Point cloud data processing method for finned tube circumferential weld and laser welding equipment

Abstract

The invention discloses a finned tube circumferential weld point cloud data processing method and laser welding equipment. The finned tube circumferential weld point cloud data processing method comprises the following steps that a visual sensor is connected with a robot; the method comprises the following steps: correspondingly moving a visual sensor to an upper area, an outer side area and a lower area of an annular welding seam step by step through a robot, correspondingly recording and storing robot pose data in a current state, and correspondingly acquiring upper original point cloud coordinates, outer original point cloud coordinates and lower original point cloud coordinates of the welding seam through the visual sensor; according to the pose data of the robot, mapping the upper original point cloud coordinates, the outer original point cloud coordinates and the lower original point cloud coordinates to a base coordinate system to obtain lower mapping point cloud coordinates; and importing the upper mapping point cloud coordinates, the outer side mapping point cloud coordinates and the lower mapping point cloud coordinates into a base coordinate system to obtain a point cloud model of the welding seam. According to the invention, the problem that the shapes of the finned tube and the U-shaped tube block the sight line of the visual sensor can be solved, the data acquisition efficiency is improved, and meanwhile, the point cloud coordinates can be directly associated with the position information in the actual space, so that the data processing efficiency is greatly improved.

Description

Technical Field

[0001] The invention belongs to the technical field of laser processing, and in particular to a point cloud data processing method for a fin tube annular weld and laser welding equipment. Background Art

[0002] The core heat exchange device of the air temperature evaporator, such as Figure 1 As shown, its main structure includes: a fin tube 1 and a U-shaped tube 2 made of aluminum alloy, the fin tube 1 is sleeved on the end of the U-shaped tube 2, and then welded by laser welding or other methods to form an annular weld 3 at the end of the fin tube 1. In the process of laser welding the fin tube 1, it is necessary to accurately collect the point cloud information of the weld 3, and construct a point cloud model of the weld 3 according to the point cloud information, and guide the laser welding equipment to complete the welding through the point cloud model of the weld 3, so as to improve the automation degree of laser welding of the fin tube 1 and ensure the quality of the weld.

[0003] Although weld 3 is a complete annular weld, the difficulty of collecting its point cloud information and constructing its point cloud model is much higher than that of a conventional annular weld. The main reasons are:

[0004] 1. Since the fin tube 1 and the U-shaped tube 2 are sleeved and then laser welded, when the sensor collects the point cloud data of the weld 3 along a certain direction, the volume and shape of the fin tube 1 and the U-shaped tube 2 will block the sensor's line of sight, resulting in the sensor being able to only collect the morphology of the side the sensor's line of sight is facing, and unable to collect the back side image of the weld 3; when the sensor circles the fin tube 3 to collect point cloud data, due to the spatial position interference between the geometric shape of the U-shaped tube 2 and the sensor, the sensor collides with the U-shaped tube 2 during the circle, and still cannot construct a complete three-dimensional morphology of the weld;

[0005] At the same time, when collecting point cloud data, the visual sensor needs to move at a uniform speed, and simply collects data continuously on a time scale based on the principle of triangulation, and then superimposes the collected local point cloud data to obtain a complete point cloud image. However, during the collection process, the movement speed of the sensor at each moment cannot be guaranteed to be always the same, resulting in deviations in the point cloud data obtained at each angle, causing the final superimposed point cloud image to be distorted. The reason is that the collected local point cloud data cannot be directly associated with the position information in the actual space;

[0006] 2. In the prior art, the local point cloud data of the weld 3 is first collected, and the coordinate system is performed, and then the data is spliced. However, when performing the coordinate system, it is necessary to set a number of marking points in advance, and then perform point-to-point matching based on the marking points to construct the three-dimensional morphology of the weld. This process is computationally huge, time-consuming, and it is difficult to ensure the modeling accuracy of the point cloud model. Summary of the invention

[0007] In order to solve the above problems, the present invention proposes a point cloud data processing method and laser welding equipment for a fin tube annular weld, which can solve the problem of the fin tube and U-tube's own shape blocking the visual sensor's line of sight, improve data acquisition efficiency, and directly associate the point cloud coordinates with the position information in the actual space, thereby greatly improving data processing efficiency.

