Automatic welding device of turbine blade visual identification robot

Abstract

The invention discloses a turbine blade visual identification robot automatic welding device, and relates to the technical field of robot welding, the turbine blade visual identification robot automatic welding device comprises a bottom plate, one side of the upper end of the bottom plate is connected with a control table, one side of the control table is located at the upper end of the bottom plate and connected with a mounting seat, and the middle of the mounting seat is hollow; a plurality of mounting holes are evenly formed in the outer side wall of the mounting base, a first motor is connected to the upper end of a bottom plate on the inner side of the mounting base, and a first movable arm is rotationally connected to the center of the upper end of the mounting base. The electric push rod drives the connecting frame to stretch out and draw back, the camera is driven to flexibly ascend and descend in the sliding groove, multi-angle weld joint dynamic tracking is achieved, the transparent protective cover is installed on the outer side of the camera in a threaded sealing mode, and compared with an existing camera fixed to the area close to a welding gun, welding spatter and smoke pollution can be effectively isolated, the image collection definition is ensured, and the welding quality is improved. And by combining a visual algorithm of a console, the welding positioning precision of the complex curved-surface blade is remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of robotic welding technology, specifically to an automated welding device for turbine blades using a vision recognition robot. Background Technology

[0002] As a core component of energy equipment, the manufacturing quality of turbine blades directly affects the overall operating efficiency and reliability of the turbine. Currently, the welding of these parts still largely relies on manual labor or semi-automatic equipment, which is not only labor-intensive and inefficient, but also results in significant fluctuations in weld quality due to individual differences among operators, making it difficult to meet high consistency requirements. In particular, when dealing with complex curved surfaces and multi-dimensional spatial welds, traditional processes lack adaptability and flexibility, often resulting in defects such as weld misalignment and incomplete welds, which seriously restricts the manufacturing level of high-performance blades.

[0003] Specifically, turbine blade profiles are highly complex and require specific aerodynamic characteristics. Weld trajectories are often located in difficult-to-observe and inaccessible positions such as deep bevels, narrow gaps, or irregular corners. Conventional mechanical positioning and offline programming methods cannot respond to dynamic changes during the welding process in real time, easily leading to trajectory deviations. On the other hand, the intense arc light, heat radiation, and spatter generated in the high-temperature welding environment pose a severe challenge to visual recognition systems. Although mainstream welding robots are equipped with visual sensing devices, most lack effective protection and heat dissipation designs for extreme working conditions. During continuous operation, cameras experience thermal drift due to temperature rise, resulting in distorted calibration parameters, blurred images, or loss of features, severely weakening the stability of weld recognition and tracking.

[0004] Furthermore, existing automated systems still lack in visual dynamic compensation and multimodal sensor fusion. Single-vision guidance is susceptible to environmental interference, and systems lacking infrared, laser, or other auxiliary sensing methods struggle to achieve robust all-weather operation. Simultaneously, most equipment lacks integrated real-time thermal management mechanisms, failing to effectively cool and protect optical components from contamination, further limiting the equipment's availability and accuracy maintenance capabilities during long-term continuous production. To address these bottlenecks, there is an urgent need to develop an automated welding system with highly adaptable visual recognition, dynamic anti-interference, and intelligent thermal control functions. This device is proposed against this backdrop, aiming to significantly improve the accuracy, efficiency, and system reliability of blade welding through innovative visual enhancement design and thermally robust control strategies. Summary of the Invention

[0005] The purpose of this invention is to provide an automatic welding device for turbine blades using a visual recognition robot. Through dynamic visual compensation and a high-efficiency dual heat dissipation system, it can effectively resist high welding temperatures and spatter interference, and achieve high-precision and high-stability automatic tracking and welding of complex curved blade welds.

[0006] To achieve the above objectives, the present invention provides the following technical solution: an automatic welding device for a turbine blade visual recognition robot, comprising a base plate, a control console fixedly connected to one side of the upper end of the base plate, a mounting base fixedly connected to one side of the control console at the upper end of the base plate, the mounting base being hollow in the middle, a plurality of mounting holes evenly provided on the outer side wall of the mounting base, a first motor fixedly connected to the upper end of the base plate on the inner side of the mounting base, a first movable arm rotatably connected to the upper center position of the mounting base, the output end of the first motor being drively connected to the lower center position of the first movable arm, a first rotating shaft rotatably connected to the upper end of the first movable arm, a second movable arm fixedly sleeved on the first rotating shaft, one end of the first rotating shaft extending through to the outside of the first movable arm, and a second motor connected to one end of the first rotating shaft, the second motor being fixedly connected to the side wall of the first movable arm, a welding component connected to the end of the second movable arm away from the first rotating shaft, an identification component provided on the side wall of the second movable arm, and a heat dissipation component shared by the second movable arm and the first movable arm. This invention achieves flexible movement and stable support of the welding device as a whole through the coordination of the control console, mounting base, first motor, first movable arm, second movable arm and other components, providing a basic structural guarantee for the automated welding of complex curved blades and improving the integration and motion accuracy of the equipment.

