Welding device for aluminum alloy door and window machining and welding method thereof

Abstract

The invention discloses a welding device for aluminum alloy door and window machining, and relates to the field of aluminum alloy door and window machining equipment.The welding device comprises a welding table, a sliding rail is arranged on one side of the top end of the welding table, a first electric control telescopic rod is arranged at one end of the sliding rail, a mechanical arm is arranged at the top of the sliding rail, and a welding gun is arranged at one end of the mechanical arm; according to the welding device for aluminum alloy door and window machining, welding seam image features are extracted through the recognition operation mechanism by means of the neural convolutional network, a welding seam area is positioned and amplified through the super-pixel technology, and precise recognition of the width of a welding seam is achieved. And when the welding seam is not recognized, the electric control telescopic rod drives the camera to traverse the surface of the workpiece to ensure that the welding seam is positioned without omission. When the gap is detected to exceed the standard, the servo motor drives the gear to transmit, the guide shaft and the permanent magnet are driven to knock a defect area, impurity welding slag is removed, a clean environment is provided for repair welding, and the continuity and accuracy of welding quality control are guaranteed.

Description

Technical Field

[0001] This invention relates to the field of aluminum alloy door and window processing equipment technology, specifically to a welding device and welding method for aluminum alloy door and window processing. Background Technology

[0002] In the welding process of aluminum alloy doors and windows, the welding of the door and window support pipes and the outer frame requires strict control of the weld gap width to ensure structural strength. However, in actual operation, the weld gap width often exceeds the standard due to workpiece positioning deviation and welding gun trajectory deviation. At the same time, in the industry's common welding process, the positioning and width detection of the weld area mostly rely on manual visual inspection, which has the problems of low detection efficiency and large error. Moreover, there is a lack of an automatic mechanism to traverse the workpiece surface, which easily leads to missed weld inspections, thereby affecting the structural stability and assembly accuracy of the finished doors and windows.

[0003] According to CN114083191B, the welding device and its welding method adopt a combination of a welding nozzle lifting control mechanism, a feeding mechanism and a welding nozzle opening and closing mechanism. The welding nozzle assembly is driven to move up and down to get closer to or away from the welding product through a second power mechanism. At the same time, the relative positions of the left and right welding nozzles are adjusted by the welding nozzle opening and closing mechanism to achieve positional adaptation between the welding nozzle and the welding product, thereby meeting the positional requirements of the welding operation and improving the basic welding quality.

[0004] However, the above technical solutions do not include an automatic identification and traversal detection mechanism for the weld area, making it impossible to autonomously locate the weld position. If the weld area on the workpiece surface is not precisely guided by human intervention, misalignment between the weld tip and the weld is likely to occur. At the same time, the lack of image recognition-based weld width detection function means that weld quality inspection still relies on manual labor.

[0005] Therefore, in order to address the existing shortcomings, we conducted research and improvements and proposed a welding device for processing aluminum alloy doors and windows. Summary of the Invention

[0006] The purpose of this invention is to provide a welding device and welding method for processing aluminum alloy doors and windows, so as to solve the problems mentioned in the background art.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a welding device for processing aluminum alloy doors and windows, comprising: a welding table, a slide rail provided on one side of the top of the welding table, a first electrically controlled telescopic rod provided at one end of the slide rail, a robotic arm provided on the top of the slide rail, a welding gun provided at one end of the robotic arm, a door and window outer frame provided on the other side of the top of the welding table, and multiple door and window support pipes arranged laterally on the inner side of the door and window outer frame; A locking plate is provided at the intersection of the door and window branch pipe and the outer frame of the door and window. A placement frame is provided on the outside of the locking plate. Flip frames are provided at the front and rear ends of the placement frame. An adapter bearing is provided at the connection between the flip frame and the placement frame. A latch is provided between the locking plate and one end of the door and window branch pipe. An identification operation mechanism is provided at the middle of the top of the welding table. The identification operation mechanism includes a Y-guide groove, which is located at the middle of the top of the welding table. A heat dissipation mechanism is provided on one side of the welding torch. The heat dissipation mechanism includes a single air outlet pipe, which is horizontally aligned on one side of the welding torch. An auxiliary mechanism is provided at the bottom of one side of the door and window branch pipe. The auxiliary mechanism includes a lifting arc block, which is located on one side of the bottom of the door and window branch pipe.

[0008] Furthermore, the identification operating mechanism also includes a pivot, servo motor, base column, push rod, transverse groove, third electrically controlled telescopic rod, half gear, iron plate, limiter, fixing plate, vertical crank arm shaft, adapter back plate, permanent magnet, guide shaft, limit piece, spring, multi-hole row plate, sliding support column, identification camera, side panel, adapter mounting frame, X-axis guide rail, and second electrically controlled telescopic rod. The third electrically controlled telescopic rod is located inside the Y-axis guide groove. A sleeve block is located at one end of the third electrically controlled telescopic rod, and an X-axis guide rail is located on the top of the sleeve block. A transverse groove is located inside the X-axis guide rail, and a second electrically controlled telescopic rod is located on one side of the transverse groove. An adapter mounting frame is located on one side of the second electrically controlled telescopic rod, and a sliding support column is located on the top of the adapter mounting frame. An identification camera is located at the upper outer end of the sliding support column near the door / window support pipe. The upper end of the flip frame on one side of the placement frame... The system is equipped with a pivot, a servo motor on one side of the pivot, a base column at the bottom of the servo motor, a half-gear on the upper end of the flip frame on the other side of the placement frame, a full gear on one side of the half-gear, an adapter backplate on the outside of the full gear, a retaining plate on one side of the adapter backplate, a limiter on the back of the retaining plate, a push rod vertically penetrating the inside of the limiter, a vertical crank arm shaft at one end of the push rod, a full gear at one end of the vertical crank arm shaft, and an iron plate at the other end of the push rod. A side panel is located on one side of the top of the welding table, a perforated plate on one side of the top of the side panel, a guide shaft vertically penetrating the inside of the holes on the surface of the perforated plate, a spring sleeved on the upper end of the guide shaft, a limit plate at the top of the guide shaft, and a permanent magnet at the bottom of the guide shaft.

