Two-dimensional welding seam starting point locating method based on laser tracking

Abstract

The invention belongs to the field of robot welding, and particularly relates to a two-dimensional welding seam starting point locating method based on laser tracking, which comprises the following steps: scanning a to-be-welded workpiece frame by frame along the extension direction of a welding seam by using a laser sensor; converting the laser stripes obtained by each time of scanning into point cloud data; inputting the laser stripes obtained by each time of scanning into a trained convolutional neural network classifier to identify the type of the weld groove, and selecting a median method or Hough transform to extract feature points of the weld groove according to the type of the weld groove and the converted point cloud data; when the number of the feature points of the weld groove reaches a set threshold value, the laser sensor is controlled to stop scanning the to-be-welded workpiece; fitting the extracted feature points of all the weld grooves into a straight line by using an orthogonal regression method based on random sampling consistency; and calculating according to the fitted straight line to obtain a starting point coordinate of the welding seam under the laser sensor coordinate system. According to the method, starting points of various two-dimensional welding seams and welding seams in different groove forms can be accurately, efficiently and automatically recognized, and the automatic welding efficiency of the welding robot is greatly improved.

Description

Technical Field

[0001] The invention belongs to the field of robot welding intelligent automation, and in particular relates to a two-dimensional weld starting point location finding method based on laser tracking. Background Art

[0002] In order to cope with the inevitable trend of high-quality and high-efficiency development of the manufacturing industry and the deepening reform of intelligent manufacturing, the development of intelligent and automated welding has become a key means to improve welding quality and efficiency. Among them, weld adaptive tracking is a key technology to achieve intelligent and automated welding. Nowadays, the welds of workpieces to be welded are in various forms such as straight welds, right-angle welds, S-shaped welds, curved welds, etc. Accurately sensing the starting point of the welds of V-shaped, I-shaped, Y-shaped, and lap-shaped weld grooves has become an important factor restricting the high-quality welding of workpieces. Therefore, it is urgent to break through the automatic positioning of the weld starting point and the adaptive tracking technology of the welding trajectory.

[0003] At present, for conventional industrial production scenarios, the main methods for using welding robots to locate the starting point of welds are: first, the teaching and reproduction working method is adopted. If the size or position of the workpiece to be welded is changed, re-teaching is required, so it has the disadvantages of high labor intensity and low work efficiency. Secondly, the starting point of the weld is located by designing tooling and fixtures. Due to clamping errors during the workpiece processing, problems such as inaccurate starting point positioning will also occur. Finally, a visual sensor such as a CCD camera is used to capture the weld image and find the starting point of the weld through relevant algorithms. However, due to the interactive form of welding work in the production environment, arc light, molten pool splashing, sparks and other phenomena will be generated, resulting in a variety of noises in the CCD camera imaging and low welding efficiency. Summary of the invention

[0004] To solve the above problems, the present invention provides a two-dimensional weld starting point positioning method based on laser tracking, which is applicable to all two-dimensional welds and groove types, is not affected by arc light, molten pool splashes, sparks, etc. generated by the interactive working environment, can be used in harsh working conditions and scenes that are out of reach of humans, and improves the efficiency, robustness and accuracy of welding.

[0005] The present invention provides a two-dimensional weld starting point location finding method based on laser tracking, comprising:

[0006] S1: Use a laser sensor to scan the workpiece to be welded frame by frame along the extension direction of the weld;

[0007] S2: Convert the laser stripes obtained from each scan into point cloud data;

[0008] S3: Input the laser stripes obtained from each scan into the trained convolutional neural network classifier to identify the type of weld groove, and select the median method or Hough transform to extract the characteristic points of the weld groove according to the type of weld groove and the converted point cloud data;

[0009] S4: When the number of characteristic points of the weld groove reaches a set threshold, the laser sensor is controlled to stop scanning the workpiece to be welded;

[0010] S5: The orthogonal regression method based on random sampling consistency is used to fit the extracted characteristic points of all weld grooves into straight lines; the starting point coordinates of the weld in the laser sensor coordinate system are calculated based on the fitted straight line.

[0011] Furthermore, the types of the weld groove include: continuous type and discontinuous type; for the continuous weld groove, Hough transform is used to extract the characteristic points of the weld groove; for the discontinuous weld groove, the median method is used to extract the characteristic points of the weld groove.

