A method and system for identifying a welding starting point of a trihedral corner joint
By using a three-point sampling and point cloud merging optimization method, the welding start point of a three-sided corner weld is identified, which solves the problem of low production efficiency in the welding of three-sided corner welds and achieves efficient and accurate welding start point identification, meeting the requirements of full automation.
Patent Information
- Application Number
- CN202211351836.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-10-31
- Publication Date
- 2025-12-09
- Estimated Expiration
- 2042-10-31
AI Technical Summary
Existing technologies for welding three-sided corner joints have low production efficiency, cannot perform large-scale automated welding, and involve tedious manual teaching work with high worker turnover, resulting in low welding accuracy and efficiency.
A method and system for identifying the welding start point of a three-sided corner weld is designed by employing three-point sampling to find a qualified straight line model, reducing the influence of noise by optimizing point cloud merging rules, transforming the coordinates of weld corner points to the robot coordinate system, and introducing weight coefficients to estimate the welding start point.
Under different sizes and workpiece deformation conditions, it can accurately identify the welding start point, meet the requirements of fully automated welding, improve algorithm efficiency by 60%, improve accuracy by 43%, and control the welding start point fitting repeatability accuracy within 0.34mm.
Smart Images

Figure CN115846945B_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of welding process, in particular to a method and system for identifying a welding starting point of a three-corner joint. BACKGROUND
[0002] With the rapid development of intelligent manufacturing technology, the research of robot intelligent welding is more and more extensive. In the absence of manual teaching, accurate identification of the welding starting point is the first step for robots to realize intelligent welding, and directly affects the performance of subsequent robot weld automatic tracking. Therefore, accurate identification of the welding starting point is one of the key technologies of intelligent welding. There are usually two methods for identifying the welding starting point: mechanical identification and visual identification. Mechanical identification is to fix the welding starting point at a position by a mechanical positioning mechanism each time welding. At this time, the welding starting point position accuracy is greatly affected by factors such as the quality of the parts, the experience of engineers, the wear degree of the positioning mechanism and the replacement of welding rods, making it difficult to weld a large number of workpieces. Visual identification is to use a visual sensor to find the welding starting point position, and then realize intelligent identification of the starting point, which has strong flexibility and is a promising identification method.
[0003] Three-corner joints are widely used in large welding parts such as truck carriage side plates and ship cabin grids. Currently, there are problems such as workpiece deformation, making it difficult to keep the size consistent with the numerical model. If automatic welding is used, manual teaching of the welding starting point of each welding part is required, which will slow down the welding process and reduce the welding accuracy. The current three-corner joint welding parts still mainly use manual welding. However, through research, the author found that the related robot teaching work is boring, the worker turnover rate is high, and the production efficiency is low. Therefore, there is an urgent need for full-automatic welding in such factories.
[0004] Therefore, there is a need to design a welding starting point identification algorithm for three-corner joint welding to meet the needs of full-automatic welding. SUMMARY
[0005] This section aims to summarize some aspects of the embodiments of the present application and briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this section and the abstract and title of the specification to avoid obscuring the purpose of this section, abstract and title, and such simplifications or omissions cannot be used to limit the scope of the present application.
[0006] In view of the above existing problems, the present application is proposed.
[0007] Therefore, the present application provides a method and system for identifying a welding starting point of a three-corner joint, which can solve the problem of low production efficiency and inability to perform mass welding work of the traditional method.
[0008] To solve the above technical problems, the present application provides the following technical solutions, a kind of three-corner joint welding starting point identification method, comprising:
[0009] Collect the point cloud profile data of the section of three-corner joint;
[0010] Three-point sampling is used to find qualified straight line model;
[0011] Point cloud merging optimization rule is used to reduce the influence of noise;
[0012] Convert the weld corner point coordinates to the robot coordinate system, and introduce the weight coefficient in the weld straight line fitting;
[0013] Estimate the welding starting point.
[0014] As a preferred scheme of the three-corner joint welding starting point identification method of the present application, wherein: the collection of point cloud profile data includes using a line structured light camera to shoot a certain section position of the three-corner joint, and repeatedly sampling points, the first group of data of the starting point position is randomly taken at a distance of 10 mm from the starting point, and the subsequent point cloud data is taken at a fixed distance interval.