[0009] Furthermore, the identification camera is equipped with a control module, which is used to identify whether the width of the welding gap between the door and window branch pipe and the outer frame of the door and window is too large through pattern recognition technology. The pattern recognition technology adopts neural convolutional networks in deep learning combined with superpixel technology of image recognition to realize the magnification of the welding gap image and accelerate the recognition process.

[0010] Furthermore, the neural convolutional network is used to extract features from the acquired weld gap image, and the superpixel technology is used to segment the weld gap image into multiple superpixel units to achieve precise positioning and image magnification of the weld gap region.

[0011] Furthermore, the control module also includes an image acquisition control unit, which controls the recognition camera to capture images of the welding gaps between the door and window branch pipes and the outer frame of the door and window, and transmits the acquired images to the pattern recognition processing unit.

[0012] Furthermore, the control module can output a judgment signal on whether the weld gap width is too large based on the recognition result, and the judgment signal can be transmitted to the inside of the servo motor.

[0013] Furthermore, when the identification camera fails to identify the door / window branch pipe and the door / window outer frame, the control module drives the third and second electrically controlled telescopic rods to move, thereby causing the sliding support column and the adapter mounting frame to slide inside the horizontal groove and Y guide groove of the X guide rail, so that the identification camera can gradually move and traverse along the weld surface of the door / window branch pipe and the door / window outer frame.

[0014] Furthermore, the heat dissipation mechanism also includes a main control fan and an auxiliary support rod. The main control fan is installed on one side of the single air outlet pipe, and the auxiliary support rod is installed on the back of the main control fan. When the recognition camera observes that the welding torch leaves the weld seam between the door and window branch pipe and the door and window outer frame, the control module inside the recognition camera receives the observation signal and sends a control signal to the main control fan, so that the main control fan first blows air along a straight line to the surface of the door and window branch pipe and the door and window outer frame. After the blowing continues for a set time, the main control fan exhausts all the internal gas through the single air outlet pipe. At this time, the surface of the main control fan stops blowing air to the door and window branch pipe and the door and window outer frame, and the gas is blown to the surface of the welding torch by the single air outlet pipe.

[0015] Furthermore, the auxiliary mechanism also includes a support plate, a rotating shaft, and vertical shaft pieces. The bottom of the lifting arc block is provided with a support plate, the bottom of the support plate is provided with a rotating shaft, and vertical shaft pieces are provided on both sides of the rotating shaft. The vertical shaft pieces are fixedly connected to the iron plate.

[0016] A welding method for a welding device used in the processing of aluminum alloy doors and windows includes the following steps: S1: Hoist the outer frame of the door and window and the support pipe of the door and window to the designated work position of the placement rack. The limiting structure on the placement rack completes the precise positioning to ensure that the areas to be welded are aligned and fitted. Activate the locking components on both sides of the placement rack to fix the outer frame of the door and window and the support pipe of the door and window to prevent displacement during the welding process. At the same time, the control module initializes the recognition camera, the second electric telescopic rod, the third electric telescopic rod, the fan and other components to complete the self-check of the equipment standby status. S2: The camera is activated and uses a deep learning neural convolutional network combined with superpixel technology to perform a preliminary scan of the area where the outer frame of the door and window meets the branch pipe. If the area to be welded is not identified, the control module drives the second and third electric telescopic rods to move, causing the sliding support to slide along the X guide rail groove and Y guide groove, so that the camera can traverse the surface of the workpiece until the weld is located. After the positioning is completed, the camera transmits the weld position signal to the welding gun control system, which guides the welding gun to move to the start end of the weld and adjusts it to the appropriate welding angle. S3: The welding torch starts welding according to the preset parameters and moves at a constant speed along the weld seam trajectory to perform welding operations. During the welding process, the recognition camera continuously collects weld seam images and monitors the changes in the weld seam width in real time through pattern recognition technology. The data is transmitted to the control module in a synchronous manner. If the weld seam width is detected to exceed the preset threshold, the control module immediately sends an early warning signal, suspends the welding operation and records the abnormal position for further processing. If the welding parameters are normal, the welding work of the entire weld seam is completed. S4: After the welding torch completes welding, it leaves the weld area. The recognition camera captures the signal of the welding torch leaving and transmits it to the control module. The control module then sends a start command to the fan. The fan blows along a straight line to the welded surface of the door and window frame and the branch pipe under the support of the support rod. During the blowing process, the airflow quickly carries away the residual heat of welding, realizing the cooling and heat dissipation of the workpiece. At the same time, it removes a small amount of welding slag and dust from the surface, providing a clean surface for subsequent inspection. The blowing time is executed according to the preset threshold of the control module. S5: After the blowing is completed, the recognition camera scans the weld seam again. The neural convolutional network extracts the weld seam image features and combines superpixel technology to magnify the image details. It accurately detects whether the gap width is qualified. If the detection is qualified, it proceeds to the next step. If defects such as excessive gap are detected, the control module drives the servo motor to start. The half gear and full gear mesh to drive the push rod and iron plate to move. With the help of the permanent magnet, the guide shaft is regularly tapped to remove impurities and welding slag, preparing for repair welding. S6: If a welding defect exists, the welding torch moves to the corresponding area based on the defect location identified by the recognition camera to perform repair welding. After the repair welding is completed, the control module sends a signal again to switch the fan direction, so that the fan airflow is completely directed to the surface of the welding torch through the single exhaust duct, providing targeted cooling to the welding torch and preventing high-temperature aging. After both the workpiece and the welding torch have cooled to a safe temperature, the placement rack locking assembly unlocks, and the workpiece is hoisted to the designated storage area, completing the entire welding process.

