Welding seam tracking method and system of laser welding robot and robot
By using vision sensors in laser welding robots to extract weld feature points in real time and calculate dynamic adjustment amounts, the problem of weld deviation during welding of thin-wall stainless steel pipes is solved, and the stability and efficiency of welding quality are achieved.
Patent Information
- Application Number
- CN202510348467.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-03-24
- Publication Date
- 2025-05-27
Smart Images
Figure CN120038434A_ABST
Abstract
Description
Technical Field
[0001] As an important type of pipe, thin-walled stainless steel pipes are widely used in industries such as petroleum, environmental protection, food processing, and medical device manufacturing. However, due to the relatively small melting range during laser welding of stainless steel pipes, when the shape of the weld track curve of the component connection is complex, it is easy to cause the phenomenon that the weld deviates from the original curve track. Coupled with the thermal expansion characteristics of the stainless steel pipe material, quality defects such as deformation, burn, crack, and weld deviation will be caused. Especially for thin-walled stainless steel pipes used in environmental detection equipment, the pipe fitting materials are all 316L stainless steel with a relatively thin thickness, and the weld track is usually the intersection line formed by the intersection of two cylinders of different sizes. When the robot welding torch performs curve feeding along the intersection line, errors and interferences are likely to cause the weld to deviate, affecting the weld quality and even resulting in welding scrap.
[0002] Therefore, it is necessary to rely on high-precision real-time monitoring tools for the weld track. The existing laser welding technology for thin-walled stainless steel pipes lacks high-precision real-time monitoring tools for the weld track and cannot quickly and accurately feedback and adjust the welding distance position, resulting in unstable weld quality and poor consistency. Summary of the Invention
[0003] The present invention aims to overcome at least one defect of the above-mentioned existing technology, and provides a weld tracking method, system and robot for a laser welding robot, which is used to solve the problem of weld deviation during the laser welding process of thin-walled stainless steel pipes in the existing technology, and realizes the real-time monitoring and precise adjustment of the weld track.
[0004] The present invention provides a weld tracking method for a laser welding robot, including the following steps:
[0005] S01. A vision sensor extracts weld feature points in real time to obtain real-time weld feature information;
[0006] S02. Send the real-time weld feature information to a controller;
[0007] S03. Calculate the dynamic adjustment amount △S(τ) during the welding process through a robot weld deviation adjustment algorithm
[0008]
[0009] where τ is the time of lag detection, t represents the current moment, and S(t) is the real-time weld deviation amount;
[0010] S04. Continuously track and adjust and correct the trajectory according to the dynamic adjustment amount △S(τ) to realize the tracking and adjustment of the robot motion trajectory.
[0011] Since the melting range of laser welding of thin-walled stainless steel pipes is relatively small, when the shape of the weld track of the component connection is complex, it is very easy to cause the deviation of the weld during the welding process. Moreover, due to the thermal expansion characteristics of stainless steel materials, quality defects such as deformation, burn and crack are likely to occur during laser welding, affecting the welding effect and even resulting in welding scrap. Therefore, during the welding process, it is necessary to monitor the weld track in real time. The real-time weld feature information described in step S01 includes the weld width and the left and right misalignment distances of the weld, which helps to quickly and accurately feedback the position of the welding torch.
[0012] The robot weld deviation adjustment algorithm described in step S03 includes:
[0013] S31. Obtain the tracking adjustment curve Y(t) of the welding torch center and the groove center curve R(t) of the actual weld according to the real-time weld feature information:
[0014]
[0015] S32. Calculate the time τ of the lag detection:
[0016]
[0017] Wherein, V represents the welding speed, and λ represents the distance between the detection point and the welding point;
[0018] S33. Calculate and obtain the delay deviation distance el(t);
[0019]
[0020] S34. Obtain the actual error e(t) between the welding torch walking time and the groove center:
[0021]
[0022] During the welding process, after the laser emitted by the laser focus unit is reflected or refracted by the workpiece, the filter retains the light of a specific wavelength emitted by the laser focus unit and filters out light of other wavelengths. When the vision sensor extracts the weld feature points in real time, it will be affected by the laser focus unit, the camera, and the welding torch. The real-time extraction of the weld feature points by the vision sensor described in step S01 further includes: the vision camera of the vision sensor generates a weld image by vertical shooting, which can maximally avoid the relative positions in space of the laser focus unit, the vision camera itself, and the welding torch in the arc light area, and can also reduce the image distortion caused by the shooting angle, so that the captured image can reflect the relative positions of the weld and the welding torch, ensuring that the subsequent extraction and analysis of the weld feature points can be based on accurate and reliable image data, and avoiding image distortion or information loss caused by improper shooting angles, simplifying the image processing flow, improving the speed and efficiency of image processing, helping to achieve the real-time extraction and analysis of the weld feature points, and ensuring that the dynamic adjustment during the welding process can respond to the changes in the weld state in a timely and accurate manner.