[0008] To achieve the above object, the technical solution adopted by the present invention is:

[0009] On the one hand, a method for processing point cloud data of a fin tube annular weld is provided, which comprises the following steps:

[0010] Connecting vision sensors to robots;

[0011] The robot moves the visual sensor to the upper area, the outer area and the lower area of ​​the annular weld in steps, and records and saves the robot posture data in the current state, and then obtains the upper original point cloud coordinates, the outer original point cloud coordinates and the lower original point cloud coordinates of the weld through the visual sensor;

[0012] According to the robot posture data, the upper original point cloud coordinates, the outer original point cloud coordinates, and the lower original point cloud coordinates are mapped to the base coordinate system O B -X B Y B Z B Next, to obtain the coordinates of the lower mapping point cloud;

[0013] Import the upper mapping point cloud coordinates, the outer mapping point cloud coordinates, and the lower mapping point cloud coordinates into the base coordinate system O B -X B Y B Z B to obtain the point cloud model of the weld.

[0014] On the other hand, a laser welding device is provided, which includes a laser welding head, a weld point cloud model generation unit and a control unit, and the weld point cloud model generation unit is used to implement the above-mentioned point cloud data processing method, and the control unit controls the movement of the laser welding head based on the obtained weld point cloud model to complete the welding of fin tubes and U-shaped tubes.

[0015] The beneficial effects of using the present invention are:

[0016] The present invention scans the weld in steps from three angles to obtain the local original point cloud coordinates of the corresponding parts, thereby effectively solving the problem of the fin tube and the U-shaped tube blocking the visual sensor's sight line, and avoiding the visual sensor from colliding and interfering with the U-shaped tube and the fin tube during movement, thereby improving safety. At the same time, the visual sensor does not need to move at a constant speed all the time, so that the total time for collecting the original point cloud coordinates is greatly shortened, thereby improving data collection efficiency.

[0017] At the same time, the present invention can quickly map the original point cloud coordinates of any angle acquired by the sensor in the camera coordinate system to the base coordinate system, thereby directly linking the point cloud coordinates with the position information in the actual space. There is no need to perform subsequent data processing steps such as coordinate conversion, point cloud registration, and overlay splicing, and the real weld point cloud model can be directly obtained. While greatly improving the data processing efficiency, the problem of point cloud image distortion caused by steps such as coordinate conversion, point cloud registration, and overlay splicing is avoided. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 It is a structural schematic diagram of a heat exchange device of an air-temperature evaporator in the prior art;

[0019] Figure 2 It is a schematic diagram of the steps of the point cloud data processing method of the fin tube annular weld in the present invention;

[0020] Figure 3 It is a schematic diagram of the state in which the visual sensor in the present invention is in the upper area of ​​the annular weld to obtain the coordinates of the upper original point cloud of the weld;

[0021] Figure 4 It is a schematic diagram of the state of the visual sensor in the present invention in the outer area of ​​the annular weld to obtain the original point cloud coordinates of the outer side of the weld;

[0022] Figure 5 It is a schematic diagram of the state in which the visual sensor in the present invention is in the upper area of ​​the annular weld to obtain the original point cloud coordinates of the lower part of the weld;

[0023] Figure 6 Schematic diagram of the positional relationship of each coordinate system in the present invention;

[0024] Figure 7 It is the weld point cloud model obtained by the point cloud data processing method of the present invention. DETAILED DESCRIPTION

[0025] In order to make the purpose, technical solution and advantages of the technical solution more clear, the technical solution is further described in detail below in conjunction with specific implementation methods. It should be understood that these descriptions are only exemplary and are not intended to limit the scope of the technical solution.

[0026] Embodiment 1:

[0027] This embodiment provides a method for processing point cloud data of a fin tube annular weld, which includes the following steps:

[0028] S1, connecting the visual sensor 100 to the flange of the robot, so that the visual sensor 100 can be moved by the robot;

[0029] The robot moves the visual sensor 100 to the upper area of ​​the annular weld 3, and makes it located in the upper area of ​​the fin tube 1 at the same time, and keeps it still, records and saves the robot posture data in the current state, and then obtains the original point cloud coordinates of the upper part of the weld 3 through the visual sensor 100;

[0030] And according to the robot posture data, the original point cloud coordinates of the upper part of the weld 3 are mapped to the base coordinate system O B -X B Y B Z B Down to obtain the coordinates of the upper mapping point cloud;

[0031] In this embodiment, the visual sensor 100 includes a line laser visual sensor and can generate a fan-shaped laser light plane 101. Before obtaining the original point cloud coordinates of the upper part of the weld 3, the laser light plane 101 is made perpendicular to the central axis of the annular weld 3.

