combining the design model parameters, small cylinders are fitted segment by segment in the pipeline point cloud with the weld design width as the step size and iterative analysis is performed, thereby realizing integration with the pipeline design information. This method makes it applicable to pipelines of different diameters, and can achieve accurate positioning of welds, and can realize high-precision positioning and identification of weld width and height. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 is a flow chart of the method of the present invention;

[0028] Figure 2 Schematic diagram of weld width calculation using the method of the present invention. DETAILED DESCRIPTION

[0029] To make the above-mentioned objects, features, and advantages of the present invention more readily apparent, the specific implementation methods of the present invention are described in detail below with reference to the accompanying drawings. The following description sets forth many specific details to facilitate a full understanding of the present invention. However, the present invention can be implemented in many other ways than those described herein, and those skilled in the art may make similar modifications without departing from the scope of the invention. Therefore, the present invention is not limited to the specific implementation methods disclosed below.

[0030] like Figure 1 FIG. 1 is a flow chart of a method for detecting the height and width of a circumferential weld of a straight section of a pipeline according to the present invention, comprising the following steps:

[0031] Capture images of welded pipelines with welds from multiple angles;

[0032] Obtain the point cloud of the pipeline weld and the straight line segments before and after the weld;

[0033] Input design model template;

[0034] Fit the small cylinder segment by segment with the weld design width as the step length;

[0035] Iteratively merge adjacent cylindrical segments of the same radius to separate straight line segments and weld rough positioning segments;

[0036] Calculate the distance from the weld rough positioning segment point cloud to the straight segment cylindrical axis to achieve accurate weld positioning;

[0037] Calculate the difference between the maximum distance between the weld point cloud and the pipeline axis and the pipeline radius to obtain the weld height;

[0038] The weld point cloud is projected onto the pipeline axis to obtain a new point cloud, and the maximum distance between the points in the new point cloud is calculated to obtain the weld width.

[0039] like Figure 2 As shown, Figure 2The schematic diagram of the weld width calculation method of the present invention is as follows: Figure 2 The following are some key steps to explain in detail:

[0040] Step 1. Collect images of welded pipes with welds from multiple angles and fuse them into original point clouds;

[0041] Three methods can be used to collect point cloud data: Arrange six positioning points around the weld pipe and use six cameras to simultaneously capture weld images from corresponding angles; use a single camera to capture weld images from corresponding angles at each of the six positioning points; or control the weld pipe to rotate about its axis, presenting different angles to a single camera at a preset position, allowing it to capture weld images from multiple angles. In this example, line structured light is used as the light source to illuminate the weld area of ​​the weld pipe, and the camera is a laser scanning camera.

[0042] The local point clouds at various angles are fused into the initial point cloud using the ICP point cloud registration method, and then the abnormal points and interference noise points are filtered out through filtering and denoising to obtain a clean original point cloud.

[0043] Step 2. Obtain the original point cloud of the pipeline weld and the straight line segments before and after the weld, including:

[0044] Obtain a complete circumferential point cloud of the straight section weld of the pipeline and a distance of not less than 4 times the weld width d in front and behind. If the theoretical design width of the weld is 5mm, then in addition to the weld design width d, a point cloud at least 20mm long should be retained in front and behind, that is, a point cloud of at least 45mm in length and width as the weld candidate area.

[0045] Step 3. Enter the design model template, including:

[0046] The method of the present invention is a calculation and identification performed in a computer three-dimensional model software, and therefore requires input of pipeline weld-related design model parameters, including the diameters of the front and rear straight segments A and B, and the design width d and height of the weld.

[0047] Step 4. Fit the small cylinder segment by segment using the weld design width as the step size, including:

[0048] In the weld candidate area, at the starting position of the input point cloud, with the weld design width d as the step size, the candidate area point cloud is divided into multiple circumferential small point clouds with a length of the weld design width d. Each circumferential small point cloud is fitted with a small cylinder to obtain the corresponding radius and axis.

[0049] Step 5. Iteratively merge adjacent cylindrical segments of the same radius to segment straight line segments and weld rough positioning segments, including:

[0050] Within the weld candidate area, starting from the small cylindrical segment at the starting position and moving towards the small cylindrical segment at the end, the radius and axis of two adjacent small cylindrical segments are analyzed and compared. When the radius difference is less than 0.1mm and the axis misalignment is less than 0.05 degrees, they are considered to be the same cylinder segment. In this way, these two point clouds are fused and a new cylinder is generated by fitting. This process is repeated until the cylindrical point clouds cannot be merged. The cylinder is fitted again to obtain the radius and axis. The cylindrical segment with the largest radius and the largest angle between the axis and other axes is the weld area, i.e., the weld rough positioning segment.

[0051] It should be noted that the radius difference threshold of 0.1mm in step 5 is obtained by comprehensively considering the weld design height (greater than 0.5mm), thermal deformation caused by welding, and the accuracy of the input point cloud. When the input point cloud accuracy is higher or the weld design height is higher, this threshold can be further relaxed; when the weld design height is lower, this threshold can be reduced accordingly.

[0052] Step 6. Calculate the distance from the weld rough positioning segment point cloud to the straight line segment cylinder axis to achieve accurate weld positioning, including:

[0053] The cylindrical point clouds before and after the rough positioning section of the weld are fused, and the cylinder equation is fitted to obtain the pipeline axis L and radius R. The distance h from the rough positioning section point cloud of the weld to the pipeline axis is calculated, and the difference between h and R is calculated. When the difference is greater than 0.25 mm, it is considered to be the precise point cloud P of the weld. At this point, the precise positioning of the weld is obtained.