[0023] Since the smoke, dust, and flying debris generated during the welding process will have a certain impact on the acquisition of the weld morphology data, therefore, a filter protection plate should be installed on each sensor to block the interference of arc light, spatter, and smoke on the laser focus unit, making the vision sensor more accurate and stable.
[0024] To ensure that the weld track during the welding process is more accurate and stable, it is necessary to preset the accuracy of the deviation of the weld track. As long as the deviation of the weld track is within the preset range, it means that the weld track conforms to the expected curve. Since the welding materials and welding requirements are usually different, the left and right threshold deviation limit ranges of the seeking point are set according to the specific welding materials, workpiece thickness, and accuracy requirements, so as to adapt to different welding scenarios and requirements and improve the flexibility and versatility of the welding operation.
[0025] By setting the left and right threshold deviation limit ranges, the system can monitor the deviation of the weld track in real time. When one of the directions of the robot exceeds the threshold deviation limit range during the welding process, the actions and positions of the robot are adjusted to limit the actual weld track within the left and right threshold deviation ranges, improving the welding quality and stability.
[0026] When the visual sensor acquires the weld seam image, there will be many noise signals mixed in. Therefore, acquiring the weld seam image also includes preprocessing the weld seam image, filtering and restoring the acquired weld seam image. Acquiring the weld seam image also includes setting the parameters of the image filter, the angle of the vision camera, the speed of continuous filtering, and the left and right boundary values of the filter to further optimize the filtering effect, ensuring that while removing noise, useful information in the weld seam image is retained as much as possible, and filtering and improving interference such as sputtering noise and electronic noise, thereby improving the accuracy and quality of the weld seam image.
[0027] The present invention also provides a laser welding seam tracking system, including:
[0028] An acquisition module: used to extract weld seam feature points in real time to obtain real-time weld seam feature information;
[0029] A transmission module: used to send the real-time weld seam feature information to the controller;
[0030] A calculation module: calculates the dynamic adjustment amount △S(τ) during the welding process through the robot weld deviation adjustment algorithm;
[0031] An adjustment module: makes continuous trajectory tracking adjustments and corrections according to the dynamic adjustment amount △S(τ) to achieve the tracking and adjustment of the motion trajectory.
[0032] The present invention also provides a laser welding robot, including:
[0033] A laser focusing device, used to emit laser light;
[0034] A laser welding torch, used to perform welding at the weld seam;
[0035] An image detection sensor, used to acquire the weld seam image;
[0036] It also includes a weld seam tracking adjustment mechanism to implement the weld seam tracking method of the laser welding robot, used to adjust the laser welding torch and the image detection sensor to form a weld seam curve trajectory that conforms to the preset.
[0037] The beneficial effects of the present invention are as follows: For the laser welding robot seam tracking method, system and robot, through the robot seam deviation adjustment algorithm, the dynamic adjustment amount △S(τ) during the welding process is obtained, and continuous trajectory tracking adjustment and correction are performed according to the dynamic adjustment amount △S(τ) to achieve real-time tracking and adjustment of the robot motion trajectory. The vision camera of the vision sensor generates a seam image by vertical shooting, which can maximize avoiding the relative positions in space of the laser concentrator, the vision camera itself and the welding torch in the arc light area, and can also reduce the image distortion caused by the shooting angle, so that the captured image can reflect the relative positions of the seam and the welding torch, ensuring that the subsequent extraction and analysis of the seam feature points can be based on accurate and reliable image data, and avoiding image distortion or information loss caused by improper shooting angles, simplifying the image processing process, improving the speed and efficiency of image processing, helping to achieve real-time extraction and analysis of the seam feature points, ensuring that the dynamic adjustment during the welding process can respond to the changes in the seam state in a timely and accurate manner, setting the limit range of the welding threshold deviation according to different welding materials and welding requirements, ensuring that the error of the welding trajectory is within the limit range of the threshold deviation, so as to adapt to different welding scenarios and requirements and improve the flexibility and versatility of the welding operation. BRIEF DESCRIPTION OF THE DRAWINGS
[0038] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.