[0032] S2, using the robot to move the visual sensor 100 to the outer area of ​​the annular weld 3, and make it located at the outer area of ​​the fin tube 1 at the same time, and keep it still, record and save the robot posture data in the current state, and then obtain the original point cloud coordinates of the outer side of the weld 3 through the visual sensor 100;

[0033] And according to the robot posture data, the original point cloud coordinates outside the weld 3 are mapped to the base coordinate system O B -X B Y B Z B Next, to obtain the outer side mapping point cloud coordinates;

[0034] Similarly, before obtaining the original point cloud coordinates of the outer side of the weld 3, the laser light plane 101 is made perpendicular to the central axis of the annular weld 3;

[0035] In this embodiment, the area between the two free ends of the U-shaped tube 2 is "inside", and the direction opposite thereto is "outside";

[0036] S3, using the robot to move the visual sensor 100 to the area below the annular weld 3, and make it located at the area below the fin tube 1 at the same time, and keep it still, record and save the robot posture data in the current state, and then obtain the original point cloud coordinates of the lower part of the weld 3 through the visual sensor 100;

[0037] And according to the robot posture data, the original point cloud coordinates of the lower part of the weld 3 are mapped to the base coordinate system O B -X B Y B Z B Next, to obtain the coordinates of the lower mapping point cloud;

[0038] Similarly, before obtaining the original point cloud coordinates of the lower part of the weld 3, the laser light plane 101 is made perpendicular to the central axis of the annular weld 3;

[0039] It should be noted that, in this embodiment, during the position conversion process from the upper area to the outer area and from the outer area to the lower area, the visual sensor 100 does not collect the original point cloud coordinates before reaching the upper area, the outer area, and the lower area of ​​the weld 3;

[0040] S4, import the upper mapping point cloud coordinates, the outer mapping point cloud coordinates, and the lower mapping point cloud coordinates into the base coordinate system O B -X B Y B Z B to obtain the point cloud model of weld 3.

[0041] Furthermore, in this embodiment, the posture data of the robot includes X, Y, Z, A, B, and C, where X, Y, and Z are the origin of the flange coordinate system O, E Relative to the origin of the base coordinate system O B , displacement in the X, Y, and Z directions; A, B, and C represent the flange coordinate system O E -X E Y E Z E Rotate to the base coordinate system O B -X B Y B Z B When the initial orientation is the same, the flange coordinate system O E -X E Y E Z E Around Z E Axis, Y E Axis, X EThe axis rotates at angles A, B, and C, and X, Y, Z, A, B, and C can all be obtained through the robot's built-in sensors and stored in the Euler angle data {X, Y, Z, A, B, C} format to facilitate subsequent matrix conversion calculations.

[0042] Furthermore, in this embodiment, the coordinate system where the visual sensor 100 is located is the camera coordinate system. C -X C Y C Z C The original point cloud coordinates of the visual sensor 100, the original point cloud coordinates of the outer side, and the original point cloud coordinates of the lower part are all in the camera coordinate system O C -X C Y C Z C On this basis, according to the robot posture data, the original point cloud coordinates of the upper part of the weld 3, the original point cloud coordinates of the outer side, and the original point cloud coordinates of the lower part are mapped to the base coordinate system O B -X B Y B Z B The method for obtaining the corresponding mapping point cloud coordinates is the same, and specifically includes the following steps:

[0043] According to formula (1), the flange coordinate system O is obtained E -X E Y E Z E , base coordinate system O B -X B Y B Z B The coordinate transformation matrix between

[0044]

[0045] in, O is the origin of the flange coordinate system E Relative to the base coordinate system O B -X B Y B Z B The position vector in To transform the flange coordinate system O E -X E Y E Z E Rotate to the base coordinate system O B -X B Y B Z B The rotation matrix of the same orientation, and That is c represents cos; s represents sin;

[0046] Specifically, the rotation matrix The derivation process is as follows:

[0047] Since the flange coordinate system O E -X E Y E Z E Around Z E Axis, Y E Axis, X E Axis rotation, according to the principle of "from left to right" to describe the corresponding rotation matrix, we can get the following expression:

[0048]