[0054] It should be noted that the threshold of the difference greater than 0.25mm in step 5 is a threshold value obtained by comprehensively considering the weld design height (greater than 0.5mm), point cloud accuracy, and the theoretical height of the weld edge (the weld is an arch with the edge lower than the middle).

[0055] Step 7. Calculate the difference between the maximum distance between the weld point cloud P and the fused pipeline axis and the pipeline radius to obtain the weld height, including:

[0056] The weld point cloud P and pipeline axis L obtained in step 6 are traversed to calculate the maximum distance from the weld point cloud P to the fused pipeline axis L, and the difference between the distance and the pipeline radius is calculated to obtain the weld height.

[0057] It should be noted that since the input point cloud in step 1 is a point cloud that has been filtered and denoised, the maximum value in step 7 will not be obviously unreasonable, such as being greater than 10 times the design height.

[0058] Step 8. Project the weld point cloud P onto the fusion pipeline axis to obtain a new point cloud P1. Calculate the maximum distance between the points in the new point cloud P1 to obtain the weld width, including:

[0059] The weld point cloud P and the fusion pipeline axis L obtained in step 6 are projected onto the fusion pipeline axis L to obtain a new point set P1. The interior of P1 is traversed, and the distance between any two points is calculated to obtain the maximum distance, that is, the weld width.

[0060] In summary, the method of the present invention provides a method for detecting the height and width of circumferential welds in straight sections of pipelines. Combined with pipeline design information, it is not restricted by pipe diameter and material, and can be applied to situations where the front and rear pipe diameters change. Moreover, by analyzing the three-dimensional point cloud, the weld position can be accurately located, with high efficiency and precision, and is suitable for use in scenarios where there is a need to detect the height and width of welds in straight sections of pipelines.

[0061] The specific embodiments described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above are only specific embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for detecting the height and width of a circumferential weld of a straight section of a pipeline, characterized in that: It is used to detect the weld positioning after welding two known straight pipe sections; the method includes the following steps: Images of welded pipes with welds are collected from multiple angles and fused into an original point cloud. The candidate regions of the weld circumferential point cloud are preliminarily estimated based on the weld design parameters. This includes: using the weld design width d required by the process as a reference, defining multiples of the weld design width d (N×d) as further identification of the weld axial span, with N > 9 integers, to narrow the point cloud range and obtain the candidate regions of the weld circumferential point cloud. Taking the weld design width as the step length, the candidate area of ​​the weld circumferential point cloud is axially segmented, and each small cylinder is fitted one by one to obtain the radius and axis of the cylinder; Iteratively merge adjacent cylindrical segments to segment the weld rough positioning segment and the two straight line segments A and B on both sides of it; including: If the radius difference and axis misalignment of adjacent cylindrical segments meet the error threshold, the point cloud data of adjacent cylindrical segments are fused and fitted to generate a new cylindrical segment, and the new radius and axis are obtained; The above process is repeated iteratively to traverse all small cylinders, and finally three new cylinder segments are generated; The cylindrical section with the largest radius and the largest angle between its axis and other axes is used as the rough positioning section of the weld, and the other two cylindrical sections are used as the straight pipe sections A and B on both sides of the rough positioning section of the weld; The two straight pipelines are further fitted into a fused pipeline passing through the weld, and the distance from the weld rough positioning segment point cloud to the fused pipeline axis is calculated to obtain the weld precise point cloud P. This includes: The point clouds of the cylindrical segments A and B before and after the weld rough positioning segment are fused, and the cylindrical equation is fitted to obtain the axis L and radius R of the fused cylindrical segment; Calculate the distance h from any point cloud in the rough weld positioning segment to the axis of the fused cylindrical segment. If the difference between h and R is greater than the threshold, it is considered to be the weld point cloud P, and the weld is accurately positioned. Calculate the distance from the precise point cloud of the weld to the axis of the fused pipeline, and take the difference between the maximum distance and the radius of the fused pipeline as the weld height; project the precise point cloud of the weld onto the axis of the fused pipeline to obtain the projected point cloud P1, and calculate the span of the projected point cloud on the axis as the weld width, so as to finally identify the weld on the welding pipeline.

2. A method for detecting the height and width of a circumferential weld of a pipeline straight section according to claim 1, characterized in that: The multi-angle acquisition of images of welded pipelines with welds is as follows: a plurality of positioning points are evenly arranged around the circumference of the welded pipeline, and multiple cameras are used to simultaneously capture weld images of corresponding angles at each positioning point, or a single camera is used to capture weld images of corresponding angles at each positioning point one by one; or a single camera is controlled to rotate with the welded pipeline centered on the axial direction so as to display different angles at a preset position, so as to capture weld images of multiple angles.

3. A method for detecting the height and width of a circumferential weld of a pipeline straight section according to claim 1, characterized in that: The fusion into the original point cloud is to fuse the local weld images collected from multiple angles, and filter out abnormal points and interfering noise points through filtering and denoising to obtain a weld pipeline point cloud with welds and circumferential integrity.

4. A method for detecting the height and width of a circumferential weld of a pipeline straight section according to claim 1, characterized in that: The span is the maximum value of the distance between any two points in the weld point cloud P and the projected point cloud P1.

Images