[0012] Furthermore, the method of extracting characteristic points of the weld groove by using the median method includes:

[0013] S311: Calculate the distance ΔI between any two adjacent points in the point cloud data i =X i -X i-1 ; Among them, X i is the coordinate value of the i-th point in the point cloud data, X i-1 is the coordinate value of the i-1th point in the point cloud data, i∈{2,3,…,N}, N represents the number of points in the point cloud data;

[0014] S312: Traverse all points in the point cloud data and define ΔI i The maximum value of ΔI Max , and let ΔI Max The corresponding two adjacent points are feature point C and feature point A;

[0015] S313: According to the coordinate values ​​of the characteristic points A and C, the characteristic point B of the weld groove is calculated by the median method. 终 ; Therefore, the characteristic point B of the weld groove 终 The coordinate value of (X B ,Z B )in

[0016] Furthermore, the method of extracting characteristic points of weld groove by using Hough transform includes:

[0017] S321: Take the point with the largest ordinate in the point cloud data as point B 始 ;

[0018] S322: The first point of the point cloud data is used as the left endpoint L, and the last point of the point cloud data is used as the right endpoint R; point R and point B 始 The connected straight line is used as the right reference line; calculate points R and B 始 The shortest distance between each point and the right reference straight line, and select the point with the largest distance as point C;

[0019] S323: Point L and point B 始 The connected straight line is used as the left reference line; calculate points L and B 始 The shortest distance between each point and the left reference straight line, and select the point with the largest distance as point A;

[0020] S324: Divide the point cloud data into LA segment and AB segment 始 Section B 始 C segment and CR segment, using Hough transform to transform AB 始 Duan He B 始 Segments C and AB are fitted as straight lines. 始 Line segment and B 始 The intersection of the straight line of segment C is the characteristic point B of the weld groove. 终 .

[0021] Furthermore, the method of fitting all extracted weld groove feature points into straight lines using the orthogonal regression method based on random sampling consistency includes: using the random sampling consistency algorithm to filter out some noise weld groove feature points; and using the orthogonal regression method to fit the remaining weld groove feature points into straight lines.

[0022] Furthermore, the characteristic points of the weld groove that are filtered out by using a random sampling consistency algorithm include:

[0023] S511: Randomly select two feature points from all the extracted weld groove feature points, connect the two feature points to form a straight line L; calculate the shortest distance between the remaining weld groove feature points and the straight line L. If the shortest distance is less than the set threshold d, define the point as an internal point and store it in data set B. jin Save in

[0024] S512: Determine whether the current number of iterations is greater than the maximum number of iterations m. If so, execute step S514; otherwise, execute step S513:

[0025] S513: Repeat steps S511-S512;

[0026] S514: Extract the data set B with the largest number of points in the iteration process jin As the optimal dataset B max , will not belong to data set B maxThe characteristic points of the weld groove are filtered out as noise points.

[0027] Furthermore, the maximum number of iterations m includes:

[0028]

[0029] Among them, η0 is the credibility, ε is the proportion of in-office points to all sample points, e=2 is the minimum number of points estimated by the model; log represents the natural logarithm.

[0030] Further, the method of fitting the remaining characteristic points of the weld groove into a straight line by using the orthogonal regression method includes:

[0031] S521: Define the remaining characteristic point set of the weld groove as P, the target straight line as z=a+bx, and construct the objective function with the minimum sum of the shortest distances from all points in the set P to the straight line z=a+bx as the optimization target Where n represents the number of points in the set P; X j and Z j represents the coordinates of point j in set P; b represents the slope of the line; a represents the intercept of the line;

[0032] S522: Objective function Find the partial derivative, that is, and And order:

[0033]

[0034] Obtain the intercept of the final fitting line Slope in,

[0035] Furthermore, the calculation of the starting point coordinates of the weld in the laser sensor coordinate system based on the fitted straight line includes: taking the X coordinate of the first feature point in the feature point set P as the X coordinate of the starting point, letting x equal the X value and bringing it into the fitted target straight line z=a+bx to obtain the Z coordinate of the starting point, then the two-dimensional coordinates of the starting point in the laser sensor coordinate system are (X, Z); and then calculating the Y coordinate of the starting point based on the number of scanning frames of the laser sensor when the feature point of the weld groove is extracted for the first time, to obtain the three-dimensional coordinates (X, Y, Z) of the starting point of the weld in the laser sensor coordinate system.