[0049] The above expression can be used as an operator for calculation, and after calculating the product of R(Z,A)R(Y,B)R(X,C), we can get:

[0050]

[0051] The flange coordinate system O is obtained by using the robot four-point method and tool calibration. E -X E Y E Z E With tool coordinate system O T -X T Y T Z T The coordinate transformation matrix between Among them, the robot four-point calibration method is an existing technology and will not be described in detail here;

[0052] Furthermore, according to formulas (2)-(3), the tool coordinate system O is obtained: T -X T Y T Z T With the camera coordinate system O C -X C Y C Z C The coordinate transformation matrix between

[0053]

[0054] Among them, P T is the tool coordinate system O T -X T Y T Z T The coordinates of a point are described below, that is P C is the camera coordinate system O C -XC Y C Z C The following is the coordinate description of the same point, that is, To transform the camera coordinate system O C -X C Y C Z C Rotate to the tool coordinate system O T -X T Y T Z T The rotation matrix of the same orientation, O is the origin of the camera coordinate system C Relative to the tool coordinate system O T -X T Y T Z T The position vector in .

[0055] When deriving, substitute formula (3) into formula (2) and replace P T , P C Expand to get formula (4):

[0056]

[0057] Further matrix product operation is performed on formula (4), and formula (5) can be obtained:

[0058]

[0059] Therefore, there are 12 unknowns in formula (5) (i.e., r1-r9 and v1-v3). Formula (5) can be further expanded to obtain the following equation:

[0060]

[0061] It can be seen that the coordinates of the same point in the tool coordinate system O T -X T Y T Z T With the camera coordinate system O C -X C Y C Z C The corresponding relationship in can obtain the above three sets of equations, from which three unknowns can be solved. Therefore, through at least 4 sets of similar corresponding relationships (each set of corresponding relationships obtains three sets of equations and solves 3 unknowns), all 12 unknowns can be solved, and the tool coordinate system O can be determined. T -X T Y T Z T With the camera coordinate system O C -X C Y CZ C The coordinate transformation matrix between

[0062] Transform the coordinate matrix Perform matrix product operation to obtain the camera coordinate system O C -X C Y C Z C With the base coordinate system O B -X B Y B Z B The coordinate transformation matrix between

[0063] The camera coordinate system O is converted into C -X C Y C Z C The upper original point cloud coordinates / outer original point cloud coordinates / lower original point cloud coordinates are converted to the base coordinate system O B -X B Y B Z B Next, the original point cloud coordinates are mapped to the base coordinate system O B -X B Y B Z B To obtain the corresponding mapping point cloud coordinates:

[0064]

[0065] Among them, x C ,y C 、z C is the original point cloud coordinate (i.e., any one of the upper original point cloud coordinate, the outer original point cloud coordinate, and the lower original point cloud coordinate); x B ,y B 、z B Map the original point cloud coordinates to the base coordinate system O B -X B Y B Z B The mapping point cloud coordinates obtained after downloading (i.e., the upper mapping point cloud coordinates, the outer mapping point cloud coordinates, and the lower mapping point cloud coordinates).

[0066] On this basis, the upper mapping point cloud coordinates, the outer mapping point cloud coordinates, and the lower mapping point cloud coordinates obtained after conversion are the real coordinates of the weld 3 in the base coordinate system. Therefore, in step S4, the upper mapping point cloud coordinates, the outer mapping point cloud coordinates, and the lower mapping point cloud coordinates are directly imported into the base coordinate system to obtain the point cloud model of the weld 3 (such as Figure 7 shown).

[0067] It should be noted that, although in this embodiment only the upper original point cloud coordinates, the outer original point cloud coordinates, and the lower original point cloud coordinates (excluding the inner original point cloud coordinates) are obtained by sensor 100 and converted to the base coordinate system accordingly, it already contains most of the information of the weld 3. Therefore, in step S4, the upper mapped point cloud coordinates, the outer mapped point cloud coordinates, and the lower mapped point cloud coordinates are directly imported into the base coordinate system, and after fitting, a partial point cloud model of the weld 3 can be obtained. This model is sufficient to guide the laser to complete the welding of the complete weld 3 to obtain a complete and accurate weld 3.