[0017] Compared with the prior art, the beneficial effects of the present invention are: 1. This invention achieves precise identification of weld gap width by using an identification mechanism, extracting weld image features with the aid of a neural convolutional network, and employing superpixel technology to locate and magnify the weld area. When no weld is detected, an electrically controlled telescopic rod drives a camera to traverse the workpiece surface, ensuring no weld is missed. If the gap exceeds the limit, a servo motor drives a gear transmission, which in turn drives a guide shaft and a permanent magnet to strike the defective area, removing impurities and slag, providing a clean environment for repair welding, and ensuring the consistency and accuracy of welding quality control. 2. This invention triggers the main control fan to start by capturing the signal of the welding torch leaving the workpiece using a recognition camera. The fan first blows on the surface of the welded workpiece to quickly dissipate residual heat and remove surface welding slag and dust, creating clean conditions for subsequent inspection. After a preset time, the fan switches the airflow direction through a single outlet duct to directionally cool the welding torch, preventing high temperatures from damaging it, extending its service life, and ensuring continuous and stable equipment operation. 3. This invention uses a lifting arc block to wrap around the outside of the door / window branch pipe. An iron plate drives the vertical shaft and support plate to move, while the lifting arc block moves synchronously across the branch pipe surface to clean away debris. The support plate can rotate around a pivot to adjust the wiping angle of the lifting arc block, ensuring thorough cleaning of the branch pipe surface. This mechanism, in conjunction with the flipping action of the placement frame, simultaneously completes the cleaning of the branch pipe, reducing additional cleaning steps, improving the overall welding process efficiency, and ensuring the smooth progress of subsequent welding operations. Attached Figure Description

[0018] Figure 1 This is a top view of the right side of the welding device for processing aluminum alloy doors and windows according to the present invention. Figure 2 The present invention relates to a welding apparatus for processing aluminum alloy doors and windows. Figure 1 A magnified structural diagram at point A; Figure 3 This is a top view of the left side of the welding device for processing aluminum alloy doors and windows according to the present invention. Figure 4 The present invention relates to a welding apparatus for processing aluminum alloy doors and windows. Figure 3 A magnified structural diagram at point B; Figure 5This is a schematic diagram of the single-sided side view of the welding device for processing aluminum alloy doors and windows according to the present invention. Figure 6 The present invention relates to a welding apparatus for processing aluminum alloy doors and windows. Figure 5 A magnified structural diagram at point C; Figure 7 This is a top view of the welding device for processing aluminum alloy doors and windows according to the present invention. Figure 8 The present invention relates to a welding apparatus for processing aluminum alloy doors and windows. Figure 7 A magnified structural diagram at point D; Figure 9 The present invention relates to a welding apparatus for processing aluminum alloy doors and windows. Figure 7 A magnified structural diagram at point E; Figure 10 This is a schematic diagram of the one-way side view of the welding device for processing aluminum alloy doors and windows according to the present invention. Figure 11 The present invention relates to a welding apparatus for processing aluminum alloy doors and windows. Figure 10 A magnified structural diagram at point F.

[0019] In the diagram: 1. Welding table; 2. First electrically controlled telescopic rod; 3. Slide rail; 4. Second electrically controlled telescopic rod; 5. X-axis guide rail; 6. Adaptor mounting frame; 7. Placement rack; 8. Side panel; 9. Main control fan; 10. Single air outlet duct; 11. Welding torch; 12. Recognition camera; 13. Robotic arm; 14. Sliding support column; 15. Y-axis guide groove; 16. Perforated plate; 17. Spring; 18. Limiting plate; 19. Door and window branch pipe; 20. Locking plate; 21. Guide shaft; 22. Door and window 23. Outer frame; 24. Permanent magnet; 25. Adaptive bearing; 26. Support rod; 27. Adaptive back plate; 28. Vertical crank arm shaft; 29. ​​Fixing plate; 30. Limiter; 31. Full gear; 32. Iron plate; 33. Half gear; 34. Third electrically controlled telescopic rod; 35. Horizontal groove; 36. Lifting arc block; 37. Support plate; 38. Push rod; 39. Base column; 40. Servo motor; 41. Pivot; 42. Flip frame; 43. Lock; 44. Rotating shaft; 45. Vertical shaft plate. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. Example

[0021] like Figures 1 to 11As shown, a welding device for processing aluminum alloy doors and windows includes: a welding table 1, a slide rail 3 is provided on one side of the top of the welding table 1, a first electrically controlled telescopic rod 2 is provided at one end of the slide rail 3, a robot arm 13 is provided on the top of the slide rail 3, a welding gun 11 is provided at one end of the robot arm 13, a door and window outer frame 22 is provided on the other side of the top of the welding table 1, and a plurality of door and window support pipes 19 are arranged horizontally on the inner side of the door and window outer frame 22. A locking plate 20 is provided at the intersection of the door and window branch pipe 19 and the door and window outer frame 22. A placement frame 7 is provided on the outside of the locking plate 20. Flip frames 41 are provided at the front and rear ends of the placement frame 7. An adapter bearing 24 is provided at the connection between the flip frame 41 and the placement frame 7. A latch 42 is provided between the locking plate 20 and one end of the door and window branch pipe 19. An identification operation mechanism is provided at the middle of the top of the welding table 1. The identification operation mechanism includes a Y guide groove 15, which is located at the middle of the top of the welding table 1. A heat dissipation mechanism is provided on one side of the welding torch 11. The heat dissipation mechanism includes a single air outlet pipe 10, which is horizontally aligned on one side of the welding torch 11. An auxiliary mechanism is provided at the bottom of one side of the door and window branch pipe 19. The auxiliary mechanism includes a lifting arc block 35, which is located on one side of the bottom of the door and window branch pipe 19. The user opens the latch 42 at one end of the locking plate 20 at the top of the placement rack 7, places the outer frame 22 of the door and window on the top of the placement rack 7, locks the latch 42 on the outside of the outer frame 22, aligns the door and window support pipe 19 with the predetermined position inside the outer frame 22, and then starts the first electrically controlled telescopic rod 2. After the first electrically controlled telescopic rod 2 is started, it drives the robot arm 13 to move outside the slide rail 3, so that the robot arm 13 drives the welding gun 11 to move on the surface of the door and window support pipe 19 and the outer frame 22. When the welding gun 11 heats up, it can start welding at the connection between the door and window support pipe 19 and the outer frame 22.