[0015] As a preferred scheme of the three-corner joint welding starting point identification method of the present application, wherein: the three-point sampling includes randomly selecting 3 sample points in the point cloud data, and the distance between any two sample points should be greater than D, 3 sample points can obtain 3 straight line models, and the calculation formula of D is represented as:
[0016]
[0017] Wherein, σ comprehensively represents the surface roughness of the welding part and the point cloud extraction error, which can be selected according to the actual situation.
[0018] As a preferred scheme of the three-corner joint welding starting point identification method of the present application, wherein: the three-point sampling further includes,
[0019] Select a straight line model, calculate the distance between it and a sample point not on the model, and according to the distance, all point clouds are divided into local points and local points of the straight line, set a reasonable error tolerance threshold t, if the distance is less than the threshold t, the model pre-inspection passes, otherwise calculate the remaining models.
[0020] As a preferred scheme of the three-corner joint welding starting point identification method of the present application, wherein: when 3 straight line models do not pass the inspection, 3 sample points are selected again for three-point sampling modeling until the model pre-inspection passes, the number of local points of the model passing the pre-inspection is calculated, and if the preset number requirement is met, the straight line model and the local points are retained.
[0021] As a preferred scheme of the three-corner angle joint welding starting point recognition method, the point cloud merging optimization comprises: selecting a straight line model with the largest number of local points as a main model, and calculating the angle between the remaining straight line models and the main model; if the included angle is less than a threshold θ, the point cloud data contained by the straight line model is merged, and the straight line model is re-estimated, and the calculation formula of the angle threshold is represented as:
[0022]
[0023] Wherein, h is the maximum straightness allowed by the workpiece, and w is the point cloud collection width of the camera on a single face of the welding piece.
[0024] As a preferred scheme of the three-corner angle joint welding starting point recognition method, the weight coefficient introduction comprises: numbering the weld corner points according to the distance from the welding starting point from near to far as 1, 2, 3, …, i; calculating the distance of the i-th point to the straight line model after introducing the weight coefficient The formula is represented as:
[0025]
[0026]
[0027] Wherein, k i is the weight of the corner point i, n is the number of local points contained by the model, A, B, and C are the coefficients of the straight line model.
[0028] As a preferred scheme of the three-corner angle joint welding starting point recognition method, the estimation of the welding starting point comprises: the two straight line models finally calculated are not intersected in space, so the midpoint of the common perpendicular line of the two straight line models is taken as the weld starting point.
[0029] A three-corner angle joint welding starting point recognition system comprises: a data acquisition module for acquiring point cloud profile data of a partial cross section of a three-corner angle joint;
[0030] A model pre-inspection module connected with the data acquisition module for pre-inspecting the straight line model;
[0031] A model optimization module connected with the model pre-inspection module for merging and optimizing the point cloud to reduce the influence of noise points;
[0032] An output module connected with the model optimization module for introducing a weight coefficient to estimate the welding starting point.
[0033] The beneficial effects of the present application: the present application can fit the welding starting point through interval sampling on both sides of the welding starting point in the case of different sizes and larger deformation of workpieces, and can fully meet the process requirements of general "three-corner joint" welding by cooperating with the tracking function of the welding robot, the algorithm of the present application improves the efficiency by 60% and the accuracy by 43% compared with the traditional RANSAC algorithm, and the repeatability of welding starting point fitting can be controlled within 0.34mm. BRIEF DESCRIPTION OF DRAWINGS
[0034] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor. Among them:
[0035] Figure 1 A three-corner joint welding starting point identification method flow chart is provided for an embodiment of the present application.
[0036] Figure 2 An angle threshold diagram of a three-corner joint welding starting point identification method is provided for an embodiment of the present application.
[0037] Figure 3 An initial point cloud sample collection diagram of a three-corner joint welding starting point identification method is provided for an embodiment of the present application.
[0038] Figure 4 An RMSE error mean diagram of welding starting point fitting of a three-corner joint welding starting point identification method is provided for an embodiment of the present application.
[0039] Figure 5 A different sampling interval RMSE mean trend diagram of a three-corner joint welding starting point identification method is provided for an embodiment of the present application.
[0040] Figure 6 A welding starting point diagram obtained by the algorithm before and after improvement of a three-corner joint welding starting point identification method is provided for an embodiment of the present application. DETAILED DESCRIPTION
[0041] In order to make the above objectives, characteristics and advantages of the present application more obvious and easy to understand, the specific embodiments of the present application are described in detail below with reference to the drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative work should fall within the protection scope of the present application.