[0039] Figure 1 It is a schematic diagram of the steps of the laser welding robot seam tracking method.
[0040] Figure 2 It is a schematic diagram of the steps of the robot seam deviation adjustment algorithm.
[0041] Figure 3 It is a diagram of the deviation adjustment of the seam tracking trajectory.
[0042] Figure 4 It is a diagram for detecting the position deviation of the laser welding seam.
[0043] Figure 5 It is a calibration diagram of the feature points during the seam tracking programming in Embodiment 4.
[0044] Figure 6 It is the camera angle, the left and right boundary values of the filter, and the filter setting in Embodiment 4.
[0045] Figure 7It is the process of weld seam tracking programming and parameter setting in Example 5.
[0046] Figure 8 It is the selection of the specific positions of the workpiece feature points in Example 5.
[0047] Figure 9 It is a schematic structural diagram of the laser welding robot.
[0048] Figure 10 It is a schematic structural diagram of the laser welding workstation for thin-walled stainless steel pipes in Example 7. Specific implementation manners
[0049] To make the objectives, technical solutions and advantages of the present invention clearer, the following will further describe in detail the embodiments of the present invention in conjunction with the accompanying drawings:
[0050] The embodiments of the present invention provide an embodiment of a weld seam tracking method, system and robot for a laser welding robot. Although the logical sequence is shown in the flowchart, under certain data, the steps shown or described can be completed in a different order than here.
[0051] Example 1: Refer to Figure 1
[0052] As Figure 1 As shown in the step schematic diagram of a weld seam tracking method for a laser welding robot shown, the present invention provides a weld seam tracking method for a laser welding robot, including the following steps:
[0053] S01. The vision sensor extracts the weld seam feature points in real time to obtain real-time weld seam feature information;
[0054] S02. Send the real-time weld seam feature information to the controller;
[0055] S03. Calculate the dynamic adjustment amount △S(τ) during the welding process through the robot weld seam deviation adjustment algorithm
[0056]
[0057] where τ is the time of lag detection, t represents the current moment, and S(t) is the real-time weld seam deviation amount;
[0058] S04. Continuously track and adjust the trajectory according to the dynamic adjustment amount △S(τ) to realize the tracking and adjustment of the robot motion trajectory.
[0059] Due to the relatively small melting range in the laser welding of thin-walled stainless steel pipes, when the shape of the weld track of the component connection is complex, it is very easy to cause the deviation of the weld during the welding process. Moreover, due to the thermal expansion characteristics of stainless steel materials, quality defects such as deformation, burn and crack are likely to occur during the laser welding process, affecting the welding effect and even resulting in welding rejection. Therefore, during the welding process, it is necessary to monitor the weld track in real time. The real-time weld feature information described in step S01 includes the weld width and the left and right misalignment distances of the weld, which helps to quickly and accurately feedback the position of the welding torch.
[0060] Example 2: Refer to Figures 2 to 3
[0061] As Figure 2 shown is the robot weld deviation adjustment algorithm. The robot weld deviation adjustment algorithm described in step S03 includes:
[0062] S31. Obtain the tracking adjustment curve Y(t) of the welding torch center and the groove center curve R(t) of the actual weld according to the real-time weld feature information:
[0063]
[0064] S32. Calculate the time τ of the lag detection:
[0065]
[0066] where V represents the welding speed and λ represents the distance between the detection point and the welding point;
[0067] S33. Calculate and obtain the delay deviation distance el(t);
[0068]
[0069] S34. Obtain the actual error e(t) between the welding torch travel time and the groove center:
[0070]
[0071] As Figure 3 shown, the floating coordinates of the welding torch center are adjusted while welding, and the actual weld track is restricted within the left and right threshold deviation ranges, forming a weld curve track that meets the preset.