[0022] Example 1: As Figures 1 to 11As shown, the recognition operating mechanism also includes a pivot 40, a servo motor 39, a base column 38, a push rod 37, a transverse groove 34, a third electrically controlled telescopic rod 33, a half-gear 32, an iron plate 31, a limiter 29, a retaining plate 28, a vertical crank arm shaft 27, an adapter back plate 26, a permanent magnet 23, a guide shaft 21, a limit piece 18, a spring 17, a perforated plate 16, a sliding support column 14, a recognition camera 12, a side panel 8, an adapter mounting frame 6, an X-axis guide rail 5, a second electrically controlled telescopic rod 4, and a Y-axis guide rail. A third electrically controlled telescopic rod 33 is provided on the inner side of the groove 15. A sleeve block is provided at one end of the third electrically controlled telescopic rod 33. An X-guide rail 5 is provided on the top of the sleeve block. A transverse groove 34 is provided on the inner side of the X-guide rail 5. A second electrically controlled telescopic rod 4 is provided on one side of the transverse groove 34. An adapter mounting frame 6 is provided on one side of the second electrically controlled telescopic rod 4. A sliding support column 14 is provided on the top of the adapter mounting frame 6. A recognition camera 12 is provided at the upper outer end of the sliding support column 14 near the door and window support pipe 19. A placement rack 7 is located on one side. A pivot 40 is provided at the upper end of the flip frame 41. A servo motor 39 is provided on one side of the pivot 40. A base column 38 is provided at the bottom of the servo motor 39. A half gear 32 is provided at the upper end of the flip frame 41 on the other side of the placement frame 7. A full gear 30 is provided on one side of the half gear 32. An adapter back plate 26 is provided on the outside of the full gear 30. A retaining plate 28 is provided on one side of the adapter back plate 26. A limiter 29 is provided on the back of the retaining plate 28. A push rod 37 is vertically inserted through the inner side of the limiter 29. A vertical crank arm shaft 27 is provided at one end of the rod 37, and a full gear 30 is provided at one end of the vertical crank arm shaft 27. An iron plate 31 is provided at the other end of the push rod 37. The side panel 8 is provided on one side of the top of the welding table 1. A perforated plate 16 is provided on one side of the top of the side panel 8. A guide shaft 21 is vertically inserted through the holes on the surface of the perforated plate 16. A spring 17 is sleeved on the upper end of the guide shaft 21. A limit piece 18 is provided on the top of the guide shaft 21. A permanent magnet 23 is provided on the bottom of the guide shaft 21. The internal control module of the recognition camera 12 is used to identify whether the width of the welding gap between the door and window branch pipe 19 and the door and window outer frame 22 is too large through pattern recognition technology. The pattern recognition technology adopts neural convolutional network in deep learning combined with superpixel technology of image recognition to realize the magnification of the welding gap image and accelerate the recognition process. Neural convolutional networks are used to extract features from the acquired weld gap images, and superpixel technology is used to segment the weld gap images into multiple superpixel units to achieve precise localization and image magnification of the weld gap region. The control module can output a judgment signal on whether the weld gap width is too large based on the recognition result. The judgment signal can be transmitted to the inside of the servo motor 39. When the recognition camera 12 fails to recognize the door and window branch pipe 19 and the door and window outer frame 22, the control module drives the third electric telescopic rod 33 and the second electric telescopic rod 4 to move, thereby driving the sliding support column 14 and the adapter mounting frame 6 to slide inside the horizontal groove 34 and the Y guide groove 15 of the X guide rail 5, so that the recognition camera 12 can gradually move and traverse along the weld surface of the door and window branch pipe 19 and the door and window outer frame 22. When the welding torch 11 completes the welding operation between the door / window branch pipe 19 and the door / window outer frame 22, the recognition camera 12 immediately starts working. If the recognition camera 12 fails to capture the weld area between the door / window branch pipe 19 and the door / window outer frame 22, its internal control module immediately sends drive signals to the second electrically controlled telescopic rod 4 and the third electrically controlled telescopic rod 33. At this time, the second electrically controlled telescopic rod 4 drives the adapter mounting frame 6 to move laterally along the transverse groove 34 inside the X-guide rail 5, while the third electrically controlled telescopic rod 33 drives the X-guide rail 5 to move vertically along the Y-guide groove 15 through the sleeve block. The coordinated action of the two can accurately adjust the spatial position of the recognition camera 12, so that the recognition camera 12 gradually approaches and aligns with the weld surface between the door / window branch pipe 19 and the door / window outer frame 22, ensuring the effectiveness of subsequent recognition operations.

[0023] Furthermore, after the recognition camera 12 extracts weld seam image features through a neural convolutional network in deep learning and achieves precise positioning and image magnification of the weld seam area using superpixel technology, if the weld seam width exceeds a preset threshold, its internal control module immediately sends a control signal to the servo motor 39 at the top of the base column 38. After the servo motor 39 starts, it drives the pivot 40 to rotate, which in turn drives the placement frame 7 to rotate around the adapter bearings 24 on both sides of the flip frame 41. The placement frame 7 simultaneously drives the combination of the door and window outer frame 22 and the door and window branch pipe 19 to flip and adjust its posture. During this process, the other end of the placement frame 7 simultaneously drives the half gear 32 to rotate. The half gear 32 meshes with the full gear 30, driving the full gear 30 to drive the vertical crank arm shaft 27 to swing back and forth. One end of the vertical crank arm shaft 27 pushes the push rod 37 to move vertically in a straight line along the inside of the limiter 29, while the other end of the push rod 37 drives the iron plate 31 to move upward. Because the iron plate 31 is ferromagnetic, it will attract the permanent magnet 23 at the bottom of the guide shaft 21 during its ascent. The permanent magnet 23 will pull the guide shaft 21 to move downward along the inside of the holes in the porous plate 16. At this time, the spring 17 sleeved on the upper part of the guide shaft 21 will be compressed and stored. The limiting piece 18 at the top of the guide shaft 21 can effectively limit the downward stroke of the guide shaft 21 and avoid excessive displacement. When the guide shaft 21 drives the permanent magnet 23 to move down to contact the welded body of the door and window branch pipe 19 and the door and window outer frame 22, a regular knocking action will be generated, which will cause impurities and slag in the area of ​​excessive weld width to fall off from the inside of the door and window branch pipe 19 and the door and window outer frame 22, providing a clean weld environment for subsequent re-inspection or repair welding. When the weld inspection is qualified or the knocking operation is completed, the control module stops sending drive signals to the servo motor 39, the spring 17 releases its stored force, and drives the guide shaft 21 and the permanent magnet 23 to reset.

