Method for monitoring TIG welding quality in real time based on molten pool camera and welding equipment

Abstract

The invention relates to the field of electric arc welding, and discloses a method for monitoring TIG welding quality in real time based on a molten pool camera, and the method comprises the following steps: S100, obtaining an image of a welding area through the molten pool camera; s200, identifying a tungsten needle and a workpiece groove in the image, and calibrating imaging parameters of the image according to the actual size of the tungsten needle; s300, the relative position of the tail end of the tungsten needle and the workpiece groove is obtained, and the position of the tungsten needle is adjusted according to the relative position; s400, TIG welding is conducted on the groove of the workpiece, and a real-time image in the welding process is obtained through a molten pool camera; and S500, relevant welding data are obtained from the real-time image, the position state of a tungsten needle and a workpiece groove and / or the working state of a welding power source are / is controlled in real time according to the relevant welding data, welding is stopped and an alarm is given when the data exceed a set threshold value range, the change of welding parameters can be monitored in real time, and the quality of a welding seam is guaranteed. The invention further discloses welding equipment for monitoring the TIG welding quality in real time based on the molten pool camera.

Description

Technical Field

[0001] The present application relates to the field of arc welding, and more particularly to a method for real-time monitoring of TIG welding quality based on a molten pool camera. In addition, the present application also relates to a welding device for real-time monitoring of TIG welding quality based on a molten pool camera. Background Art

[0002] Arc welding is a welding method that fuses workpieces together through the heat generated by the welding arc. The welding arc is powered by a welding power source. Under a certain voltage, a strong and long-lasting discharge phenomenon is generated between the welding electrode and the workpiece. Arc discharge has the characteristics of low voltage, high current, high temperature and strong luminescence. With the help of the high temperature generated by the arc discharge, the workpiece at the welding point can be melted, or the solder and the workpiece can be melted together, and a weld can be formed after condensation, firmly connecting the workpieces together. The welding arc can be generated between the workpieces to be welded, or between the welding gun and the workpiece. During the welding process, a shielding gas is usually output in the weld area to prevent the material in the welding area from oxidizing under the action of high temperature, which will affect the quality of the weld.

[0003] TIG (Tungsten Inert Gas) welding is a welding method that uses tungsten or tungsten alloy as the welding gun electrode and inert gas, usually argon, as the shielding gas. It is commonly used to weld thin plates of stainless steel and non-ferrous metals such as aluminum, magnesium, and copper alloys. It has the advantages of convenient control of welding point position and high welding quality. With the development of automatic control technology, automated welding equipment such as robot TIG welding, pipeline trolley TIG welding, and special machine TIG welding have been derived. Automated welding has high requirements for product consistency. It is necessary to monitor the welding area in real time during the welding process to promptly detect changes in weld quality caused by factors such as workpiece groove size, tungsten electrode shape, or arc shape during automated TIG welding.

[0004] Existing TIG welding equipment usually monitors welding quality through manual visual methods, and the monitoring results are greatly affected by human factors. Recently, some TIG welding equipment that uses molten pool cameras to monitor welding quality has also appeared. However, due to the large difference in brightness of the welding area, existing TIG welding equipment can usually only monitor the molten pool with higher brightness in real time, or monitor the welding quality by comparing the images before and after welding, and it is difficult to effectively monitor the real-time changes of welding parameters. Summary of the invention

[0005] In order to monitor the changes in welding parameters in real time and ensure the quality of the weld, the present application provides a method and welding equipment for real-time monitoring of TIG welding quality based on a molten pool camera.

[0006] The method for real-time monitoring of TIG welding quality based on a molten pool camera provided in the present application adopts the following technical solution: a method for real-time monitoring of TIG welding quality based on a molten pool camera, comprising the following steps: S100, acquiring an image of the welding area through a molten pool camera; S200, identifying the tungsten needle and the workpiece groove in the image, and calibrating the imaging parameters of the image according to the actual diameter and actual length of the tungsten needle; S300, acquiring the relative position between the end of the tungsten needle and the workpiece groove, and adjusting the initial position of the tungsten needle accordingly; S400, performing TIG welding on the workpiece groove, and acquiring a real-time image of the welding process through a molten pool camera; S500, acquiring the data of the tungsten needle, workpiece groove, molten pool and welding arc from the real-time image, controlling the position status of the tungsten needle and the workpiece groove and / or the working status of the welding power supply in real time according to the acquired data, and stopping welding when the data exceeds the set threshold range, and issuing an alarm message.

[0007] By adopting the above technical solution, using the image of the welding area obtained before welding, the initial diameter and initial length of the tungsten needle can be used to calibrate the imaging parameters such as the imaging angle and imaging distance of the image, so as to ensure the accuracy of the data such as the tungsten needle, workpiece, molten pool, welding arc, etc. obtained by the molten pool camera, and move the tungsten needle to a suitable initial position before welding. By using the molten pool camera to obtain real-time images during the welding process, real-time detection data of the tungsten needle, workpiece groove, molten pool and welding arc can be obtained. When the data deviates slightly, the moving trajectory of the welding gun and / or the welding power supply line can be adjusted in real time, so that welding processing can be carried out under better welding parameters, ensuring the quality of the weld during automated welding, and stopping the welding processing when it is found that the monitoring data deviates seriously and forms an abnormality.

[0008] In a specific feasible implementation scheme, in step S200, the method for calibrating the imaging parameters of the image according to the initial diameter and length of the tungsten needle includes the following steps: S210, obtaining the pixel value W of the tungsten needle in the image in the diameter direction; S220, using the formula Scale=D / W to calculate the length value Scale corresponding to each pixel in the corresponding direction of the image, where D is the diameter of the tungsten needle; S230, obtaining the pixel value W1 of the tungsten needle in the length direction of the image; S240, using the formula L1=Scale*W1 to calculate the imaging length L1 of the tungsten needle; S250, using the formula cosθ=L1 / L to calculate the installation angle θ between the molten pool camera and the tungsten needle, and using this to calibrate the actual length corresponding to the pixels in different directions of the image.

[0009] By adopting the above technical solution, the length corresponding to each pixel in the image is calibrated using the initial diameter of the tungsten needle, and the length of the object is estimated by the number of pixels of the object in the image, so that the size of the object can be easily judged by the image in the image. By using the ratio between the length of the tungsten needle estimated according to the number of pixels in the image and the initial length of the tungsten needle, the angle between the imaging camera and the length direction of the tungsten needle can be conveniently inferred from the length reduction ratio of the tungsten needle when imaging, so that the size and position of the object in the image in different directions can be adjusted accordingly, and the size and position of different objects in the image can be accurately obtained.