[0036] The present invention has at least the following beneficial effects:

[0037] The present invention uses a laser sensor to accurately scan the workpiece to be welded, and can efficiently obtain two-dimensional point cloud data of the workpiece to be welded; the type of weld groove is obtained by a pre-trained convolutional neural network classifier and the characteristic points of the weld groove are extracted; the straight line fitted by the orthogonal regression method of random sampling consistency ensures the accuracy and timeliness of extracting the starting point of the weld, and avoids problems such as inaccurate positioning of the starting point due to errors; the present invention obtains point cloud data by laser radar scanning, and can adapt to welds of different sizes and sizes for starting point positioning. The present invention not only improves the robustness and generalization of automatic positioning of the weld starting point, and no repeated teaching and reproduction work is required for different welds, but also overcomes the influence of arc light, molten pool splashing, sparks, etc. on starting point positioning through laser sensor scanning and positioning, and significantly improves the efficiency of automatic welding of welding robots, and is particularly suitable for two-dimensional welds and various weld grooves commonly used in engineering, and is not affected by harsh working conditions and industrial scenes that cannot be touched by humans. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 A schematic diagram of scanning a workpiece to be welded using a laser sensor;

[0039] Figure 2 This is a step diagram for locating the starting point of a two-dimensional weld by laser tracking;

[0040] Figure 3 A flowchart of the laser tracking process for finding the starting point of a two-dimensional weld;

[0041] Figure 4 Schematic diagram of extracting weld groove feature points for various weld grooves;

[0042] Figure 5 This is a schematic diagram of the result of filtering out noise points using the Ransac algorithm using matlab software;

[0043] Figure 6 This is a schematic diagram of the results of fitting a straight line through orthogonal regression;

[0044] Figure 7 This is a schematic diagram of the result of fitting a straight line using the least squares method;

[0045] Figure 8 It is a schematic diagram of the result of fitting a straight line by iterative weighted least squares method;

[0046] Fig. 9 Schematic diagram of eye in hand calibration. DETAILED DESCRIPTION

[0047] The following describes the embodiments of the present invention through specific examples. Those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the illustrations provided in the following embodiments only illustrate the basic concept of the present invention in a schematic manner. In the absence of conflict, the following embodiments and features in the embodiments can be combined with each other. It is particularly noted that the present invention is described in terms of the coordinate system marked in the accompanying drawings. The purpose is to facilitate the understanding of relevant technical personnel, and does not represent the only limitation of the present invention. Those skilled in the art can convert the coordinate system marked in the drawings of this specification according to relevant common sense knowledge.

[0048] Among them, the drawings are only used for illustrative explanations, and they only represent schematic diagrams rather than actual pictures, and should not be understood as limitations on the present invention. In order to better illustrate the embodiments of the present invention, some parts of the drawings may be omitted, enlarged or reduced, and do not represent the size of actual products. For those skilled in the art, it is understandable that some well-known structures and their descriptions in the drawings may be omitted.

[0049] The same or similar numbers in the drawings of the embodiments of the present invention correspond to the same or similar parts; in the description of the present invention, it should be understood that if the terms "upper", "lower", "left", "right", "front", "rear", etc. indicate the orientation or position relationship, they are based on the orientation or position relationship shown in the drawings, which is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operate in a specific orientation. Therefore, the terms describing the position relationship in the drawings are only used for illustrative purposes and cannot be understood as limiting the present invention. For ordinary technicians in this field, the specific meanings of the above terms can be understood according to specific circumstances.

[0050] See also Figures 1 to 9 The present invention provides a two-dimensional weld starting point location method based on laser tracking, comprising:

[0051] S1: Use a laser sensor to scan the workpiece to be welded frame by frame along the extension direction of the weld;

[0052] S2: Convert the laser stripes obtained from each scan into point cloud data;

[0053] S3: Input the laser stripes obtained from each scan into the trained convolutional neural network classifier to identify the type of weld groove, and select the median method or Hough transform to extract the characteristic points of the weld groove according to the type of weld groove and the converted point cloud data;

[0054] S4: When the number of characteristic points of the weld groove reaches a set threshold, the laser sensor is controlled to stop scanning the workpiece to be welded;

[0055] S5: The orthogonal regression method based on random sampling consistency is used to fit the extracted characteristic points of all weld grooves into straight lines; the starting point coordinates of the weld in the laser sensor coordinate system are calculated based on the fitted straight line.