[0072] The dotted line Y(t) therein is the tracking adjustment curve of the welding torch center, which is also regarded as the output quantity fed back by the system to the actuator, that is:
[0073]
[0074] The offset e1(t) of the weld groove point B detected by the sensor is actually the deviation of point B on the R(t) curve relative to point A on the Y(t), that is
[0075]
[0076] Assume the welding speed is V (mm / s), then the time lag of the welding point A behind the detection point B is:
[0077]
[0078] Assume further that △S(τ) is the tracking adjustment amount of the torch center from the moment t - τ to the moment t, that is:
[0079]
[0080] Then the actual error between the torch travel time and the groove center should be:
[0081]
[0082] Therefore, the system synchronously compares and calculates the position of the feature point specified in advance according to the torch movement intersection curve with the feature point of the actual movement route captured by the vision sensor to obtain the dynamic adjustment amount △S(τ) that needs to be corrected, and then obtains the adjustment amount of the specific trajectory point according to the left and right threshold deviations, and performs continuous trajectory tracking adjustment and correction, so as to achieve high-precision tracking of the robot movement trajectory.
[0083] Example 3: Refer to Figure 4
[0084] During the welding process, the laser emitted by the laser focus unit passes through the workpiece by reflection or refraction, and after passing through the filter, the light of the specific wavelength emitted by the laser focus unit is retained, and the light of other wavelengths is filtered out. When the vision sensor extracts the weld feature points in real time, it will be affected by the laser focus unit, the camera and the welding torch. The real-time extraction of the weld feature points by the vision sensor in step S01 further includes: the vision camera of the vision sensor generates a weld image by vertical shooting, which can maximally avoid the relative positions in space of the laser focus unit, the vision camera itself and the welding torch in the arc light area, and can also reduce the image distortion caused by the shooting angle, so that the captured image can reflect the relative position of the weld and the welding torch, ensuring that the subsequent extraction and analysis of the weld feature points can be based on accurate and reliable image data, and avoiding image distortion or information loss caused by improper shooting angles, simplifying the image processing process, improving the speed and efficiency of image processing, contributing to the real-time extraction and analysis of the weld feature points, and ensuring that the dynamic adjustment during the welding process can respond to the changes in the weld state in a timely and accurate manner.
[0085] As Figure 4As shown in the figure, after the light band is reflected or refracted by the workpiece, the filter retains the light of a specific wavelength emitted by the laser, while filtering out light of other wavelengths. Then, the camera generates a weld image, and the weld feature points are extracted through an optical algorithm to obtain data such as the weld width and the left-right misalignment distance of the weld, and the actual deviation distance is calculated. Based on the preset positions of the feature points on the torch center trajectory, a synchronous comparison calculation is performed with the actual movement route captured by the vision sensor to obtain the dynamic adjustment amount that needs to be corrected, and then continuous trajectory tracking adjustment and correction are performed according to the adjustment amount, so as to achieve high-precision tracking of the robot's movement trajectory.
[0086] During the welding process, the generated soot and flying debris will have a certain impact on the acquisition of weld morphology data. Therefore, when each sensor is used, a filter protection plate is installed to block the interference of arc light, flying debris, and soot on the laser concentrator, making the vision sensor more accurate and stable.
[0087] Example 4: Refer to Figures 5 to 6
[0088] The present invention also provides a laser welding seam tracking system, including:
[0089] An acquisition module: used to extract weld feature points in real time to obtain real-time weld feature information;
[0090] A transmission module: used to send the real-time weld feature information to the controller;
[0091] A calculation module: calculates the dynamic adjustment amount △S(τ) during the welding process through the robot weld deviation adjustment algorithm;
[0092] An adjustment module: performs continuous trajectory tracking adjustment and correction according to the dynamic adjustment amount △S(τ) to achieve the tracking and adjustment of the movement trajectory.
[0093] During the laser welding process of stainless steel pipes, when using the Sairong weld visual tracking control system for adaptive tracking, it is necessary to select and agree on the positions of the detection feature points of the stainless steel pipe weld trajectory in order to achieve real-time control of the weld tracking process. Through the programming window of the Sairong software, specific feature point positions are selected on the stainless steel pipe workpiece for calibration and corresponding programming settings. As Figure 5 shown, which includes search positions, weld types, and sampling periods, etc., so that the laser can detect and track the weld position according to the search conditions and feature points, and compare the search feature point results with the coordinates of the preset trajectory to continuously and dynamically generate deviation data.