[0010] In a specific feasible implementation scheme, after step 250, the following steps are also included: S260, calculating the groove depth and groove width of the workpiece according to the installation angle θ and the pixel length Scale value of the molten pool camera; S270, verifying and adjusting the calibration result according to the actual groove depth and actual groove width of the workpiece.

[0011] By adopting the above technical solution, the corresponding data extracted from the image according to the calibration result is verified using the actual data of the groove of the workpiece, which can improve the accuracy of the data extracted from the image, thereby ensuring the quality of TIG welding.

[0012] In a specific feasible implementation method, step S300 includes the following steps: S310, obtaining the depth of the workpiece groove, the width of the groove bottom, the groove width and the length of the slopes on both sides; S320, obtaining the position of the end of the tungsten needle at the groove of the workpiece, and calculating the distance between the end of the tungsten needle and the bottom of the groove and the slopes on both sides; S330, controlling the movement of the tungsten needle so that the tungsten needle moves to the initial position where the end of the tungsten needle is a set distance from the bottom of the groove and at least one side of the slope.

[0013] By adopting the above technical solution, using the acquired data such as the depth of the groove of the workpiece, the width of the groove bottom, the width of the groove and the length of the slopes on both sides, it is possible to determine whether the groove can be filled by moving the tungsten needle in a straight line along the length direction of the groove, so that the optimal tungsten needle movement direction during welding can be selected. Using the relative position between the tungsten needle and the groove obtained from the image, the movement direction of the tungsten needle can be designed so that the tungsten needle automatically moves to a suitable initial position before welding.

[0014] In a specific implementation scheme, in step S400, a monochromatic light source is used to illuminate the welding area, and a filter with the same wavelength as the monochromatic light source is set in front of the molten pool camera to obtain a real-time image formed by the monochromatic light source during the welding process.

[0015] By adopting the above technical solution, using a monochromatic light source to illuminate the welding area, and setting a filter with the same wavelength as the monochromatic light in front of the molten pool camera, the molten pool camera can only image the light emitted by the monochromatic light source reflected by the objects in the welding area, while shielding the high-brightness light generated by the arc discharge, effectively reducing the brightness difference of the welding area imaging, making the exposure of the welding area imaging more uniform and the imaging of different objects clearer.

[0016] In a specific feasible implementation scheme, in step S500, the method for obtaining tungsten needle data from a real-time image includes the following steps: S511, extracting the edge contour line of the tungsten needle in the real-time image; S512, distinguishing the needle rod part and the end part in the edge contour line of the tungsten needle; S513, calculating the distance between the edge lines on both sides of the needle rod part, that is, the tungsten needle diameter, and the length of the contour line within the set length of the tungsten needle end; S514, outputting the tungsten needle diameter value, and comparing the length of the contour line of the tungsten needle end with the set contour line length. When the length of the contour line of the tungsten needle end is greater than the set contour line length, it is judged that the tungsten needle data exceeds the set threshold range.

[0017] By adopting the above technical solution, by extracting the edge contour line of the tungsten needle from the real-time image during the welding process, the real-time changes in the diameter and the end shape of the tungsten needle during the welding process can be obtained, thereby monitoring the deformation and loss of the tungsten needle during the welding process in real time; by comparing the contour line within the set length of the end of the tungsten needle with the set contour line length, the degree of deformation such as balling at the end of the tungsten needle can be monitored, and when the deformation of the end of the tungsten needle is sufficient to affect the quality of the weld, it is determined that the tungsten needle data exceeds the set threshold range, and welding is stopped and an alarm message is issued.

[0018] In a specific feasible implementation scheme, in step S500, the method for obtaining workpiece groove data from a real-time image includes the following steps: S521, extracting the surface contour line of the groove portion of the workpiece in the real-time image; S522, respectively calculating the groove width, bottom width and two side widths of the workpiece groove, and outputting the detection results; S523, calculating the distance between the end of the tungsten needle and the bottom and two side surfaces of the workpiece groove according to the end position of the tungsten needle in the image, and controlling the moving trajectory of the tungsten needle according to the calculation results.

[0019] By adopting the above technical solution, using the contour lines of different surfaces of the workpiece groove part extracted from the real-time image, it is possible to monitor the data changes of different positions of the workpiece groove in the length direction during the welding process, so that the welding parameters can be adjusted in real time according to the different groove data during welding to ensure the quality of the weld at different positions of the groove. By monitoring the real-time distance between the end of the tungsten needle and the bottom of the groove and both sides of the groove during welding, the position of the tungsten needle can be adjusted in real time to ensure that the welding arc generated by the end of the tungsten needle can act on the bottom and both sides of the groove at the same time, ensuring that the solder is evenly integrated with the workpiece in the groove, and ensuring the quality of the weld.

[0020] In a specific feasible implementation scheme, in step S523, the method for controlling the moving trajectory of the tungsten needle according to the distance between the end of the tungsten needle and the bottom and two side surfaces of the workpiece groove includes the following steps: calculating the equidistant distance when the end of the tungsten needle is equal to the bottom and two side surfaces of the groove according to the groove width, bottom width and two side surface widths of the workpiece groove; comparing the equidistant distance with the set maximum distance; if the equidistant distance is not greater than the set maximum distance, controlling the tungsten needle to move to a position equal to the bottom and two side surfaces of the groove, controlling the voltage of the welding power supply according to the equidistant distance, and controlling the tungsten needle to move along the length direction of the groove; if the equidistant distance is greater than the set maximum distance, moving the end of the tungsten needle to a position where the distances from the bottom and one side of the groove are both set distances, and controlling the tungsten needle to move back and forth in a direction perpendicular to the length of the groove while moving along the length direction of the groove.

[0021] By adopting the above technical scheme, by calculating the equidistant distance when the distance between the end of the tungsten needle and the bottom of the groove and the two sides of the groove is equal, it is possible to determine whether the arc emitted by the end of the tungsten needle can cover the bottom and both sides of the groove, and weld both sides of the groove at the same time. When the equidistant distance is not greater than the set maximum distance, it is determined that the arc can cover both sides of the groove at the same time. At this time, the tungsten needle is controlled to move along the length direction of the groove of the workpiece, so that the distance between the end of the tungsten needle and the bottom of the groove and both sides of the groove is always equal, and the welding power supply voltage is controlled according to the distance to ensure the stability of the weld quality; when the equidistant distance is greater than the set maximum distance, it is determined that the arc cannot weld both sides of the groove at the same time. At this time, the tungsten needle is controlled to move along the length direction of the groove of the workpiece, and at the same time, it is repeatedly moved back and forth along the width direction of the groove to form a weld covering both sides of the groove.