[0068] Therefore, in this embodiment, the visual sensor scans the weld in steps from three angles to obtain the local original point cloud coordinates of the corresponding parts, and the scanning positions of the visual sensor are not on the inner side of the U-shaped tube, so the problem of the fin tube and the U-shaped tube's own shape blocking the visual sensor's sight line can be effectively solved, and the visual sensor can be prevented from colliding and interfering with the U-shaped tube and the fin tube during movement, thereby improving safety. At the same time, the visual sensor starts to collect point cloud data only after it moves to a predetermined position (for example, the upper area, the outer area, and the lower area of ​​the weld 3), and no data collection is performed during the position change process. Therefore, the collection position change of the visual sensor can be completed quickly in a short time, and there is no need for the visual sensor to always move at a uniform speed (in the prior art, the visual sensor needs to move at a uniform speed and continuously in a certain direction, and the point cloud data of the weld 3 needs to be continuously collected during the movement, and the total data collection time is relatively long), so that the total collection time of the original point cloud coordinates is greatly shortened to improve data collection efficiency;

[0069] In addition, according to the above-mentioned coordinate conversion relationship, the present embodiment can quickly map the original point cloud coordinates of any angle acquired by the sensor in the camera coordinate system to the base coordinate system, and the coordinates mapped to the base coordinate system (i.e., the mapped point cloud coordinates) are the real coordinates of the weld in the base coordinate system. In this way, the point cloud coordinates can be directly associated with the position information in the actual space, and there is no need to perform subsequent data processing steps such as coordinate conversion, point cloud registration, and overlay splicing. The real weld point cloud model can be directly obtained, which greatly improves the data processing efficiency while avoiding the problem of point cloud image distortion caused by steps such as coordinate conversion, point cloud registration, and overlay splicing.

[0070] Embodiment 2:

[0071] The present embodiment provides a laser welding device, which includes a laser welding head, a weld point cloud model generation unit and a control unit, and the weld point cloud model generation unit can implement the point cloud data processing method described in Example 1. The control unit controls the movement of the laser welding head based on the obtained weld point cloud model to complete the welding of the fin tube 1 and the U-shaped tube 2.

[0072] In summary, in the present invention, the visual sensor scans the weld in steps from three angles to obtain the local original point cloud coordinates of the corresponding parts, thereby effectively solving the problem of the fin tube and the U-shaped tube blocking the visual sensor's sight line, and avoiding the visual sensor from colliding and interfering with the U-shaped tube and the fin tube during movement, thereby improving safety. At the same time, there is no need for the visual sensor to always move at a constant speed, which greatly shortens the total time for collecting the original point cloud coordinates, thereby improving data collection efficiency.

[0073] At the same time, the present invention can quickly map the original point cloud coordinates of any angle acquired by the sensor in the camera coordinate system to the base coordinate system, thereby directly linking the point cloud coordinates with the position information in the actual space. There is no need to perform subsequent data processing steps such as coordinate conversion, point cloud registration, and overlay splicing, and the real weld point cloud model can be directly obtained. While greatly improving the data processing efficiency, the problem of point cloud image distortion caused by steps such as coordinate conversion, point cloud registration, and overlay splicing is avoided.

[0074] The above contents are only preferred embodiments of the present invention. For ordinary technicians in this field, many changes can be made in the specific implementation methods and application scopes based on the ideas of the present technical content. As long as these changes do not deviate from the concept of the present invention, they all fall within the scope of protection of this patent.

Claims

1. A point cloud data processing method for a fin tube annular weld, characterized in that: The steps include: Connecting vision sensors to robots; The robot moves the visual sensor to the upper area, the outer area and the lower area of ​​the annular weld in steps, and records and saves the robot posture data in the current state, and then obtains the upper original point cloud coordinates, the outer original point cloud coordinates and the lower original point cloud coordinates of the weld through the visual sensor; According to the robot posture data, the upper original point cloud coordinates, the outer original point cloud coordinates, and the lower original point cloud coordinates are mapped to the base coordinate system O B -X B Y B Z B Next, to obtain the coordinates of the lower mapping point cloud; Import the upper mapping point cloud coordinates, the outer mapping point cloud coordinates, and the lower mapping point cloud coordinates into the base coordinate system O B -X B Y B Z B to obtain the point cloud model of the weld.

2. The point cloud data processing method according to claim 1, characterized in that: The vision sensor includes a line laser vision sensor.