[0056] This embodiment takes the weld type as S-type and the weld groove as a continuous V-shaped groove as an example. Figure 1 As shown, the robot welding gun 1 and the laser sensor 4 are fixed together through the connecting frame 3, and the laser sensor 4 is kept parallel to the robot welding gun 1 and located close to the workpiece to be welded. At the same time, the height of the laser sensor 4 and the workpiece to be welded is adjusted in advance and the laser sensor 4 is kept parallel to the workpiece to be welded to ensure that the scanning range of the laser sensor 4 includes the longitudinal distance of the entire workpiece to be welded. During the positioning process, the actuator drives the robot welding gun 1 and the laser sensor 4 to scan the workpiece to be welded frame by frame along the extension direction of the weld to ensure that the range of the laser stripes includes the longitudinal distance of the entire workpiece to be welded. Figure 3 This is a flowchart of the laser tracking process of finding the starting point of a two-dimensional weld. The entire process includes scanning, extracting feature point cloud data, filtering out noise points in the point cloud data, fitting a straight line to the data set after filtering out the noise points, and calculating the starting point of the weld.

[0057] In this embodiment, the laser sensor is used to accurately scan the workpiece to be welded, so that the two-dimensional point cloud data of the workpiece to be welded can be efficiently obtained; the type of weld groove is obtained by the pre-trained convolutional neural network classifier and the characteristic points of the weld groove are extracted; the straight line fitted by the orthogonal regression method of random sampling consistency ensures the accuracy and timeliness of extracting the starting point of the weld, and avoids the problem of inaccurate positioning of the starting point due to errors; the present invention obtains point cloud data by laser radar scanning, and can adapt to welds of different sizes and sizes for starting point positioning. The present invention not only improves the robustness and generalization of automatic positioning of the weld starting point, but also eliminates the need for repeated teaching and reproduction work for different welds. At the same time, the laser sensor scanning and positioning method overcomes the influence of arc light, molten pool splashing, sparks, etc. on starting point positioning, and significantly improves the efficiency of automatic welding of welding robots. It is particularly suitable for two-dimensional welds and various weld grooves commonly used in engineering and is not affected by harsh working conditions and industrial scenes that cannot be touched by humans.

[0058] In this embodiment, suitable methods are used to extract weld groove feature points for different weld groove types. Figure 4 As shown in (a), for weld grooves with corner or V-shaped point cloud data continuity, Hough transform is used to extract weld groove feature points; Figure 4As shown in (b), the median method is used to extract weld groove feature points for point cloud data discontinuities such as Y-type or I-type weld grooves. By introducing a pre-trained convolutional neural network classifier, different weld groove types can be accurately identified, laying a good foundation for subsequent weld groove feature point extraction.

[0059] Furthermore, the types of the weld groove include: continuous type and discontinuous type; for the continuous weld groove, Hough transform is used to extract the characteristic points of the weld groove; for the discontinuous weld groove, the median method is used to extract the characteristic points of the weld groove.

[0060] Furthermore, the method of extracting characteristic points of the weld groove by using the median method includes:

[0061] S311: Calculate the distance ΔI between any two adjacent points in the point cloud data i =X i -X i-1 ; Among them, X i is the coordinate value of the i-th point in the point cloud data, X i-1 is the coordinate value of the i-1th point in the point cloud data, i∈{2,3,…,N}, N represents the number of points in the point cloud data;

[0062] S312: Traverse all points in the point cloud data and define ΔI i The maximum value of ΔI Max , and let ΔI Max The corresponding two adjacent points are feature point C and feature point A;

[0063] S313: According to the coordinate values ​​of the characteristic points A and C, the characteristic point B of the weld groove is calculated by the median method. 终 ; Therefore, the characteristic point B of the weld groove 终 The coordinate value of (X B ,Z B )in

[0064] Furthermore, the method of extracting characteristic points of weld groove by using Hough transform includes:

[0065] S321: Take the point with the largest ordinate in the point cloud data as point B 始 ;

[0066] S322: The first point of the point cloud data is used as the left endpoint L, and the last point of the point cloud data is used as the right endpoint R; point R and point B 始 The connected straight line is used as the right reference line; calculate points R and B 始 The shortest distance between each point and the right reference straight line, and select the point with the largest distance as point C;

[0067] S323: Point L and point B 始 The connected straight line is used as the left reference line; calculate points L and B 始 The shortest distance between each point and the left reference straight line, and select the point with the largest distance as point A;

[0068] S324: Divide the point cloud data into LA segment and AB segment 始 Section B 始 C segment and CR segment, using Hough transform to transform AB 始 Duan He B 始 Segments C and AB are fitted as straight lines. 始 Line segment and B 始 The intersection of the straight line of segment C is the characteristic point B of the weld groove. 终 .