[0022] In a specific feasible implementation scheme, in step S500, the method for obtaining molten pool data and arc data from a real-time image includes the following steps: extracting the shape and brightness of the welding arc, as well as the shape and texture of the molten pool from the real-time image; comparing the arc shape with the standard arc shape to obtain a deviation value of the arc shape, and determining that the data is abnormal when the arc shape deviation value is greater than the arc shape deviation threshold; comparing the change amplitude and average brightness of the arc brightness within a set time period, and comparing the change amplitude with the brightness change threshold, and comparing the average brightness with the brightness threshold, and comparing the brightness with the brightness threshold. When the change amplitude is greater than the brightness change threshold, or the average brightness is greater than the brightness threshold, it is judged as a data anomaly; the welding power supply is adjusted according to the arc morphology deviation value and the deviation value between the average brightness and the standard brightness; the molten pool morphology is compared with the standard molten pool morphology to obtain the deviation value of the molten pool morphology, and it is judged as a data anomaly when the molten pool morphology deviation value is greater than the molten pool morphology deviation threshold; the molten pool texture is compared with the standard molten pool texture to obtain the deviation value of the molten pool texture, and it is judged as a data anomaly when the molten pool texture deviation value is greater than the molten pool texture deviation threshold; the welding power supply is adjusted according to the molten pool morphology deviation value.

[0023] By adopting the above technical solution, the shape change and position deviation of the arc can be judged by using the deviation value between the arc shape and the standard arc shape obtained in real time, so as to judge whether the arc can ensure the normal progress of welding; by using the real-time monitoring of the change of arc brightness and the average brightness value, the stability of the arc and whether the arc can meet the welding requirements can be judged. When the shape and brightness of the arc deviate from the required value, the welding power supply is adjusted. When the deviation value exceeds the set limit and the welding cannot be effectively carried out, an abnormal information is issued. By using the real-time acquired molten pool shape and molten pool texture information, the deviation between the molten pool shape and the standard molten pool shape, as well as the deviation between the molten pool texture and the standard molten pool texture can be monitored. When the deviation is greater than the set threshold, it is judged that it does not meet the welding requirements and an abnormal data information is issued; and when the deviation does not exceed the set threshold, the welding power supply is adjusted so that the molten pool shape and texture are closer to the standard molten pool shape and standard molten pool texture, so as to ensure the consistency of welding quality.

[0024] The welding equipment for real-time monitoring of TIG welding quality based on a molten pool camera provided in the present application can realize the method for real-time monitoring of TIG welding quality based on a molten pool camera provided in the present application, and also has the corresponding advantages of the method for real-time monitoring of TIG welding quality based on a molten pool camera provided in the present application.

[0025] In summary, the present application includes at least one of the following beneficial technical effects: 1. By comparing the tungsten needle data extracted from the image of the welding area obtained before welding with the actual tungsten needle data, the size and relative position of each object in the welding area image captured by the molten pool camera can be calibrated, so that the real-time status of the welding process can be monitored through the image of the welding area captured in real time during the welding process, and effective monitoring and real-time adjustment of the welding process can be achieved.

[0026] 2. By acquiring data on the tungsten needle, workpiece groove, molten pool and arc from real-time images, the main factors affecting the quality of the weld can be effectively monitored. When the data is abnormal enough to cause welding defects, welding is stopped and an alarm message is issued to ensure that the welding quality meets the basic welding requirements. When there is a data deviation, the trajectory of the relative movement between the welding gun and the workpiece and / or the welding power supply are adjusted in real time, so that the welding parameters are closer to the standard process requirements, thereby improving the consistency of welding quality.

[0027] 3. By setting a monochromatic lighting source for the molten pool camera and setting a filter that can only pass the monochromatic light in front of the camera, the molten pool camera can only image the monochromatic light reflected by the objects in the welding area, thereby eliminating the high-brightness light generated by the arc and the ultra-high exposure caused by the high-brightness arc light reflected by the objects in the welding area, reducing the difference in exposure of the objects in the image, making the imaging of the objects in the image clearer, so that the edge contours of the objects can be clearly identified, and effective identification of objects in the low-exposure area of ​​the image can be achieved.

[0028] 4. By real-time monitoring of the groove data during welding, when the groove width of the workpiece changes, the welding parameters can be adaptively adjusted by actually adjusting the tungsten needle movement trajectory and / or welding power supply to ensure that the weld quality is not affected by the change of the workpiece groove width, thus ensuring the stability and reliability of the welding quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 This is a flowchart of an embodiment of the method for real-time monitoring of TIG welding quality based on a molten pool camera of the present application.

[0030] Figure 2 This is a flowchart of calibrating data information in an image obtained by a molten pool camera by a tungsten needle in one embodiment of the method for real-time monitoring of TIG welding quality based on a molten pool camera of the present application.

[0031] Figure 3 This is a schematic diagram of the groove of the workpiece to be welded in one embodiment of the method for real-time monitoring of TIG welding quality based on a molten pool camera of the present application.

[0032] Figure 4 This is a flowchart of adjusting the position of the tungsten needle before welding in one embodiment of the method for real-time monitoring of TIG welding quality based on a molten pool camera of the present application.

[0033] Figure 5 This is a schematic diagram of the position status of the tungsten needle and the groove of the workpiece to be welded in one embodiment of the method for real-time monitoring of TIG welding quality based on a molten pool camera of the present application.

[0034] Figure 6 This is a schematic diagram of a welding area image captured by a molten pool camera in one embodiment of the method for real-time monitoring of TIG welding quality based on a molten pool camera of the present application.

[0035] Figure 7 This is a flow chart of real-time monitoring of the state of the tungsten needle during the welding process in one embodiment of the method for real-time monitoring of TIG welding quality based on a molten pool camera of the present application.

[0036] Figure 8 This is a schematic diagram of the change in the morphology of the tungsten needle after welding in one embodiment of the method for real-time monitoring of TIG welding quality based on a molten pool camera of the present application.

[0037] Fig. 9 This is a schematic diagram of the arc shape during welding in one embodiment of the method for real-time monitoring of TIG welding quality based on a molten pool camera of the present application.

[0038] Explanation of the accompanying reference numerals: 1, workpiece; 11, groove; 2, welding gun; 21, tungsten needle; 3, welding arc; 31, standard arc. DETAILED DESCRIPTION

[0039] The specific implementation of the present application is described in detail below in conjunction with the accompanying drawings. It should be understood that the specific implementation described here is only used to illustrate and explain the present application, and is not used to limit the present application.