[0094] According to the specific welding materials, workpiece thickness, and precision requirements, the left and right threshold deviation limit ranges of the seam finding points are both set to 0.3 mm in the system. The system calculates the deviation between the detected weld seam and the ideal trajectory based on the actual position of the tracking feature points. If the deviation in one direction exceeds the 0.3 mm threshold deviation range, adjustment is required. Then, the motion control mechanism corrects the center position of the welding torch according to the output deviation data, precisely guiding the welding gun to automatically correct the deviation, thus achieving intelligent and precise real-time laser weld seam tracking welding.
[0095] To ensure that the weld seam trajectory during the welding process is more accurate and stable, it is necessary to preset the precision of the deviation of the weld seam trajectory. As long as the deviation of the weld seam trajectory is within the preset range, it indicates that the weld seam trajectory conforms to the expected curve. Since welding materials and welding requirements are usually different, according to the specific welding materials, workpiece thickness, and precision requirements, the left and right threshold deviation limit ranges of the seam finding points are set to adapt to different welding scenarios and requirements, improving the flexibility and versatility of welding operations.
[0096] By setting the left and right threshold deviation limit ranges, the system can monitor the deviation of the weld seam trajectory in real-time. When the deviation of the robot in one direction exceeds the threshold deviation limit range during the welding process, the actions and positions of the robot are adjusted to limit the actual weld seam trajectory within the left and right threshold deviation ranges, improving the welding quality and stability.
[0097] As Figure 6 shown, when the vision sensor acquires the weld seam image, many noise signals are attached. Therefore, acquiring the weld seam image also includes preprocessing the weld seam image, filtering and restoring the acquired weld seam image. Acquiring the weld seam image also includes setting the parameters of the image filter, the angle of the vision camera, the speed of continuous filtering, and the left and right boundary values of the filter to further optimize the filtering effect, ensuring that while removing noise, as much useful information in the weld seam image as possible is retained, filtering and improving interference such as sputtering noise and electronic noise, thereby improving the precision quality of the weld seam image.
[0098] Example 5: Refer to Figures 7 to 8
[0099] As Figure 7 shown is the programming setting of the Sairong weld seam tracking. Based on the positions of the weld seam feature points collected by the vision sensor, the dynamic deviation correction and adjustment of the welding torch also require the use of the KUKA laser welding programming WeldCcom software and the Sairong weld seam tracking control software to achieve the extraction and tracking adjustment of the continuous characteristic parameters during the welding process.
[0100] Set various laser parameters and welding process parameters in the programming of the Sairong software, and generate the Figure 8Feature points such as P1, P2, P3, and P4. The program for the first circular arc intersection weld tracking trajectory of the stainless steel pipe is as follows:
[0101] 1 → INI / / Initialize;
[0102] 2 vbadap = false; deactivate ADAP function / / Set vbadap to false and deactivate the ADAP function;
[0103] 3 PTP HOMEAA Vel = 20% DEFAULT Tool[1]:HQ Base[1]:Kinematik Machine2 / / The welding torch will move to the HOMEAA position at a speed of 20%, with Kinematik Machine2 as the reference coordinate;
[0104] 4 SeamTrack Init / / Initialize seam tracking;
[0105] 5 LIN P9 Vel =.2m / s CPDAT6 Tool[1]:HQ Base[1]:Kinematik Machine2 / / Linearly move to the P9 position at a speed of 0.2 m / s, using CPDAT6 data, with Kinematik Machine2 as the reference coordinate; 6 SeamTrack On Dist = 0mn Set = S1 / / Turn on seam tracking, distance is 0, set is S1;
[0106] 7 LIN P2 Vel =.2m / s CPDAT1 Tool[1]:HQ Base[1]:Kinematik Machine2 / / The welding torch linearly moves to the P2 position at a speed of 0.2 m / s;
[0107] 8 startadap(); start adaptive welding function / / Start the adaptive welding function;
[0108] 9 ARCON WDAT1 LIN P1 Vel = 0.2m / s CPDAT2 Tool[1]:HQ Base[1]:Kinematik Machine2 / / Linearly start laser welding to the P1 position at a speed of 0.2 m / s;