[0069] Specifically, the method of extracting weld groove feature points by using Hough transform includes:

[0070] Step S321: Figure 4 As shown in (a), the point with the largest ordinate in the point cloud data is taken as point B 始 ;

[0071] Step S322: Figure 4 As shown in (c), the first point of the point cloud data is taken as the left endpoint L, and the last point of the point cloud data is taken as the right endpoint R; point R and point B 始 The connected straight line is used as the right reference line; calculate points R and B 始 The shortest distance between each point and the right reference straight line, and select the point with the largest distance as point C;

[0072] Step S323: Point L and point B 始 The connected straight line is used as the left reference line; calculate points L and B 始 The shortest distance between each point and the left reference straight line, and select the point with the largest distance as point A;

[0073] Step S324: Figure 4 As shown in (d), the point cloud data is divided into LA segment and AB segment. 始 Section B 始 C segment and CR segment, using Hough transform to transform AB 始 Duan He B 始 Segments C and AB are fitted as straight lines. 始 Line segment and B 始 The intersection of the straight line of segment C is the characteristic point B of the weld groove. 终 .

[0074] Further, specifically, when the number of characteristic points of the weld groove reaches a set threshold, the step S4 controls the laser sensor to stop scanning the workpiece to be welded, specifically including:

[0075] A counter M is provided in the laser sensor, and the initial value of the counter M is 0;

[0076] Depend on Figure 1 , Figure 2 and Figure 3 It can be seen that:

[0077] Step S401: The laser sensor scans the workpiece to be welded along the extension direction of the weld.

[0078] Step S402: If the laser sensor collects point cloud data and successfully extracts the weld groove feature point, the counter M=M+1 and stores the feature point in point set B j Otherwise, return to step S401:

[0079] Step S403: Determine that the counter M is ≥ 30. If so, the laser sensor stops scanning the workpiece to be welded to obtain a point set B of the complete collection of weld groove feature points. j ; Otherwise return to step S401.

[0080] Furthermore, the method of fitting all extracted weld groove feature points into straight lines using the orthogonal regression method based on random sampling consistency includes: using the random sampling consistency algorithm to filter out some noise weld groove feature points; and using the orthogonal regression method to fit the remaining weld groove feature points into straight lines.

[0081] Furthermore, the characteristic points of the weld groove that are filtered out by using a random sampling consistency algorithm include:

[0082] S511: In the feature point set B of all weld grooves extracted j Randomly select two feature points from the dataset and connect them to form a straight line L. Calculate the shortest distance between the feature points of the remaining weld grooves and the straight line L. If the shortest distance is less than the set threshold d, the point is defined as an internal point and stored in the dataset B. jin Save in

[0083] S512: Determine whether the current number of iterations is greater than the maximum number of iterations m. If so, execute step S514; otherwise, execute step S513:

[0084] S513: Repeat steps S511-S512;

[0085] S514: Extract the data set B with the largest number of points in the iteration process jin As the optimal dataset B max , will not belong to data set Bmax The characteristic points of the weld groove are filtered out as noise points.

[0086] Furthermore, the maximum number of iterations m includes:

[0087]

[0088] Among them, η0 is the credibility (generally 0.95-0.99), ε is the proportion of in-office points to all sample points, e=2 is the minimum number of points estimated by the model; log represents the natural logarithm. Figure 5 This is a schematic diagram of the results of filtering out noise points using the Ransac algorithm through MATLAB software. The results show that the Ransac algorithm is effective and successfully filters out noise outliers.