[0040] In the description of this application, it should be noted that, unless otherwise clearly specified and limited, the terms "disposed" and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a direct connection, or an indirect connection through an intermediate medium, it can be the internal connection of two elements or the interaction relationship between two elements. For ordinary technicians in this field, the specific meanings of the above terms in this application can be understood according to specific circumstances.

[0041] An embodiment of the method for real-time monitoring of TIG welding quality based on a molten pool camera of the present application is as follows: Figure 1 As shown, the following steps are included: S100. Obtain an image of the welding area through a molten pool camera. The molten pool camera can use various industrial cameras that can effectively image high temperature and high brightness areas. An anti-high temperature protection lens and an infrared filter can be installed on the molten pool camera to reduce the impact of high temperature on the molten pool camera and the imaging results of the molten pool camera.

[0042] S200, identifying the tungsten needle 21 on the welding gun 2 in the image through machine vision software, and identifying the diameter direction and length direction of the tungsten needle 21 according to the shape and direction of the tungsten needle 21, extracting the imaging diameter D1 and imaging length L1 of the tungsten needle 21 in the image, and calibrating the imaging parameters of the image according to the actual diameter and actual length of the tungsten needle before welding, that is, the ratio of the size in the image after imaging to the actual size, so that the actual size of the object can be accurately obtained according to the size of the image of the object in the image.

[0043] Generally, the imaging focal length and imaging distance of the melt pool camera will affect the ratio of the image size to the actual size. When the lens of the melt pool camera is tilted at an angle to the welding plane, the ratio of the image size to the actual size of the object in different directions in the image will be different to a certain extent. Therefore, it is necessary to calibrate the imaging parameters of the melt pool camera so that the actual size of the object can be accurately restored according to the image size of the object in different directions in the image.

[0044] The workpiece 1 and the groove 11 for welding on the workpiece 1 in the image are identified, and the actual size of the groove 11 is obtained through the imaging size of the groove 11 in the image.

[0045] S300, determine the relative position between the end of the tungsten needle 21 and the groove 11 on the workpiece 1 according to the calibration result of the previous step, including but not limited to the position of the end of the tungsten needle 21 in the length direction of the groove 11, the distance between the end of the tungsten needle 21 and the bottom of the groove 11, the distance between the end of the tungsten needle 21 and the two sides of the groove 11, etc. And according to the relative position between the end of the tungsten needle 21 and the groove 11 on the workpiece 1, determine whether it is necessary to adjust the position of the tungsten needle 21 before welding, so that the end of the tungsten needle 21 is in a suitable initial welding position, which is conducive to the smooth and efficient welding process.

[0046] S400, input inert shielding gas, such as argon, into the welding area, start the power supply of the welding gun 2, generate a welding arc 3 between the end of the tungsten needle 21 and the groove 11 on the workpiece 1, and input solder into the welding area on the groove 11. The welding arc 3 melts the solder and the workpiece material adjacent to the groove 11 to form a molten pool, and TIG welding is performed on the groove 11 of the workpiece 1. After the molten pool solidifies, the solder and the workpiece material are mixed and connected between the workpieces 1 on both sides of the groove 11 to form a weld.

[0047] During the welding process, the welding area is continuously photographed by the molten pool camera installed on the welding gun 2 to obtain a real-time image of the welding area during the welding process. The real-time image is usually transmitted to a control system, such as an industrial computer, in real time, and the real-time image is recognized and processed using machine vision software. The machine vision software can be compiled by personnel in related fields according to existing technologies, or the corresponding commercial software can be purchased from outside. This application will not further explain its working principle in detail.

[0048] S500, the data such as the diameter, length, end shape, and distance between the end and the groove 11 of the tungsten needle 21 obtained from the real-time image; the data such as the groove depth, groove width, width of both sides, and side inclination angle of the workpiece 11 in the welding area; the data such as the shape, area size, and texture of the molten pool; the data such as the shape of the welding arc 3, the position of the welding arc 3, the brightness of the welding arc 3, and the brightness change range, so as to monitor the TIG welding process in real time.

[0049] When the acquired data changes, the moving trajectory of the welding gun 2 can be adjusted according to the type of data that changes and the amplitude of the data change, so as to adjust the distance between the end of the tungsten needle 21 and the bottom and two sides of the groove 11, and ensure that the welding arc 3 generated by the end of the tungsten needle 21 can weld different positions of the groove 11; the voltage and current of the welding power supply can be adjusted, so as to adjust the distribution range and energy of the welding arc 3, and ensure the depth of the weld and the reliability of the welding; the delivery amount of the solder can be adjusted to ensure the effective filling of the groove 11 area by the solder; it can also be a comprehensive adjustment of the moving trajectory of the welding gun 2, the voltage and current of the welding power supply, and the delivery amount of the solder. Through the above adjustments, the welding parameters during the welding process are maintained to be adapted to the state of the groove 11 at different positions of the workpiece 1, and the stability and consistency of the weld quality are ensured.

[0050] It should be noted that the moving trajectory of the welding gun 2 mentioned in the present application refers to the trajectory formed by the relative movement between the welding gun 2 and the workpiece 1, which can be formed not only by the movement of the welding gun 2, but also by the movement of the workpiece 1 relative to the welding gun 2 when the welding gun 2 remains fixed.

[0051] When the data of the tungsten needle, workpiece groove, molten pool and welding arc obtained from the real-time image exceeds the set threshold range and cannot ensure the normal welding process, the control system controls the welding gun 2 to stop working and stop TIG welding of the workpiece 1, and at the same time issues an alarm message to prompt the staff to intervene and perform corresponding processing according to the data type that exceeds the threshold range. In this way, welding defects can be prevented during the welding process and the reliability of welding quality can be ensured during the welding process.

[0052] In some embodiments of the method for real-time monitoring of TIG welding quality based on a molten pool camera of the present application, as Figure 2 As shown in step S200, the method for calibrating the imaging parameters of the image according to the initial diameter and length of the tungsten needle specifically includes the following steps: S210, extracting the number of pixels of the tungsten needle 21 image in the width direction, that is, the diameter direction, that is, the pixel value W of the diameter of the tungsten needle 21 from the image taken by the molten pool camera.

[0053] S220. According to the actual diameter D of the tungsten needle 21 and the pixel value W of the diameter of the tungsten needle 21 in the image, the length value in the real space corresponding to each pixel in the diameter direction of the tungsten needle 21 is calculated using the formula Scale=D / W, also called pixel length Scale.