[0024] Example 2: Figures 1 to 11 As shown, the heat dissipation mechanism also includes a main control fan 9 and an auxiliary support rod 25. The main control fan 9 is installed on one side of the single air outlet duct 10, and the auxiliary support rod 25 is installed on the back of the main control fan 9. When the recognition camera 12 observes that the welding torch leaves the weld between the door and window branch pipe 19 and the door and window outer frame 22, the control module inside the recognition camera 12 receives the observation signal and sends a control signal to the main control fan 9, so that the main control fan 9 first blows the surface of the door and window branch pipe 19 and the door and window outer frame 22 in a straight line. After the blowing continues for a set time, the main control fan 9 exhausts all the internal gas through the single air outlet duct 10. At this time, the surface of the main control fan 9 stops blowing air to the door and window branch pipe 19 and the door and window outer frame 22, and the gas is blown to the surface of the welding torch 11 by the single air outlet duct 10. When the welding torch 11 completes the welding operation between the door / window branch pipe 19 and the door / window outer frame 22 and leaves the weld area, the recognition camera 12 accurately captures the departure state of the welding torch 11 through image acquisition and pattern recognition technology. Its internal control module then receives this status signal and sends a start control signal to the main control fan 9. The main control fan 9 starts running under the stable support of the support rod 25. It first blows along a straight line on the welded surface of the door / window branch pipe 19 and the door / window outer frame 22, quickly removing the residual heat after welding, thus cooling and dissipating the heat from the welded workpiece. At the same time, it can remove a small amount of welding slag and dust adhering to the surface. When the blowing operation continues for the preset time threshold, the control module sends a wind direction switching signal to the main control fan 9, causing the airflow guidance structure inside the main control fan 9 to switch and stop the airflow from the surface of the main control fan 9 to the door and window branch pipe 19 and the outer frame 22 of the door and window. At this time, all the airflow generated by the main control fan 9 is discharged through the single air outlet pipe 10 connected to it. Since the single air outlet pipe 10 and the welding torch 11 are set to be horizontally aligned, the discharged airflow can be accurately blown to the surface of the welding torch 11, quickly reducing the working temperature of the welding torch 11, avoiding the welding torch 11 from aging or being damaged due to high temperature, extending the service life of the welding torch 11, and ensuring the stability of subsequent welding operations.

[0025] Example 3: Figures 1 to 11 As shown, the auxiliary mechanism also includes a support plate 36, a rotating shaft 43, and a vertical shaft plate 44. The bottom of the lifting arc block 35 is provided with a support plate 36, the bottom of the support plate 36 is provided with a rotating shaft 43, and the two sides of the rotating shaft 43 are provided with vertical shaft plates 44. The vertical shaft plates 44 are fixedly connected to the iron plate 31. When the placement frame 7 tilts and rotates the welded body of the door and window branch pipe 19 and the outer frame 22 to one side, the lifting arc block 35 will wrap around the outside of the door and window branch pipe 19, and the iron plate 31 will drive the vertical shaft plate 44 to move upward. The vertical shaft plate 44 will drive the support plate 36 to move upward. The support plate 36 will drive the lifting arc block 35 to move outside the door and window branch pipe 19. In this way, the lifting arc block 35 will clean the surface of the door and window branch pipe 19. The user can rotate the support plate 36. At this time, the support plate 36 will drive the rotating shaft 43 to rotate inside the vertical shaft plate 44, thereby adjusting the wiping angle of the lifting arc block 35.

[0026] The core functional code in Examples 1 and 2 is presented, and the function is implemented based on Python.