[0089] Further, the method of fitting the remaining characteristic points of the weld groove into a straight line by using the orthogonal regression method includes:

[0090] S521: Define the remaining characteristic point set P of the weld groove, fit the target straight line z=a+bx, and construct the objective function with the minimum sum of the shortest distances from all points in the set P to the straight line z=a+bx as the optimization target Where n represents the number of points in the set P; X j and Z j represents the coordinates of point j in set P; b represents the slope of the line; a represents the intercept of the line;

[0091] S522: Objective function Find the partial derivative, that is, and And order:

[0092]

[0093] Obtain the intercept of the final fitting line Slope in,

[0094] Figure 6 , Figure 7 and Figure 8 This is a schematic diagram of the results of fitting the remaining weld groove feature point set P using different common straight line fitting methods in engineering through matlab. After calculating the sum of square errors (SSE) and root mean square error (RMSE), the final comparison concluded that the orthogonal regression method of fitting straight lines has smaller sum of square errors (SSE) and root mean square error (RMSE) than other methods, and the fitting speed is comparable to the least squares, and is faster than the iterative weighted least squares, so the fitting goodness of fit is the highest and the fitting accuracy is good, which lays a good foundation for the subsequent calculation of the weld starting point.

[0095] Furthermore, the calculation of the starting point coordinates of the weld in the laser sensor coordinate system based on the fitted straight line includes: taking the X coordinate of the first feature point in the feature point set P as the X coordinate of the weld starting point, letting x equal the X value and bringing it into the fitted target straight line z=a+bx to obtain the Z coordinate of the weld starting point, then the two-dimensional coordinates of the weld starting point in the laser sensor coordinate system are (X, Z); then calculating the Y coordinate of the weld starting point based on the number of scanning frames of the laser sensor when the feature point of the weld groove is extracted for the first time, and obtaining the three-dimensional coordinates (X, Y, Z) of the weld starting point in the laser sensor coordinate system.

[0096] Furthermore, if Fig. 9 The hand-eye calibration shown in the figure adopts the eye-in-hand method. The robot base coordinate system is {B}, the robot tool coordinate system is {T}, the laser sensor coordinate system is {C}, and the coordinate system of the center of the robot end flange is {E}. The transformation matrix of the robot tool coordinate system relative to the robot end flange center coordinate system is: The transformation matrix between the laser sensor coordinate system and the robot tool coordinate system is: The transformation matrix between the robot end flange center coordinate system and the robot base coordinate system is: Assume that the coordinates of any point P in the space in the laser sensor coordinate system are P( c X c , Y c , Z c ), the coordinates in the robot base coordinate system are P B (X B , Y B , Z B ), therefore, the following equation is constructed based on the equal position of any point P in the space in the same coordinate system: The coordinate value of the weld starting point in the laser sensor coordinate system is transformed into the coordinate value of the robot base coordinate system. Finally, the welding robot is controlled to move to the weld starting point to ensure the stability and consistency of the welding process and complete the adaptive search for the weld starting point and positioning work.

[0097] In summary, the present invention uses a laser sensor to accurately scan the workpiece to be welded, and can efficiently obtain the two-dimensional point cloud data of the workpiece to be welded; the type of weld groove is obtained by a pre-trained convolutional neural network classifier and the characteristic points of the weld groove are extracted; the straight line fitted by the orthogonal regression method of random sampling consistency ensures the accuracy and timeliness of extracting the starting point of the weld, and avoids the problem of inaccurate positioning of the starting point due to errors; the present invention obtains point cloud data by laser radar scanning, and can adapt to welds of different sizes and sizes for starting point positioning. The present invention not only improves the robustness and generalization of automatic positioning of the weld starting point, but also eliminates the need for repeated teaching and reproduction work for different welds. At the same time, the laser sensor scanning and positioning method overcomes the influence of arc light, molten pool splashing, sparks, etc. on the starting point positioning, and significantly improves the efficiency of automatic welding of welding robots. It is particularly suitable for two-dimensional welds and various weld grooves commonly used in engineering, and is not affected by harsh working conditions and industrial scenes that cannot be touched by humans.

[0098] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the purpose and scope of the technical solution, which should be included in the scope of the claims of the present invention.

Claims

1. A two-dimensional weld starting point location method based on laser tracking, characterized in that: include: S1: Use a laser sensor to scan the workpiece to be welded frame by frame along the extension direction of the weld; S2: Convert the laser stripes obtained from each scan into point cloud data; S3: Input the laser stripes obtained from each scan into the trained convolutional neural network classifier to identify the type of weld groove, and select the median method or Hough transform to extract the characteristic points of the weld groove according to the type of weld groove and the converted point cloud data; S4: When the number of characteristic points of the weld groove reaches a set threshold, the laser sensor is controlled to stop scanning the workpiece to be welded; S5: The orthogonal regression method based on random sampling consistency is used to fit the extracted characteristic points of all weld grooves into straight lines; the starting point coordinates of the weld in the laser sensor coordinate system are calculated based on the fitted straight line.