[0054] S230 , extracting the number of pixels of the image of the tungsten needle 21 in the length direction from the image, that is, the pixel value W1 of the length of the tungsten needle 21 .

[0055] S240, using the formula L1=Scale*W1, calculate the imaging length L1 of the tungsten needle 21 in the real space corresponding to the pixel value of the length of the tungsten needle 21. The imaging length L1 of the tungsten needle 21 represents the length of the tungsten needle 21 projected on the focal plane of the molten pool camera when the molten pool camera is taking pictures.

[0056] S250. According to the actual length L of the tungsten needle 21 and the imaging length L1 of the tungsten needle 21, the projection angle of the tungsten needle 21 on the focal plane of the molten pool camera is calculated using the formula cosθ=L1 / L, that is, the installation angle θ of the molten pool camera relative to the tungsten needle 21.

[0057] According to the installation angle θ of the molten pool camera, the actual lengths corresponding to the pixels in different directions in the image are calibrated to ensure that the length data in different directions extracted from the image are consistent with the actual data in the real space, thereby ensuring the reliability of various monitoring numerical controls extracted from the images obtained by the molten pool camera.

[0058] In a preferred embodiment of the method for real-time monitoring of TIG welding quality based on a molten pool camera of the present application, before step S210, step S205 is performed first, and a visual reference is set with the installation point of the tungsten needle 21 on the welding gun 2, and a rectangular coordinate system is established with the installation point as the center of the circle, the installation surface of the welding gun 2 as the X-axis, and the tungsten needle 21 as the Y-axis, and the size of each object in the image and the distance between the corresponding points are calculated through the coordinates of each pixel in the image.

[0059] like Figure 2 As shown, after step 250, the following steps are also included: S260. Select corresponding points on the groove 11 that are easy to calculate from the image, obtain the coordinates of each point in the rectangular coordinate system, adjust the Y value of the coordinate according to the installation angle θ of the molten pool camera, recalibrate the coordinates of each point, calculate the pixel value of the groove depth and the pixel value of the groove width of the groove 11 on the workpiece 1 according to the endpoint coordinates, and then use the pixel length Scale value to calculate the groove depth and groove width of the groove 11.

[0060] S270, compare the groove depth and groove width of the groove 11 calculated from the image with the actual groove depth and actual groove width of the groove 11 on the actual workpiece 1. If the difference exceeds the welding accuracy requirement, return to step S205, adjust the visual reference and coordinate system, and on this basis, adjust the calibration results of the pixel length Scale and the molten pool camera installation angle θ until the verification error meets the accuracy requirement.

[0061] In some embodiments of the method for real-time monitoring of TIG welding quality based on a molten pool camera of the present application, as Figures 3 to 5 As shown, step S300 specifically includes the following steps: S310 , obtaining the groove depth, slope bottom width, groove width, and slope lengths on both sides of the groove 11 on the workpiece 1 from the image, and obtaining position information of the groove 11 on the workpiece 1 .

[0062] S320, identifying the position of the end of the tungsten needle 21 in the image, extracting the orientation of the end of the tungsten needle 21 at the groove 11, and calculating the distance between the end of the tungsten needle and the bottom of the groove 11 and the distance from the slopes on both sides of the groove 11 according to the image imaging parameters calibrated in S200.

[0063] S330, according to the groove 11 data information such as groove depth, groove bottom width, groove width and length of both sides of the groove 11, and the actual position of the end of the tungsten needle 21 at the groove 11, the movement of the tungsten needle 21 is controlled. If it is determined according to the data information of the groove 11 that the welding arc 3 generated by the end of the tungsten needle 21 can weld the bottom and both sides of the groove 11 at the same time, the tungsten needle 21 is controlled to move to a position where the distance from the groove bottom and both sides of the groove 11 is equal and is located at a suitable welding distance; if the welding arc 3 generated by the end of the tungsten needle 21 cannot weld the bottom and both sides of the groove 11 at the same time, the tungsten needle 21 is moved to the initial position where the distance from the end to the groove bottom and one side of the slope is located at a set distance suitable for welding.

[0064] In some embodiments of the method for real-time monitoring of TIG welding quality based on a molten pool camera of the present application, in step S400, when a real-time image of the welding area during welding is obtained by the molten pool camera, a monochromatic light source is set on the molten pool camera or on the side of the molten pool camera. The monochromatic light source can use various light sources that can emit pure monochromatic light, such as a multi-point divergent laser light source and an LED light source. The welding area is illuminated by a monochromatic light source.

[0065] A monochromatic filter with the same wavelength as the monochromatic light source is set on the lens of the molten pool camera or in front of the lens, and the monochromatic filter can filter out all wavelengths of light except the wavelength of the monochromatic light emitted by the monochromatic light source. In this way, the light that can enter the molten pool camera for imaging is mainly the illumination light of the monochromatic light source reflected by the objects in the welding area, while the high-brightness light from the welding arc 3 and the high-brightness light generated by the welding arc 3 reflected by the objects in the welding area are mostly filtered out by the filter, so the image taken by the molten pool camera is no longer overexposed due to the high-brightness light of the welding arc 3, which effectively improves the uniformity of image exposure and the clarity of the image.

[0066] The molten pool camera uses the method of this embodiment to take an image of the welding area as shown in FIG. Figure 6 As shown, it can be seen that the image no longer shows the overexposed image of the welding arc 3. After removing the interference of the ultra-high brightness light formed by the welding arc 3, the imaging of the tungsten needle 21 and the groove 11 in the welding area is clearer, so that the edge contours of the tungsten needle 21 and the groove 11 can be effectively extracted. Since the light generated by the welding arc 3 is a mixed color light composed of light of various wavelengths, it also contains a small amount of light with the same wavelength as the monochromatic light source. These lights can also pass through the monochromatic filter to enter the molten pool camera for imaging, so the shape of the welding arc 3 can also be clearly displayed in the image. However, most of the light generated by the welding arc 3 has been filtered out, which greatly reduces the exposure of the welding arc 3 in the image and no longer affects the clear display of other nearby structures.

[0067] In some embodiments of the method for real-time monitoring of TIG welding quality based on a molten pool camera of the present application, in step S500, the method for obtaining tungsten needle data from a real-time image includes the following steps: S511, identifying the position of the tungsten needle 21 from the real-time image taken by the molten pool camera, distinguishing the image of the tungsten needle 21 from the images of other surrounding structures, and extracting the edge contour line of the tungsten needle 21 along the edge of the image of the tungsten needle 21.