[0109] 10 ARCSWI WDAT5 CIRC P12 P13 CPDAT8 Tool[1]:HQ Base[1]:Kinematik Machine2 / / Arc welding from the P2 position to the P3 position;
[0110] 11 ARC SWI WDAT6 CIRC P4 P5 CPDAT9 Tool[1]:HQ Base[1]:kinematic Machine2 / / Arc welding from position P4 to P5;
[0111] 12 ARC SWI WDAT7 CIRC P6 P7 CPDAT10 Tool[1]:HQ Base[1]:Kinematik Machine2 / / Arc welding from position P6 to P7;
[0112] 13 ARC SWI WDAT7 CIRC P7 P8 CPDAT10 Tool[1]:HQ Base[1]:Kinematik Machine2 / / Arc welding from position P7 to P8;
[0113] 14 ARC OFF WDAT9 LIN P8 CPDAT12 Tool[1]:HQ Base[1]:kinematic Machine2 / / Turn off arc welding linearly to position P9;
[0114] 15 Seam Track OFF Dist = 0mm Keep offset = FALSE / / Turn off seam tracking, do not keep offset 16 LIN P25 Vel =.1m / s CPDAT14 Tool[1]:HQ Base[1]:Kinematik Machine2 / / Move linearly to the preparation position at a speed of 0.1 m / s;
[0115] 17 finishadap(); finish adaptive welding function 18 Seam Track Clear / / Clear seam tracking data;
[0116] 19 PTP P10 Ve1 = 20 PDAT2 Tool[1]:HQ Base[1]:Kinematik Machine2 / / The welding torch moves to the starting position at a speed of 20%;
[0117] 20 END / / End of program;
[0118] Example 6: Refer to Figure 9
[0119] As Figure 9 shown, the present invention also provides a laser welding robot, comprising:
[0120] Laser focusers, which are used to emit lasers;
[0121] Laser welding torches, which are used to perform welding at the weld seams;
[0122] Image detection sensors, which are used to obtain weld seam images;
[0123] It further includes a weld seam tracking and adjustment mechanism, which implements the described weld seam tracking method of a laser welding robot and is used to adjust the laser welding torch and the image detection sensor to form a weld seam curve trajectory that conforms to the preset.
[0124] Weld seam tracking sensors are usually installed in front of the welding torch at a preset distance. The sensor head includes a CCD camera and one or two semiconductor lasers. The lasers are arranged obliquely as structured light sources to project laser stripes onto the surface of the workpiece below the sensor at a preset angle. The weld seam tracking sensor adopts the principle of laser triangulation reflection. The laser beam is magnified to form a laser line projected onto the surface of the object to be measured. The reflected light passes through a high-quality optical system and is projected onto the imaging matrix. The laser emitted by the laser focuser forms a light sheet through a cylindrical lens and irradiates the workpiece to form a very narrow light band.
[0125] Example 7: Refer to Figure 10
[0126] As Figure 10Shown is a laser welding workstation for stainless steel components of environmental protection detection equipment, which consists of a horizontally sealed cylindrical tube with a diameter of 80 mm and a length of 900 mm for the liquid to be detected, plus 5 detection reagent addition tubes with a diameter of 25 mm. It is required that the welded pipe fittings have sealing performance, corrosion resistance and durability after forming. The laser welding workstation consists of a fiber laser generator, a KUKA welding robot of model KR60HA, a KR-C2 robot control cabinet, a Fronius TPS400 welding machine, a Sailong weld visual tracking control system, a stainless steel pipe positioner fixture, etc. Through the coordinated control of the KR-C2 robot controller, the welding robot can work in coordination with various devices such as the welding machine and the Sailong weld tracking system to ensure that the welding torch performs laser welding along the correct weld track. The KUKA robot controller is equipped with a high-speed CPU and has a fast operation speed. The profinet communication module has rich communication interfaces and communication protocols, which is convenient for data communication control with the vision system, the robot welding torch, the welding power source, etc. The control signal is transmitted to the servo motor and the joint mechanism through the robot controller to control the movement of the welding torch. The newly launched KR60HA small arc welding intelligent robot of KUKA is selected. The welding robot has a large working range, and the movement range of the welding torch can reach a radius of 1811 mm. At the same time, it has a flexible six-axis structure, which can meet the requirements of complex welding track changes of the special-shaped curve of the stainless steel pipe. During welding, the laser sensor continuously obtains weld feature data at a high detection speed of 100 frames per second for calculation, and feeds the results back to the system for trajectory correction and adjustment.