2. A two-dimensional weld starting point location method based on laser tracking according to claim 1, characterized in that: The types of weld grooves include: continuous type and discontinuous type; for the continuous weld groove, the characteristic points of the weld groove are extracted by using Hough transform; for the discontinuous weld groove, the characteristic points of the weld groove are extracted by using median method.

3. The method for locating the starting point of a two-dimensional weld based on laser tracking according to claim 2, characterized in that: The method of extracting the characteristic points of the weld groove by using Hough transform includes: S321: Take the point with the largest ordinate in the point cloud data as point B 始 ; S322: The first point of the point cloud data is used as the left endpoint L, and the last point of the point cloud data is used as the right endpoint R; point R and point B 始 The connected straight line is used as the right reference line; calculate points R and B 始 The shortest distance between each point and the right reference straight line, and select the point with the largest distance as point C; S323: Point L and point B 始 The connected straight line is used as the left reference line; calculate points L and B 始 The shortest distance between each point and the left reference straight line, and select the point with the largest distance as point A; S324: Divide the point cloud data into LA segment and AB segment 始 Section B 始 C segment and CR segment, using Hough transform to transform AB 始 Duan He B 始 Segments C and AB are fitted as straight lines. 始 Line segment and B 始 The intersection of the straight line of segment C is the characteristic point B of the weld groove. 终 .

4. A two-dimensional weld starting point location method based on laser tracking according to claim 3, characterized in that: The method of fitting all extracted weld groove feature points into straight lines by using the orthogonal regression method based on random sampling consistency includes: filtering out some noise weld groove feature points by using the random sampling consistency algorithm; and fitting the remaining weld groove feature points into straight lines by using the orthogonal regression method.

5. The method for locating the starting point of a two-dimensional weld based on laser tracking according to claim 4, characterized in that: The characteristic points of the weld grooves that are filtered out by using a random sampling consistency algorithm include: S511: Randomly select two feature points from all the extracted weld groove feature points, connect the two feature points to form a straight line L; calculate the shortest distance between the remaining weld groove feature points and the straight line L. If the shortest distance is less than the set threshold d, define the point as an internal point and store it in data set B. jin Save in S512: Determine whether the current number of iterations is greater than the maximum number of iterations m. If so, execute step S514; otherwise, execute step S513: S513: Repeat steps S511-S512; S514: Extract the data set B with the largest number of points in the iteration process jin As the optimal dataset B max , will not belong to data set B max The characteristic points of the weld groove are filtered out as noise points.

6. A two-dimensional weld starting point location method based on laser tracking according to claim 5, characterized in that: The maximum number of iterations m includes: Among them, η0 is the credibility, ε is the proportion of in-office points to all sample points, e=2 is the minimum number of points estimated by the model; log represents the natural logarithm.

7. A two-dimensional weld starting point location method based on laser tracking according to claim 6, characterized in that: The method of fitting the remaining characteristic points of the weld groove into a straight line by using the orthogonal regression method comprises: S521: Define the remaining characteristic point set of the weld groove as P, the target straight line as z=a+bx, and construct the objective function with the minimum sum of the shortest distances from all points in the set P to the straight line z=a+bx as the optimization target Where n represents the number of points in the set P; X j and Z j represents the coordinates of point j in set P; b represents the slope of the line; a represents the intercept of the line; S522: Objective function Find the partial derivative, that is, and And order: Obtain the intercept of the final fitting line Slope in, 8. The method for locating the starting point of a two-dimensional weld based on laser tracking according to claim 7, characterized in that: The method of calculating the starting point coordinates of the weld in the laser sensor coordinate system according to the fitted straight line includes: taking the X coordinate of the first feature point in the feature point set P as the X coordinate of the starting point, substituting x equal to the X value into the fitted target straight line z=a+bx to obtain the Z coordinate of the starting point, and then the two-dimensional coordinates of the starting point in the laser sensor coordinate system are (X, Z); and then calculating the Y coordinate of the starting point according to the number of scanning frames of the laser sensor when the feature point of the weld groove is extracted for the first time, and obtaining the three-dimensional coordinates (X, Y, Z) of the starting point of the weld in the laser sensor coordinate system.

Images