[0068] S512, distinguishing the needle bar portion and the end portion in the edge contour line of the tungsten needle. Generally, the portion of the edge contour line of the tungsten needle 21 where the line segments extend in parallel is identified as the needle bar portion, and the portion where the edge contour line changes from parallel to approaching is identified as the end portion.

[0069] S513, calculating the distance between the edge lines on both sides of the needle rod part to obtain the diameter of the tungsten needle 21.

[0070] The end point of the end of the tungsten needle 21 is located, and a set length is extended from the end point toward the needle rod direction, and the contour length of the tungsten needle 21 within the length range is extracted, which is called the end contour length.

[0071] S514. Display the diameter value of the tungsten needle on the output display screen.

[0072] The end contour length of the tungsten needle 21 is compared with the set end length threshold. If the end contour length of the tungsten needle 21 is not greater than the set end length threshold, information indicating that the end shape of the tungsten needle 21 is normal is displayed on the output display screen; if the end contour length of the tungsten needle 21 is greater than the set end length threshold, it is determined that the tungsten needle data exceeds the set threshold range, welding is stopped, and an abnormal tungsten needle end shape alarm message is output on the output display screen, and an audible and visual alarm signal can also be generated by the alarm device.

[0073] The workflow diagram of the method for real-time monitoring of TIG welding quality based on a molten pool camera in this embodiment is as follows: Figure 7 When TIG welding is initially performed, the end of the tungsten needle 21 installed on the welding gun 2 is in a regular conical shape, and the end contour lines extending from the end point of the tungsten needle 21 to the needle rod direction for a set length on both sides to the end point are all straight lines, as shown in FIG. Figure 8 As shown in Figure a, the length of the end contour line is the shortest. After a period of TIG welding, under the action of the welding arc 3 and the high temperature generated by it, the end of the tungsten needle 21 will be deformed by wear or balling. The deformation of the end of the tungsten needle 21 will have a significant impact on the generation of the welding arc 3 and the welding effect of the welding arc 3, and the deformation of the end of the tungsten needle 21 will also cause the end contour line of the tungsten needle 21 to lengthen. The end length threshold can be set according to the deformation type of the end of the tungsten needle 21, or a same end length threshold can be set. Once the length of the end contour line of the tungsten needle 21 is greater than the end length threshold, it can be considered that the tungsten needle 21 can no longer meet the TIG welding quality requirements. In order to ensure the welding quality, it is necessary to stop welding and replace the tungsten needle 21.

[0074] In a preferred embodiment of the method for real-time monitoring of TIG welding quality based on a molten pool camera of the present application, in step S500, the method for obtaining workpiece groove data from a real-time image specifically includes the following steps: S521, extracting the surface contour line of the upper groove 11 of the workpiece 1 in the real-time image. The surface contour line is the intersection line between different surfaces on the workpiece 1. The upper and lower edges of the side surfaces of the upper groove 11 of the workpiece 1 can be identified through the surface contour line. In the case where the lower edges of the side surfaces of the groove 11 do not intersect, or in the case where a layer of solder has been welded and filled at the bottom of the groove 11, the edges of the bottom surface of the groove 11 can also be identified.

[0075] S522, select the corresponding points on the upper contour line of the sides of the groove 11 as the reference, and calculate the groove width of the groove 11 at the corresponding part of the workpiece 1; select the corresponding points on the lower contour line of the sides of the groove 11 as the reference, and calculate the bottom width of the groove 11 at the corresponding part; select the corresponding points on the upper and lower contour lines of the same side of the groove 11 as the reference, and calculate the side width of the groove 11 at the corresponding part. Usually, when detecting the same detection object, multiple groups of corresponding points are selected on the corresponding contour line for multiple detections, and the average value of the multiple detection results is taken as the final detection result for processing.

[0076] The calculation results are transmitted to the output display screen and displayed at the set position on the output display screen.

[0077] S523, extract the end position of the tungsten needle 21 in the image, and calculate the distance between the end of the tungsten needle 21 and the bottom of the groove 11 on the workpiece 1 and the distance between the end of the tungsten needle 21 and the sides of the groove 11 according to the calibration results. When the groove width, bottom width or side direction of the groove 11 at different positions of the workpiece 1 changes, resulting in a change in the distance between the end of the tungsten needle 21 and the bottom or side of the groove 11, the tungsten needle 21 is controlled to change its moving trajectory in real time to ensure that the end of the tungsten needle 21 maintains a suitable distance from the bottom of the groove 11 and the side of the groove 11.

[0078] When changing the moving trajectory of the tungsten needle 21 causes the distance between the end of the tungsten needle 21 and the bottom and side of the groove 11 to change, the voltage and current of the welding power supply can also be adjusted accordingly to ensure that an effective welding arc 3 can be formed between the tungsten needle 21 and the groove 11.

[0079] As a specific implementation of the method for real-time monitoring of TIG welding quality based on a molten pool camera of the present application, in step S523, controlling the moving trajectory of the tungsten needle according to the distance between the end of the tungsten needle and the bottom and two side surfaces of the workpiece groove includes the following steps: 1. According to the groove width, bottom width and two side widths of the groove 11 at the corresponding position of the workpiece 1, when the end of the tungsten needle 21 is at a position where the distances from the bottom of the groove and the two side surfaces are equal, the distance between the end of the tungsten needle 21 and the bottom of the groove is calculated, and this distance is called the equidistant distance.

[0080] 2. The maximum distance at which the bottom and both side surfaces of the groove 11 can be welded simultaneously is set as the set maximum distance, and the equidistant distance is compared with the set maximum distance. According to the comparison result, the third or fourth operation is performed.

[0081] 3. If the equidistant distance is not greater than the set maximum distance, control the tungsten needle 21 to move to a position equal to the distance between the bottom and both sides of the groove, and control the voltage of the welding power supply according to the size of the equidistant distance to ensure that a welding arc 3 can be formed between the end of the tungsten needle 21 and the bottom and both sides of the groove 11, and weld the bottom and both sides of the groove 11 at the same time, and control the solder delivery amount to ensure that the solder meets the filler requirements when welding the groove 11.

[0082] The tungsten needle 21 is controlled to move along the length direction of the groove 11 , and the welding requirement of the groove 11 is met through the unidirectional movement of the tungsten needle 21 .

[0083] 4. If the equidistant distance is greater than the set maximum distance, the welding arc 3 generated by the tungsten needle 21 cannot weld the bottom and both sides of the groove 11 at the same time. The control system controls the tungsten needle 21 to move to the set distance position at which the welding arc 3 can be formed at the end of the tungsten needle 21 from the bottom and one side of the groove for a better welding effect. The control system controls the tungsten needle 21 to move along the length direction of the groove 11 while reciprocating in a direction perpendicular to the length of the groove 11, that is, in the width direction of the groove 11.