[0027] import cv2 import numpy as np import time import torch import torch.nn as nn from typing import Tuple, Optional # -------------------------- 1. Hardware Abstraction Layer (simulates hardware interface, actually replaced by real driver) -------------------------- class HardwareInterface: "Hardware interface class, corresponding to the part number in the embodiment, simulating the actions of the electrically controlled telescopic pole, fan, and servo motor." def __init__(self): # Initialize part state self.telescopic_rod_4 = {"status": "stop", "position": 0} #Second electrically controlled telescopic rod self.telescopic_rod_33 = {"status": "stop", "position": 0} #Third electrically controlled telescopic rod self.fan_9 = {"status": "stop", "mode": "workpiece", "wind_speed": 0} # Fan self.servo_motor_39 = {"status": "stop", "rotation_angle": 0} # Servo motor self.single_air_pipe_10 = {"status": "closed"} # Single air outlet duct def drive_telescopic_rod_4(self, direction: str, distance: int) ->None: """Drive the second electrically controlled telescopic rod (4) to move along the X-guide rail (5) transverse groove (34)"" self.telescopic_rod_4["status"] = "running" self.telescopic_rod_4["position"] += distance if direction == "right" else -distance print(f"The second electrically controlled telescopic rod (4) moves along {direction} {distance} mm, current position: {self.telescopic_rod_4['position']} mm") def drive_telescopic_rod_33(self, direction: str, distance: int) ->None: """Drive the third electrically controlled telescopic rod (33) to move along the Y-guide groove (15)"" self.telescopic_rod_33["status"] = "running" self.telescopic_rod_33["position"] += distance if direction == "up" else -distance print(f"The third electrically controlled telescopic rod (33) moves {distance} mm along {direction}, current position: {self.telescopic_rod_33['position']} mm") def start_fan(self, mode: str, wind_speed: int = 5) -> None: """Start the blower (9), mode: workpiece - blow on the workpiece, welding_gun - blow on the welding gun (single outlet duct)"" self.fan_9["status"] = "running" self.fan_9["mode"] = mode self.fan_9["wind_speed"] = wind_speed if mode == "workpiece": self.single_air_pipe_10["status"] = "closed" print(f"Fan (9) started, mode: blowing workpiece (door and window frame 22 / branch pipe 19), wind speed: {wind_speed} level") else: self.single_air_pipe_10["status"] = "open" print(f"Fan (9) started, mode: welding torch (11), single air duct (10) opened, wind speed: {wind_speed} level") def stop_fan(self) -> None: "Stop the fan (9)" self.fan_9["status"] = "stop" self.single_air_pipe_10["status"] = "closed" print("Fan (9) has stopped running") def drive_servo_motor_39(self, angle: int, frequency: int = 2) ->None: """Drive the servo motor (39) to rotate, thereby driving the striking mechanism to move. frequency: striking frequency (times / second)"" self.servo_motor_39["status"] = "running" self.servo_motor_39["rotation_angle"] = angle print(f"Servo motor (39) started, rotation angle: {angle}°, tapping frequency: {frequency} times / second") def stop_servo_motor_39(self) -> None: """Stop servo motor (39)"" self.servo_motor_39["status"] = "stop" print("Servo motor (39) has stopped running, and the striking mechanism has reset") # -------------------------- 2. Image Recognition Module (Convolutional Network + Superpixel Technology) -------------------------- class WeldSeamRecognition(nn.Module): """Weld seam recognition module: based on convolutional network + superpixel segmentation, corresponding to the core function of camera (12) in embodiments 1 and 2""" def __init__(self, threshold: float = 0.8): super().__init__() self.threshold = threshold # Threshold for weld width exceeding the limit (unit: mm) self.camera_12 = cv2.VideoCapture(0) # Simulate camera recognition (12) # Simplified CNN model (can actually be replaced with a lightweight ResNet / YOLO model) self.cnn = nn.Sequential( nn.Conv2d(3, 16, 3, padding=1), nn.ReLU(), nn.MaxPool2d(2), nn.Conv2d(16, 32, 3, padding=1), nn.ReLU(), nn.MaxPool2d(2), nn.Flatten(), nn.Linear(32 * 64 * 64, 2) # Output: probability of weld presence, predicted width) def superpixel_segmentation(self, img: np.ndarray) -> np.ndarray: "Superpixel segmentation: Locating the weld seam area, corresponding to the image magnification and region positioning functions in Example 1." segments = cv2.ximgproc.createSuperpixelSLIC(img, region_size=10, ruler=10.0) segments.iterate(10) mask = segments.getLabelContourMask() result = cv2.bitwise_and(img, img, mask=cv2.bitwise_not(mask)) return result def detect_weld_seam(self, img: np.ndarray) -> Tuple[bool, float]: """Inspect weld: Return (whether a weld was detected, weld width)"" # Superpixel segmentation preprocessing seg_img = self.superpixel_segmentation(img) # Image Normalization + CNN Inference img_tensor = torch.from_numpy(seg_img.transpose(2, 0, 1)).float() / 255.0 img_tensor = img_tensor.unsqueeze(0) with torch.no_grad(): output = self.cnn(img_tensor) seam_exist_prob = torch.sigmoid(output[0, 0]).item() seam_width = torch.abs(output[0, 1]).item() * 10 # Map to actual width (mm) return seam_exist_prob > 0.7, round(seam_width, 2) def detect_welding_gun_leave(self, img: np.ndarray) -> bool: """Detect whether the welding torch (11) has left the weld area, based on color + contour recognition"" hsv = cv2.cvtColor(img, cv2.COLOR_BGR2HSV) # Welding torch color threshold (Example: metallic silver-gray) lower_silver = np.array([0, 0, 100]) upper_silver = np.array([180, 50, 255]) mask = cv2.inRange(hsv, lower_silver, upper_silver) contours, _ = cv2.findContours(mask, cv2.RETR_EXTERNAL,cv2.CHAIN_APPROX_SIMPLE) return len(contours) < 2 # If there is no welding torch outline, it is considered to have left the image. def capture_image(self) -> Optional[np.ndarray]: """Camera (12) captures images""" ret, frame = self.camera_12.read() return frame if ret else None # -------------------------- 3. Main Control Module (Integrating the core logic of Examples 1 and 2) -------------------------- class WeldingControlSystem: def __init__(self): self.hardware = HardwareInterface() self.recognition = WeldSeamRecognition(threshold=0.8) # Weld width exceeding threshold: 0.8mm self.traverse_step = 5 # Traverse step size (mm) self.blow_workpiece_time = 10 # Blowing time for the workpiece (seconds) def traverse_search_seam(self) -> None: Example 1 Function: When no weld is detected, the telescopic rod is driven to traverse and search. print("\n=== Start weld seam traversal search===") while True: img = self.recognition.capture_image() if img is None: continue seam_exist, _ = self.recognition.detect_weld_seam(img) if seam_exist: print(f"Weld seam detected, iteration stops") break # X-direction traversal (priority) self.hardware.drive_telescopic_rod_4("right",self.traverse_step) time.sleep(0.5) # Move in the Y direction after reaching the end in the X direction. if self.hardware.telescopic_rod_4["position"] >= 500: self.hardware.drive_telescopic_rod_4("left", 500) self.hardware.drive_telescopic_rod_33("down",self.traverse_step) time.sleep(0.3) def weld_seam_detection_and_process(self) -> None: Example 1 Function: Weld width inspection and defect handling (slag removal by tapping) print("\n=== Start weld width detection===") img = self.recognition.capture_image() if img is None: print("Image acquisition failed") return seam_exist, seam_width = self.recognition.detect_weld_seam(img) if not seam_exist: print("No weld seam found, initiating a traversal search") self.traverse_search_seam() return print(f"Detected weld width: {seam_width}mm, threshold: {self.recognition.threshold}mm") if seam_width > self.recognition.threshold: print("Weld width exceeds standard, activate the slag removal mechanism") self.hardware.drive_servo_motor_39(angle=30, frequency=2) time.sleep(5) # Tap for 5 seconds self.hardware.stop_servo_motor_39() def fan_control_after_welding(self) -> None: Example 2 Function: Post-weld blower control (blows the workpiece first, then the welding torch) print("\n=== Start post-weld fan control===") while True: img = self.recognition.capture_image() if img is None: continue welding_gun_leave = self.recognition.detect_welding_gun_leave(img) if welding_gun_leave: print("Welding torch (11) has been detected leaving the weld area") # Step 1: Blow air into the outer frame of the door and window (22) and the branch pipe (19) self.hardware.start_fan(mode="workpiece", wind_speed=6) time.sleep(self.blow_workpiece_time) # Step 2: Switch to single air outlet (10) and blow welding torch (11) self.hardware.start_fan(mode="welding_gun", wind_speed=5) time.sleep(8) # The welding torch lasts for 8 seconds self.hardware.stop_fan() break time.sleep(0.5) def run(self) -> None: "System Operation Main Flow" print("=== Aluminum Alloy Door and Window Welding Control System Started===") # 1. Weld Inspection and Defect Handling (Example 1) self.weld_seam_detection_and_process() # 2. Post-weld fan cooling control (Example 2) self.fan_control_after_welding() print("=== Welding process ended===") if __name__ == "__main__": system = WeldingControlSystem() system.run() The embodiments of the present invention are given for illustrative and descriptive purposes only, and are not intended to be exhaustive or to limit the invention to the forms disclosed. Many modifications and variations will be apparent to those skilled in the art. The embodiments were chosen and described in order to better illustrate the principles and practical application of the invention, and to enable those skilled in the art to understand the invention and to design various embodiments with various modifications suitable for a particular purpose.