[0042] In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present application. The present application, however, can be practiced in a variety of ways beyond the specific details set forth herein without departing from the scope of the present application. It can be appreciated by those skilled in the art that the present application can be practiced without such specific details.
[0043] Secondly, "one embodiment" or "an embodiment" referred to herein means that a specific feature, structure or characteristic described can be included in at least one implementation of the present application. "In one embodiment" appearing in different places in the specification does not mean the same embodiment, nor does it mean that the embodiment is separate or alternative to other embodiments.
[0044] The present application is described in detail with reference to the drawings. In the detailed description of the embodiments of the present application, the cross-sectional view of the device structure is partially enlarged without the general proportion for the convenience of description, and the schematic view is only an example, which should not limit the scope of protection of the present application. In addition, three-dimensional spatial dimensions including length, width and depth should be included in actual manufacture.
[0045] Meanwhile, in the description of the present application, it should be noted that the terms "upper, lower, inner and outer" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the device or element referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as a limitation on the present application. In addition, the terms "first, second or third" are only for the purpose of description, and cannot be understood as indicating or implying relative importance.
[0046] In the present application, unless otherwise explicitly specified and limited, the terms "mounting, connecting, connection" should be understood broadly, for example: it can be fixed connection, detachable connection or integral connection; it can also be mechanical connection, electrical connection or direct connection, it can also be indirectly connected through intermediate medium, or it can be the communication inside two elements. For those skilled in the art, the specific meaning of the above terms in the present application can be understood according to the specific circumstances.
[0047] Example 1
[0048] Reference Figure 1For the first embodiment of the present application, the embodiment provides a three-corner joint welding starting point identification method and system, comprising:
[0049] S1: Collect point cloud profile data of a partial section of the three-corner joint;
[0050] It should be noted that the line structured light camera is used to shoot a certain section position of the three-corner joint, and the points are repeatedly collected to obtain the point cloud profile data of the partial section of the joint.
[0051] Further, 10 groups of point cloud data are taken every certain distance on both sides of the starting point position O of the workpiece, and a total of 20 groups of data are used as initial samples. However, the first group of data of the starting point position should be randomly taken near 10 mm from the starting point O, and the subsequent point cloud data should be taken at a fixed distance interval.
[0052] S2: Three-point sampling is used to find a qualified straight line model;
[0053] Further, 3 sample points are randomly selected in the point cloud data, and the distance between any two sample points should be greater than D. 3 straight line models can be obtained by connecting the 3 sample points in pairs, and the calculation formula of D is as follows:
[0054]
[0055] Among them, sigma comprehensively represents the surface roughness of the joint and the point cloud extraction error, which can be selected according to the actual situation.
[0056] It should be noted that the distance between the candidate model and the points not on the model is calculated, all point clouds are divided into local points and local points of the straight line according to the distance between the points and the straight line model, and a reasonable error tolerance threshold t is set. If the distance is less than the threshold t, the model pre-inspection passes, otherwise the remaining models are calculated. If none of the three straight line models passes the model pre-inspection, the point selection and model pre-inspection work are re-executed, and if it is greater than the threshold, the model passes the pre-inspection.
[0057] S3: Point cloud merging optimization rule is used to reduce the influence of noise points;
[0058] Further, the straight line model with the largest number of local points is selected as the main model, and the angle between the remaining straight line models and the main model is calculated. If the included angle is less than the threshold value theta, the point cloud data contained in the straight line model is merged, and the model is re-estimated using the least square method or the like. The calculation formula of the angle threshold value is as follows:
[0059]
[0060] Among them, h is the maximum straightness allowed by the workpiece, and w is the point cloud collection width of the camera on a single face of the joint. The above steps are continued in the unmerged model.
[0061] S4: Convert the weld corner point coordinates to the robot coordinate system, and introduce a weight coefficient in the weld straight line fitting.
[0062] Further, the weld corner point coordinates are converted to the same coordinate system by using the hand-eye relationship matrix. When the starting point of welding is calculated by using the corner point fitting straight line model, the corner points far from the starting point O have lower reliability. If all the corner points have the same contribution, it will lead to larger fitting error. If the corner points far from the starting point O are removed, it will also damage the integrity of the point cloud information.