[0084] By reciprocating the tungsten needle 21 in the width direction of the groove 11, reciprocating welding can be performed between the side surface on one side of the groove 11 and the side surface on the other side, and the solder is welded between the side surfaces and the bottom surface of the groove 11 to form a fill weld covering the entire range of the groove 11, thereby ensuring the overall quality of the weld.

[0085] In some embodiments of the method for real-time monitoring of TIG welding quality based on a molten pool camera of the present application, in step S500, the method for obtaining molten pool data and arc data from the real-time image is: 1. Identify the image of the welding arc from the real-time image, extract the contour of the welding arc, and obtain the arc shape; extract the imaging brightness in the middle of the arc image and use it as the arc brightness.

[0086] The image of the molten pool is identified from the real-time image, and the contour line of the edge of the molten pool image and the texture line of the light and dark changes inside the molten pool are extracted.

[0087] Second, compare the acquired arc shape with the standard arc shape stored in the control system to obtain the deviation value of the arc shape. Specifically, Fig. 9As shown, there is usually a deviation between the contour line of the welding arc 3 and the contour line of the standard arc 31. For example, at a certain Y-axis coordinate position, the X-axis coordinate of the contour line of one side edge of the welding arc 3 is l, while the X-axis coordinate of the contour line of the same side edge of the standard arc 31 is l1, and there is a deviation l between the two. The deviation l at different Y-axis coordinate positions is statistically analyzed to obtain the deviation value between the arc shape of the welding arc 3 and the standard shape of the standard arc 31. The size of the deviation value reflects the size of the difference in shape between the welding arc 3 and the standard arc 31. When the difference exceeds a certain value, it can be considered that the shape of the welding arc 3 can no longer guarantee the quality of TIG welding, and the deviation value is selected as the arc shape deviation threshold.

[0088] Compare the arc shape of the welding arc 3 in the real-time image with the arc shape deviation value of the standard arc 31 and the arc shape deviation threshold. When the arc shape deviation value is greater than the arc shape deviation threshold, it is judged that the welding arc data exceeds the set threshold range, the welding process is stopped and an alarm message is issued.

[0089] 3. The brightness of the arc usually represents the amount of energy contained in the arc. The amount of arc energy and its stability have a great impact on the quality of TIG welding. In order to ensure the normal progress of TIG welding, the arc brightness of welding arc 3 needs to be kept within the set range and the arc brightness needs to be stable. The brightness range is set as the brightness threshold range, and the arc brightness change range that can ensure the normal progress of welding is set as the brightness change threshold.

[0090] The arc brightness of the welding arc 3 extracted from the real-time image is compared with the brightness threshold range, and the change range of the arc brightness within the set time period is recorded, and then the recorded arc brightness change range is compared with the brightness change threshold. When the recorded brightness change range is greater than the brightness change threshold, or the arc brightness of the welding arc 3 extracted from the real-time image exceeds the brightness threshold range, it is judged that the arc data exceeds the set threshold range. At this time, the welding is stopped and the abnormal alarm information of the welding process is output.

[0091] 4. When the deviation value of the arc shape, the arc brightness and the arc brightness change amplitude do not exceed the corresponding threshold range, the welding power supply is adjusted according to the deviation value of the arc shape and the deviation value between the real-time arc brightness and the set standard brightness. The distance between the end of the tungsten needle 21 and the side surfaces of the groove 11 can also be adjusted according to the deviation value of the arc shape.

[0092] When the arc shape is too small, it is usually necessary to increase the voltage value of the welding power supply, and when the arc shape is too large, it is usually necessary to reduce the voltage value of the welding power supply; when the arc shape deviates to one side of the groove 11, it is usually necessary to increase the distance between the end of the tungsten needle 21 and the side of the groove. When the arc brightness is too large, it is usually necessary to reduce the current of the welding power supply; when the arc brightness is too small, it is usually necessary to increase the current of the welding power supply.

[0093] 5. Compare the molten pool morphology obtained from the real-time image with the standard molten pool morphology stored in the control system to obtain the deviation value of the molten pool morphology. The specific method of obtaining the molten pool morphology is similar to the method of obtaining the arc morphology of welding arc 3. Compare the deviation value of the molten pool morphology with the molten pool morphology deviation threshold that can ensure the welding quality. When the molten pool morphology deviation value is greater than the molten pool morphology deviation threshold, it is judged that the molten pool data exceeds the set threshold range, the welding is stopped, and the abnormal alarm information of the welding process is output.

[0094] 6. Extract the melt pool texture data from the multiple light and dark texture lines in the melt pool extracted from the real-time image. The melt pool texture data can be one or more of the various texture feature data such as the tortuous shape of the texture line, the length of the texture line, the distribution density of the texture line, the direction of the texture line, and the brightness difference on both sides of the texture line. It can also be comprehensive feature information obtained through artificial intelligence technologies such as deep learning. Compare the acquired real-time melt pool texture with the texture data of the standard melt pool texture to obtain the deviation value of the melt pool texture data. Then compare the deviation value of the melt pool texture between the real-time melt pool and the standard melt pool with the maximum melt pool texture deviation threshold that can ensure normal welding. When the melt pool texture deviation value is greater than the melt pool texture deviation threshold, it is judged that the melt pool data exceeds the set threshold range, the welding is stopped, and the abnormal alarm information of the welding process is output.

[0095] When the molten pool texture deviation value is not greater than the molten pool texture deviation threshold, the welding power supply is adjusted according to the molten pool morphology deviation value, and the specific adjustment direction can be determined according to the results of deep learning.

[0096] An embodiment of a welding device for real-time monitoring of TIG welding quality based on a molten pool camera of the present application includes all necessary structures required to implement the method for real-time monitoring of TIG welding quality based on a molten pool camera of any embodiment of the present application, and can implement the method for real-time monitoring of TIG welding quality based on a molten pool camera of any embodiment of the present application.

[0097] In the description of the present application, the description with reference to the terms "one embodiment", "specific embodiment", "preferred embodiment", etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In the present application, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0098] The above are all preferred embodiments of the present application, and the protection scope of the present application is not limited thereto. Therefore, any equivalent changes made according to the structure, shape, and principle of the present application should be included in the protection scope of the present application.