Claims

1. A welding device for aluminum alloy door and window processing, comprising: The utility model provides a welding platform (1), its characterized in be provided with slide rail (3) on one side of the top of welding platform (1), one end of slide rail (3) is provided with first electric control telescopic link (2), the top of slide rail (3) is provided with mechanical arm (13), one end of mechanical arm (13) is provided with welding torch (11), the other side of the top of welding platform (1) is provided with door and window outer frame (22), the inside of door and window outer frame (22) is provided with a plurality of door and window branch pipes (19) laterally, The intersection of door and window branch pipe (19) and door and window outer frame (22) is provided with lock plate (20), the outside of lock plate (20) is provided with rack (7), the front and rear ends of rack (7) are provided with turnover frame (41), the connecting place of turnover frame (41) and rack (7) is provided with adaptive bearing (24), the one end between lock plate (20) and door and window branch pipe (19) is provided with lock (42) and is clamped, The top of welding platform (1) is provided with the identification operation mechanism in the middle, and the identification operation mechanism comprises Y guide groove (15), and the Y guide groove (15) is arranged in the middle of the top of the welding platform (1); One side of the welding torch (11) is provided with a heat dissipation mechanism, and the heat dissipation mechanism comprises a single-outlet pipe (10), and the single-outlet pipe (10) is horizontally aligned on one side of the welding torch (11); The bottom of the door and window branch pipe (19) is provided with an auxiliary mechanism on one side, and the auxiliary mechanism comprises a lifting arc block (35), and the lifting arc block (35) is arranged on one side of the bottom of the door and window branch pipe (19).

2. The welding device for aluminum alloy door and window processing according to claim 1, characterized in that, The identification operation mechanism further comprises a pivot (40), a servo motor (39), a bottom column (38), a push rod (37), a transverse groove (34), a third electric control telescopic rod (33), a half gear (32), an iron plate (31), a limiter (29), a retaining plate (28), a vertical curved arm shaft (27), an adaptive back plate (26), a permanent magnet (23), a guide shaft (21), a limiting piece (18), a spring (17), a multi-hole column plate (16), a sliding support (14), an identification camera (12), a side panel (8), an adaptive mounting frame (6), an X-direction guide rail (5), a second electric control telescopic rod (4), the inner side of the Y-direction guide groove (15) is provided with the third electric control telescopic rod (33), one end of the third electric control telescopic rod (33) is provided with a sleeve block, the top of the sleeve block is provided with the X-direction guide rail (5), the inner side of the X-direction guide rail (5) is provided with the transverse groove (34), one side of the transverse groove (34) is provided with the second electric control telescopic rod (4), one side of the second electric control telescopic rod (4) is provided with the adaptive mounting frame (6), the top of the adaptive mounting frame (6) is provided with the sliding support (14), the outer upper end of the sliding support (14) close to one end of the door and window support pipe (19) is provided with the identification camera (12), the upper end of the folding frame (41) on one side of the placing rack (7) is provided with the pivot (40), one side of the pivot (40) is provided with the servo motor (39), the bottom of the servo motor (39) is provided with the bottom column (38), the upper end of the folding frame (41) on the other side of the placing rack (7) is provided with the half gear (32), one side of the half gear (32) is provided with a full gear (30), the outside of the full gear (30) is provided with the adaptive back plate (26), one side of the adaptive back plate (26) is provided with the retaining plate (28), the back of the retaining plate (28) is provided with the limiter (29), the inner side of the limiter (29) is vertically penetrated and provided with the push rod (37), one end of the push rod (37) is provided with the vertical curved arm shaft (27), one end of the vertical curved arm shaft (27) is provided with the full gear (30), the other end of the push rod (37) is provided with the iron plate (31), the side panel (8) is arranged on one side of the top of the welding table (1), one side of the top of the side panel (8) is provided with the multi-hole column plate (16), the inner side of the hole on the surface of the multi-hole column plate (16) is vertically penetrated and provided with the guide shaft (21), the outer upper end of the guide shaft (21) is sleeved with the spring (17), the top of the guide shaft (21) is provided with the limiting piece (18), and the bottom of the guide shaft (21) is provided with the permanent magnet (23).

3. The welding device for aluminum alloy door and window processing according to claim 2, characterized in that, The inside of the identification camera (12) is provided with a control module, the control module is used for identifying whether the welding gap width between the door and window support pipe (19) and the door and window outer frame (22) is too large through a pattern recognition technology, the pattern recognition technology adopts a neural convolution network in deep learning combined with a super pixel technology of image recognition, so as to realize magnification of a welding gap image and acceleration of an identification process.

4. The welding device for aluminum alloy door and window processing according to claim 3, characterized in that, The neural convolutional network is used for feature extraction of the collected welding gap image, and the superpixel technology is used for segmenting the welding gap image into a plurality of superpixel units, so as to realize accurate positioning and image enlargement of the welding gap region.