[0063] By introducing a weight coefficient k in the weld straight line fitting, the contribution degree is allocated according to the distance of the corner point from the welding starting point O, the tolerance of the algorithm to the point cloud data far from the welding starting point is enhanced, and the fitting accuracy is improved. The weld corner points are numbered 1, 2, 3, …, i according to the distance from the welding starting point from near to far; the distance of the i-th point to the straight line model after introducing the weight coefficient is calculated , which is expressed by the formula:
[0064]
[0065] wherein k i is the weight of the corner point i, n is the number of local points contained in the model, A, B and C are the coefficients of the straight line model.
[0066] S5: Estimate the welding starting point.
[0067] It should be noted that the two straight line models obtained at this time are not intersected in space. It is verified by experiments that the midpoint of the common perpendicular line of the two straight line models can be taken as the welding starting point, and the distance from the midpoint of the common perpendicular line to the two straight lines is the shortest.
[0068] Further, a three-corner joint welding starting point identification system comprises:
[0069] A data acquisition module for acquiring point cloud profile data of a partial cross section of a three-corner joint;
[0070] A model pre-inspection module connected with the data acquisition module for pre-inspecting the straight line model;
[0071] A model optimization module connected with the model pre-inspection module for merging and optimizing the point cloud to reduce the influence of noise points;
[0072] An output module connected with the model optimization module for introducing a weight coefficient to estimate the welding starting point.
[0073] Embodiment 2
[0074] Reference Figures 2-6For an embodiment of the present application, a three-corner joint welding starting point identification method is provided. In order to verify the beneficial effects of the present application, scientific demonstration is carried out through experiments.
[0075] In the data collection process, in addition to the first group of point clouds, subsequent point cloud data is sampled at a fixed distance of 2mm, 5mm, 8mm, 10mm, 12mm, 15mm, 18mm, and 20mm, respectively, as shown in the figure. By comparing the RMSE mean trend chart of different sampling intervals, it can be seen that when the interval distance reaches 10mm, the RMSE error mean of the starting point coordinates found is the smallest, only 0.34mm, so the sampling interval is selected as 10mm. Figure 5
[0076] A workpiece with dimensions of 200x200x50mm is used as the experimental object. First, sample every 10mm on the two welds, collect 20 groups of data for each weld, and a total of 40 groups of data as the overall sample. From the overall sample, randomly select 6, 8, 10, 12, and 14 groups of data for each of the two welds, and use the improved and unimproved RANSAC algorithms to repeatedly experiment 20 times on each group of point clouds to obtain the RMSE error mean of the welding starting point coordinates (x, y, z average), as shown in the figure. The experimental results show that as the number of samples increases, the error mean decreases, and when the number of samples reaches 10, the error mean tends to be stable. Figure 3
[0077] The RMSE error mean of the welding starting point coordinates fitted by the 2S-RANSAC algorithm at different sampling intervals is shown in the figure. Consistent with the qualitative derivation, the error first decreases rapidly and then increases slowly with the increase of the interval, and when the interval distance is about 10mm, the error mean is the smallest, and the maximum error is not more than 0.34mm. In summary, only 10 groups of data with an interval of 10mm on both sides of the welding starting point of the three-corner joint can complete the efficient and high-quality sampling step that can reflect the overall characteristics of the workpiece. Figure 4 In order to further verify the effectiveness of the improved 2S-RANSAC algorithm in this paper relative to the traditional RANSAC, refer to the environment in section 4.1, and use the two algorithms to repeatedly experiment 20 times to obtain the welding starting point coordinates, as shown in the figure. Obviously, the theoretical position of the welding starting point obtained by the 2S-RANSAC algorithm under different sampling methods is more concentrated, which proves that the improved 2S-RANSAC algorithm proposed in this paper has higher repetition accuracy compared with the traditional RANSAC algorithm.
[0078] Figure 6
[0079] It should be noted that the above-mentioned embodiments are only used to explain the technical solutions of the present application, but not to limit the present application. Although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or equivalent replaced without departing from the spirit and scope of the present application, and all modifications and equivalent replacements should be included in the scope of the claims of the present application.