Claims

1. A method for real-time monitoring of TIG welding quality based on a molten pool camera, characterized in that: The steps include: S100, obtaining an image of the welding area through a molten pool camera; S200, identifying the tungsten needle and the groove of the workpiece in the image, and calibrating the imaging parameters of the image according to the actual diameter and actual length of the tungsten needle; S300, obtaining the relative position of the end of the tungsten needle and the groove of the workpiece, and adjusting the initial position of the tungsten needle accordingly; S400, TIG welding is performed on the groove of the workpiece, and the real-time image of the welding process is obtained through the molten pool camera; S500, acquiring data of the tungsten needle, workpiece groove, molten pool and welding arc from the real-time image, controlling the position status of the tungsten needle and the workpiece groove and / or the working status of the welding power supply in real time according to the acquired data, and stopping welding and issuing an alarm message when the data exceeds a set threshold range.

2. The method for real-time monitoring of TIG welding quality based on a molten pool camera according to claim 1 is characterized in that: In step S200, the method for calibrating the imaging parameters of the image according to the initial diameter and length of the tungsten needle includes the following steps: S210, obtaining a pixel value W of the tungsten needle in the image in the diameter direction; S220, using the formula Scale=D / W to calculate the length value Scale corresponding to each pixel in the corresponding direction in the image, where D is the diameter of the tungsten needle; S230, obtaining a pixel value W1 of the tungsten needle in the length direction in the image; S240, calculating the imaging length L1 of the tungsten needle using the formula L1=Scale*W1; S250, using the formula cosθ=L1 / L to calculate the installation angle θ between the molten pool camera and the tungsten needle, and use this to calibrate the actual length corresponding to the pixels in different directions of the image.

3. The method for real-time monitoring of TIG welding quality based on a molten pool camera according to claim 2 is characterized in that: After step 250, the following steps are also included: S260, calculating the groove depth and groove width of the workpiece according to the installation angle θ and the pixel length Scale value of the molten pool camera; S270. Verify and adjust the calibration result according to the actual groove depth and actual groove width of the workpiece.

4. The method for real-time monitoring of TIG welding quality based on a molten pool camera according to claim 1 is characterized in that: Step S300 includes the following steps: S310, obtaining the depth, bottom width, width of the groove and length of the slopes on both sides of the workpiece groove; S320, obtaining the position of the end of the tungsten needle at the groove of the workpiece, and calculating the distance between the end of the tungsten needle and the bottom of the groove and the slopes on both sides; S330, controlling the movement of the tungsten needle so that the tungsten needle moves to an initial position where the end of the tungsten needle is a set distance from the bottom of the groove and at least one side of the slope.

5. The method for real-time monitoring of TIG welding quality based on a molten pool camera according to claim 1 is characterized in that: In step S400, a monochromatic light source is used to illuminate the welding area, and a filter with the same wavelength as the monochromatic light source is set in front of the molten pool camera to obtain a real-time image formed by the monochromatic light source during the welding process.

6. The method for real-time monitoring of TIG welding quality based on a molten pool camera according to claim 1 is characterized in that: In step S500, the method for acquiring tungsten needle data from a real-time image includes the following steps: S511, extracting the edge contour line of the tungsten needle in the real-time image; S512, distinguishing the needle shaft part and the end part in the edge contour line of the tungsten needle; S513, calculating the distance between the edge lines on both sides of the needle rod part, that is, the diameter of the tungsten needle, and the length of the inner contour line of the set length of the end of the tungsten needle; S514, output the tungsten needle diameter value, and compare the length of the tungsten needle end contour line with the set end length. When the length of the tungsten needle end contour line is greater than the set end length, it is determined that the tungsten needle data exceeds the set threshold range.

7. The method for real-time monitoring of TIG welding quality based on a molten pool camera according to claim 6 is characterized in that: In step S500, the method for acquiring workpiece groove data from a real-time image includes the following steps: S521, extracting the surface contour line of the groove portion of the workpiece in the real-time image; S522, respectively calculating the groove width, bottom width and two side widths of the groove of the workpiece, and outputting the detection results; S523, calculating the distance between the end of the tungsten needle and the bottom and two side surfaces of the groove of the workpiece according to the end position of the tungsten needle in the image, and controlling the moving trajectory of the tungsten needle according to the calculation result.

8. The method for real-time monitoring of TIG welding quality based on a molten pool camera according to claim 7 is characterized in that: In step S523, the method of controlling the moving track of the tungsten needle according to the distance between the end of the tungsten needle and the bottom and two side surfaces of the workpiece groove includes the following steps: According to the groove width, bottom width and two side widths of the workpiece groove, calculate the equidistant distance when the end of the tungsten needle is at the same distance from the groove bottom and two side surfaces; Compare the equidistant distance with the set maximum distance; If the equidistant distance is not greater than the set maximum distance, the tungsten needle is controlled to move to a position equal to the distance between the bottom and the two sides of the groove, the voltage of the welding power source is controlled according to the size of the equidistant distance, and the tungsten needle is controlled to move along the length direction of the groove; If the equidistant distance is greater than the set maximum distance, move the end of the tungsten needle to a position where the distance from the bottom and one side of the groove is the set distance, and control the tungsten needle to move back and forth in a direction perpendicular to the length of the groove while moving along the length of the groove.

9. The method for real-time monitoring of TIG welding quality based on a molten pool camera according to claim 1 is characterized in that: In step S500, the method for acquiring molten pool data and arc data from a real-time image comprises the following steps: Extract the shape and brightness of the welding arc, as well as the shape and texture of the molten pool from real-time images; The arc shape is compared with the standard arc shape to obtain the deviation value of the arc shape, and when the arc shape deviation value is greater than the arc shape deviation threshold, it is determined that the data is abnormal; The brightness of the welding arc is compared with the set brightness threshold, the change range of the arc brightness within the set time period is extracted, and the change range is compared with the set brightness change threshold. When the brightness change range is greater than the brightness change threshold, or the arc brightness is greater than the brightness threshold, it is determined that the welding arc data exceeds the set threshold range; Adjust the welding power supply according to the arc shape deviation value and the deviation value between the average brightness and the standard brightness; The molten pool morphology is compared with the standard molten pool morphology to obtain a deviation value of the molten pool morphology, and when the molten pool morphology deviation value is greater than the molten pool morphology deviation threshold, it is determined that the data is abnormal; The melt pool texture is compared with the standard melt pool texture to obtain a deviation value of the melt pool texture, and when the melt pool texture deviation value is greater than the melt pool texture deviation threshold, it is determined to be data abnormal; Adjust the welding power supply according to the molten pool morphology deviation value.

10. A welding device for real-time monitoring of TIG welding quality based on a molten pool camera, characterized in that: A method for real-time monitoring of TIG welding quality based on a molten pool camera according to any one of claims 1-9.

Images