[0127] According to the geometric shape range and accuracy requirements of the intersecting line track of the stainless steel pipe laser welding, a single-axis positioner fixture that can rotate 0-120° is used to clamp the stainless steel pipe workpiece. It is necessary to ensure that the welded steel pipe is firmly clamped during the rotational movement of the positioner. Therefore, special left and right clamping molds are designed to clamp and support the stainless steel pipe, and after positioning, it is clamped and fixed with a cylinder. During welding, the positioner fixture dynamically adjusts the position of the workpiece in real time according to the change of the welding position, as well as the rotation and swing speed and angle, and coordinates with the action change of the robot welding torch, which is convenient for welding at different curve positions and ensures high-precision positioning and stable weld quality during the welding process.
Claims
1. A laser welding robot weld tracking method, characterized in that: The following steps are involved: S01, the visual sensor extracts the weld feature points in real time to obtain real-time weld feature information; S02, sending the real-time weld feature information to a controller; S03. Calculate the dynamic adjustment amount △S (τ) during welding by using the robot weld deviation adjustment algorithm Among them, τ is the time of delayed detection, t represents the current moment, and S(t) is the real-time deviation of the weld; S04. Continuously perform trajectory tracking adjustment and correction according to the dynamic adjustment amount ΔS(τ) to achieve tracking and adjustment of the robot's motion trajectory.
2. A laser welding robot weld seam tracking method according to claim 1, characterized in that: The real-time weld feature information in step S01 includes the weld width and the left and right offset distances of the weld.
3. A laser welding robot weld seam tracking method according to claim 1, characterized in that: The robot weld deviation adjustment algorithm in step S03 includes: S31, obtaining the tracking adjustment curve Y(t) of the welding torch center and the groove center curve R(t) of the actual weld according to the real-time weld feature information: S32, calculate the time τ of the hysteresis detection: Where V represents the welding speed, and λ represents the distance between the detection point and the welding point; S33, calculate and obtain the delay deviation distance el(t); S34, after obtaining the actual error e(t) between the welding torch travel time and the groove center:
4. A laser welding robot weld seam tracking method according to claim 2, characterized in that: The real-time extraction of weld feature points by the visual sensor in step S01 also includes: the visual camera of the visual sensor generates a weld image by vertical shooting, ensuring that the shot image can reflect the relative position of the weld and the welding gun.
5. A laser welding robot weld seam tracking method according to claim 3, characterized in that: The robot weld deviation adjustment algorithm also includes: setting the left and right threshold deviation limit ranges of the position finding according to the specific welding material, workpiece thickness and accuracy requirements.
6. A laser welding robot weld seam tracking method according to claim 5, characterized in that: When one direction of the robot exceeds the threshold deviation limit range during the welding process, the robot's movement and position are adjusted to limit the actual weld trajectory within the left and right threshold deviation ranges.
7. A laser welding robot weld seam tracking method according to claim 2, characterized in that: Acquiring the weld image also includes preprocessing the weld image, filtering and restoring the acquired weld image, and improving the accuracy of the weld image.
8. A laser welding robot weld seam tracking method according to claim 7, characterized in that: The method of acquiring the weld image also includes setting parameters of the image filter, the angle of the visual camera, the speed of continuous filtering, and the left and right boundary values of the filter.
9. A laser welding seam tracking system, characterized in that: include: Acquisition module: used to extract weld feature points in real time and obtain real-time weld feature information; Transmission module: used to send real-time weld feature information to the controller; Calculation module: Calculate the dynamic adjustment amount △S(τ) during welding through the robot weld deviation adjustment algorithm; Adjustment module: performs continuous trajectory tracking adjustment and correction according to the dynamic adjustment amount ΔS(τ) to achieve tracking and adjustment of the motion trajectory.
10. A laser welding robot, comprising: A laser focuser, used to emit laser light; A laser welding gun, used to weld at the seam; An image detection sensor for acquiring an image of the weld; It is characterized by further comprising: The weld seam tracking adjustment mechanism realizes a weld seam tracking method of a laser welding robot as described in claims 1-8, and also includes a single-axis positioner fixture for adjusting the angle and position of the laser welding gun and the image detection sensor to form a preset weld seam curve trajectory.