[0080] Those skilled in the art can understand that the embodiments of the present application can be provided as a method, a system or a computer program product. Therefore, the present application can take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. In addition, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. The solutions in the embodiments of the present application can be implemented in various computer languages, such as object-oriented programming language Java and interpreted scripting language JavaScript.
[0081] The present application is described with reference to flowcharts and / or block diagrams according to the methods, devices (systems) and computer program products of the embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and the combination of the flows and / or blocks in the flowcharts and / or block diagrams can be implemented by computer program instructions. These computer program instructions can be provided to a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing devices to produce a machine, so that the instructions executed by the computer or other programmable data processing devices produce a device for implementing the functions specified in the flowcharts and / or block diagrams. Figure 1 The functions specified in one flow or multiple flows and / or blocks Figure 1 The functions specified in one flow or multiple flows and / or blocks
[0082] Although the preferred embodiments of the present application have been described, those skilled in the art can make additional changes and modifications to the embodiments once they know the basic inventive concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications falling within the scope of the present application.
[0083] Obviously, those skilled in the art can make various modifications and variations to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalent technologies, the present application is also intended to include these modifications and variations.
Claims
1. A method for identifying the welding start point of a three-sided fillet weld, characterized in that: include, Collect point cloud contour data of a section of a three-sided fillet weld; Three-point sampling is used to find a qualified straight line model; Point cloud merging optimization rules are adopted to reduce the impact of noise. Transform the coordinates of the weld corner points to the robot coordinate system and introduce weighting coefficients in the weld line fitting. Estimate the welding start point; The three-point sampling involves randomly selecting three sample points from the point cloud data, where the distance between any two sample points should be greater than D. Connecting the three sample points pairwise yields three straight-line models. The formula for calculating D is as follows: Among them, σ comprehensively represents the surface roughness of the weldment and the point cloud extraction error, which can be selected according to the actual situation; The three-point sampling also includes selecting a straight line model, calculating the distance between it and a sample point not on the model, dividing all point clouds into inliers and outliers of the straight line based on the distance, setting a reasonable error tolerance threshold t, and if the distance is less than the threshold t, the model pre-test passes; otherwise, the remaining models are calculated. If all three linear models fail the test, three sample points are selected again for three-point sampling modeling until the model pre-test passes. The number of in-place points is calculated for the models that pass the pre-test. If the preset number requirement is met, the linear model and the in-place points are retained. The point cloud merging optimization includes selecting the line model with the most local points as the master model, and calculating the angle between the remaining line models and the master model. If the angle is less than a threshold... Then, the point cloud data contained in the straight line model are merged, and the straight line model is re-estimated. The formula for calculating the angle threshold is expressed as: , Where h is the maximum allowable straightness of the workpiece, and w is the point cloud acquisition width of the camera on a single surface of the weldment; The introduction of the weighting coefficient includes numbering the weld corner points 1, 2, 3, ..., i according to their distance from the welding start point from near to far; and calculating the distance from the i-th point after introducing the weighting coefficient to the straight line model. The formula is expressed as: in, Let i be the weight of the corner point i, n be the number of inliers in the model, and A, B, and C be the coefficients of the straight line model. The estimation of the welding start point includes the fact that the two straight line models obtained in the final calculation do not intersect in space, so the midpoint of the common perpendicular of the two straight line models is taken as the weld start point.
2. The method for identifying the welding start point of a three-sided fillet weld as described in claim 1, characterized in that: The acquisition of the point cloud contour data includes using a line structured light camera to photograph a certain cross-sectional position of the three-sided corner welded part and repeatedly taking points. The first set of data at the starting point position is randomly taken at a distance of 10mm from the starting point, and subsequent point cloud data are taken at fixed distance intervals.
3. A welding start point identification system for three-sided fillet welds, employing the welding start point identification method for three-sided fillet welds as described in claim 1, characterized in that... include: The data acquisition module is used to collect point cloud contour data of the cross-section of the three-sided fillet weldment. The model pre-validation module, connected to the data acquisition module, is used to pre-validate the linear model. The model optimization module, connected to the model pre-validation module, is used to merge and optimize point clouds to reduce the impact of noise. The output module, connected to the model optimization module, is used to introduce weight coefficients to estimate the welding start point.
Citation Information
Patent Citations
CCD-based method and system for automatic identification and track planning of pipeline space weld seams
CN104400279A
Welding seam tracking control device and method based on cross laser device
CN107414253A