5. The welding device for aluminum alloy door and window processing according to claim 2, characterized in that, The control module further comprises an image acquisition control unit for controlling the identification camera (12) to shoot the image of the welding gap between the door and window branch pipe (19) and the door and window outer frame (22), and transmitting the collected image to the pattern recognition processing unit.

6. The welding device for aluminum alloy door and window processing according to claim 2, characterized in that, The control module can output a judgment signal of whether the welding gap width is too large according to the identification result, and the judgment signal can be transmitted to the inside of the servo motor (39).

7. The welding device for aluminum alloy door and window processing according to claim 2, characterized in that, When the identification camera (12) does not identify the door and window branch pipe (19) and the door and window outer frame (22), the control module drives the third electric telescopic rod (33) and the second electric telescopic rod (4) to act, and then drives the sliding support (14) and the adaptive placement frame (6) to slide in the transverse groove (34) of the X-direction guide rail (5) and the Y-direction guide groove (15) inside, so as to realize the gradual movement of the identification camera (12) along the welding surface of the door and window branch pipe (19) and the door and window outer frame (22).

8. The welding device for aluminum alloy door and window processing according to claim 1, characterized in that, The heat dissipation mechanism further comprises a main control fan (9) and a supporting rod (25), one side of the single air outlet pipe (10) is provided with the main control fan (9), the back of the main control fan (9) is provided with the supporting rod (25), when the identification camera (12) observes that the welding torch leaves the welding seam between the door and window branch pipe (19) and the door and window outer frame (22), the control module inside the identification camera (12) receives the observation signal and sends a control signal to the main control fan (9), so that the main control fan (9) first blows the surface of the door and window branch pipe (19) and the door and window outer frame (22) along a straight line, and when the blowing lasts for a set time, the main control fan (9) completely discharges the internal gas through the single air outlet pipe (10), at this time, the main control fan (9) stops blowing air to the door and window branch pipe (19) and the door and window outer frame (22), and instead blows air to the surface of the welding torch (11) through the single air outlet pipe (10).

9. The welding device for aluminum alloy door and window processing according to claim 1, characterized in that, The auxiliary mechanism further comprises a support plate (36), a rotating shaft (43) and a vertical shaft piece (44), the bottom of the lifting arc block (35) is provided with the support plate (36), the bottom of the support plate (36) is provided with the rotating shaft (43), the two sides of the rotating shaft (43) are provided with the vertical shaft piece (44), and the vertical shaft piece (44) is fixedly connected with the iron plate (31).

10. A welding method of a welding device for aluminum alloy door and window processing, characterized by, The method comprises the following steps: S1: The door and window outer frame (22) and the door and window branch pipe (19) are lifted to the designated station of the placing rack (7), the precise positioning is completed through the limiting structure on the placing rack (7), it is ensured that the two to-be-welded regions are aligned and fitted, the locking assembly on both sides of the placing rack (7) is started, the door and window outer frame (22) and the door and window branch pipe (19) are fixed, displacement during welding is prevented, and at the same time, the control module initializes the identification camera (12), the second electric telescopic rod (4), the third electric telescopic rod (33), the fan (9) and other components, and completes self-checking of the standby state of the equipment; S2: The recognition camera (12) is started, the preliminary scanning of the fitting area of the door and window frame (22) and the branch pipe (19) is performed through the deep learning neural convolution network combined with the super pixel technology, if the welding area to be welded is not recognized, the control module drives the second electric control telescopic rod (4) and the third electric control telescopic rod (33) to move, drives the sliding support to slide along the X-direction guide rail (5) transverse groove (34) and the Y-direction guide groove (15), makes the camera (12) traverse the workpiece surface until the welding seam is positioned, after the positioning is completed, the camera (12) transmits the welding seam position signal to the control system of the welding gun (11), guides the welding gun (11) to move to the starting end of the welding seam and adjusts to the welding angle, and the welding gun (11) is adjusted to the welding angle. S3: The welding gun (11) starts welding according to the preset parameters, moves uniformly along the welding seam track to perform the welding operation, in the welding process, the recognition camera (12) continuously collects the welding seam image, the mode recognition technology is used to monitor the welding gap width change in real time, the data is transmitted to the control module synchronously, if it is detected that the gap width exceeds the preset threshold value, the control module immediately sends the early warning signal, suspends the welding operation and records the abnormal position, waits for subsequent processing, if the welding parameters are normal, the welding work of the whole welding seam is continuously completed. S4: After the welding gun (11) completes the welding, the recognition camera (12) captures the welding gun (11) leaving signal and transmits it to the control module, the control module immediately sends the starting instruction to the fan (9), the fan (9) blows the welding surface of the door and window frame (22) and the branch pipe (19) along the straight line under the support of the supporting rod (25), in the blowing process, the airflow quickly takes away the welding residual heat, realizes the workpiece cooling and heat dissipation, at the same time, a small amount of welding slag and dust on the surface is removed, a clean surface is provided for subsequent detection, the blowing time is executed according to the preset threshold value of the control module. S5: After the blowing is completed, the recognition camera (12) scans the welding seam again, extracts the welding seam image features through the neural convolution network, enlarges the image details combined with the super pixel technology, accurately detects whether the gap width is qualified, if the detection is qualified, the next step is entered, if the gap is too large and other defects are detected, the control module drives the servo motor (39) to start, drives the push rod (37) and the iron plate (31) to move through the half-tooth gear (32) and the full-tooth gear (30) meshing transmission, and the permanent magnet (23) is used to pull the guide shaft (21) to knock the defect area regularly, remove the impurities and welding slag, and prepare for repair welding. S6: If there is welding defect, the welding gun (11) moves to the corresponding area according to the defect position positioned by the recognition camera (12), performs repair welding operation, after the repair welding is completed, the control module sends the signal again to switch the wind direction of the fan (9), makes the airflow of the fan (9) be discharged to the surface of the welding gun (11) through the single air outlet pipe (10), and the welding gun (11) is cooled. After the workpiece and the welding gun (11) are cooled to the safe temperature, the locking assembly of the placing rack (7) is unlocked, the workpiece is lifted and transported to the specified storage area, and the whole welding process is completed.

Images