3. The point cloud data processing method according to claim 2, characterized in that: Before obtaining the original point cloud coordinates of the upper part of the weld, the original point cloud coordinates of the outer part, and the original point cloud coordinates of the lower part, the laser light plane of the line laser vision sensor is made perpendicular to the central axis of the annular weld.

4. The point cloud data processing method according to claim 1, characterized in that: During the position conversion process from the upper area to the outer area and from the outer area to the lower area, the original point cloud coordinates are not obtained before reaching the upper area, the outer area and the lower area of ​​the weld.

5. The point cloud data processing method according to claim 1, characterized in that: According to the robot posture data, the upper original point cloud coordinates, the outer original point cloud coordinates, and the lower original point cloud coordinates are mapped to the base coordinate system O B -X B Y B Z B To obtain the corresponding mapping point cloud coordinates, the following steps are included: Get flange coordinate system O E -X E Y E Z E With the base coordinate system O B -X B Y B Z B The coordinate transformation matrix between Flange coordinate system O E -X E Y E Z E With tool coordinate system O T -X T Y T Z T The coordinate transformation matrix between Tool coordinate system O T -X T Y T Z T With the camera coordinate system O C -X C Y C Z C The coordinate transformation matrix between Transform the coordinate matrix Perform matrix product operation to obtain the camera coordinate system O C -X C Y C Z C With the base coordinate system O B -X B Y B Z B The coordinate transformation matrix between According to the camera coordinate system O C -X C Y C Z C With the base coordinate system O B -X B Y B Z B The coordinate transformation matrix between Map the coordinates of each original point cloud to the base coordinate system O B -X B Y B Z B to obtain the corresponding mapping point cloud coordinates.

6. The point cloud data processing method according to claim 5, characterized in that: According to formula (1), the flange coordinate system O is obtained E -X E Y E Z E , base coordinate system O B -X B Y B Z B The coordinate transformation matrix between in, O is the origin of the flange coordinate system E Relative to the base coordinate system O B -X B Y B Z B The position vector in , and To transform the flange coordinate system O E -X E Y E Z E Rotate to the base coordinate system O B -X B Y B Z B The rotation matrix of the same orientation, and That is c represents cos; s represents sin; X, Y, and Z are the origin of the flange coordinate system O E Relative to the origin of the base coordinate system O B , displacement in the X, Y, and Z directions; A, B, and C represent the flange coordinate system O E -X E Y E Z E Rotate to the base coordinate system O B -X B Y B Z B When the initial orientation is the same, the flange coordinate system O E -X E Y E Z E Around Z E Axis, Y E Axis, X E The axis rotates by angles A, B, and C.

7. The point cloud data processing method according to claim 5, characterized in that: Obtain flange coordinate system O according to the robot four-point method E -X E Y E Z E With tool coordinate system O T -X T Y T Z T The coordinate transformation matrix between 8. The point cloud data processing method according to claim 5, characterized in that: According to formula (2)-(3), the tool coordinate system O is obtained T -X T Y T Z T , camera coordinate system O C -X C Y C Z C The coordinate transformation matrix between Among them, P T is the tool coordinate system O T -X T Y T Z T The coordinates of a point are described below, that is P C is the camera coordinate system O C -X C Y C Z C The following is the coordinate description of the same point, that is, To transform the camera coordinate system O C -X C Y C Z C Rotate to the tool coordinate system O T -X T Y T Z T The rotation matrix of the same orientation, O is the origin of the camera coordinate system C Relative to the tool coordinate system O T -X T Y T Z T The position vector in .

9. The point cloud data processing method according to claim 5, characterized in that: The camera coordinate system O is converted into C -X C Y C Z C The upper original point cloud coordinates / outer original point cloud coordinates / lower original point cloud coordinates are converted to the base coordinate system O B -X B Y B Z B To obtain the corresponding mapping point cloud coordinates: Among them, x C ,y C 、z C is the original point cloud coordinate; x B ,y B 、z B Map the original point cloud coordinates to the base coordinate system O B -X B Y B Z B The mapped point cloud coordinates obtained after downloading.

10. A laser welding device, characterized in that: It includes a laser welding head, a weld point cloud model generation unit and a control unit, and the weld point cloud model generation unit is used to implement the point cloud data processing method described in any one of claims 1 to 9. The control unit controls the movement of the laser welding head based on the obtained weld point cloud model to complete the welding of fin tubes and U-shaped tubes.

Images