[0007] Preferably, the welding assembly includes a third motor and a mounting plate. A second rotating shaft is rotatably connected to the upper end of the second movable arm. One end of the second rotating shaft is fixedly connected to the center of the mounting plate. A welding head is fixedly mounted on the edge of the mounting plate away from the second rotating shaft. The other end of the second rotating shaft is connected to the third motor, and the third motor is fixedly connected to the side wall of the second movable arm. This invention drives the second rotating shaft and the mounting plate with the third motor, enabling the welding head to achieve multi-angle adjustment. This significantly improves the adaptability of the welding head to the position of the surface to be welded, ensuring the accuracy and consistency of weld tracking and the welding process.

[0008] Preferably, the recognition component includes an electric actuator and a connecting frame. An inner cavity is formed inside the second movable arm, and the electric actuator is fixedly connected to the center of one bottom end of the inner cavity. The output end of the electric actuator is connected to the connecting frame. A sliding groove is formed at the upper end of the second movable arm on one side of the welding head. The connecting frame extends outward from the second movable arm through the sliding groove, and a mounting plate is fixedly connected to the end of the connecting frame. In this invention, the electric actuator pushes the connecting frame to smoothly rise and fall within the sliding groove, achieving flexible adjustment of the camera height. This adapts to the weld seam recognition needs at different heights and angles, enhancing the dynamic response capability and environmental adaptability of the vision system.

[0009] Preferably, a camera is fixedly connected to the mounting plate on the side away from the connecting frame, and a transparent protective cover is threaded onto the outside of the camera. Several heat dissipation holes are provided on the side of the mounting plate near the slide groove. The transparent protective cover with its threaded seal effectively isolates welding spatter and fumes, ensuring clear image acquisition. The heat dissipation holes on the mounting plate initially improve the camera's heat dissipation capacity and extend its service life.

[0010] Preferably, the heat dissipation assembly includes a mounting pipe and a fan. The heat dissipation holes are evenly distributed on the mounting plate, with the hole diameter gradually increasing from the side closer to the mounting pipe to the side farther away. Furthermore, the axis of each heat dissipation hole is inclined towards the camera housing. This gradient hole diameter and inclined orientation design effectively guides airflow to concentrate on the camera housing, enhancing heat dissipation efficiency and preventing image drift or equipment malfunction caused by heat accumulation.

[0011] Preferably, a spiral heat dissipation pipe is sleeved on the outer side of the electric actuator. The heat dissipation assembly includes a mounting pipe and a fan. The mounting pipe is fixedly installed on the end of the second movable arm near the first rotating shaft. The fan is fixedly connected inside the mounting pipe. A filter plate is detachably connected to the end of the mounting pipe away from the second movable arm. This invention forms an active air supply structure with a fan and mounting pipe, and places the filter plate at the air inlet. This effectively blocks dust while achieving cooling, ensuring heat dissipation cleanliness, and improving the reliability of the system under harsh operating conditions.

[0012] Preferably, the mounting pipe is connected to the heat dissipation holes on the mounting plate via an air duct. This connection between the mounting pipe and the heat dissipation holes forms a directional air cooling path, achieving efficient active heat dissipation for the camera module and further improving image acquisition stability and equipment durability.

[0013] Preferably, a spiral heat dissipation tube is sleeved on the outer side of the electric actuator, and fixing blocks are inserted and connected to both ends of the spiral heat dissipation tube. The two fixing blocks are fixedly connected to the inner wall of the inner cavity. The spiral heat dissipation tube is tightly wound around the outer side of the electric actuator, which can efficiently absorb the heat generated during its operation, and is stably connected to the heat dissipation circuit through the fixing blocks, thereby improving the heat dissipation efficiency and operational reliability of the electric actuator.

[0014] Preferably, a slot is provided on one side of the first movable arm, and a water tank is detachably installed in the slot. Corrugated pipes are connected to both ends of the spiral heat dissipation pipe, and both corrugated pipes are connected to the interior of the water tank. A water pump is built into the water tank, and the outlet of the water pump is connected to one of the corrugated pipes. This invention forms a circulating water cooling system using the water tank, corrugated pipes, and spiral heat dissipation pipes, enabling efficient thermal management of the electric actuator and the internal cavity area. The detachable design of the water tank facilitates maintenance, and the flexible connection of the corrugated pipes adapts to mechanical movement, ensuring continuous and stable operation of the cooling circuit.

[0015] Preferably, both bellows are made of metal corrugated hoses. By using metal corrugated hoses for the bellows, excellent heat resistance, fatigue resistance, and sealing performance are achieved, ensuring smooth coolant circulation under high temperature and frequent operation conditions, further improving the reliability and service life of the overall heat dissipation system.

[0016] Compared with the prior art, the beneficial effects of the present invention are:

[0017] 1. This turbine blade vision recognition robot automatic welding device uses an electric actuator to drive the connecting frame to perform precise telescopic movements, which in turn drives the camera to rise and fall smoothly within the wear-resistant groove. This enables flexible tracking and dynamic capture of multi-angle welds on complex curved surfaces. A high-strength transparent protective cover is installed on the outside of the camera using a threaded sealing method, which effectively isolates the spatter particles and dust pollution generated during the welding process, significantly improving the clarity and stability of image acquisition. Combined with the advanced vision recognition algorithm integrated in the control console, the system can compensate for visual deviations caused by workpiece shape and position or environmental interference in real time, thereby greatly improving the positioning accuracy of the blade weld and the consistency of welding quality.

[0018] 2. In terms of thermal management, the device introduces a dual heat dissipation mechanism combining active air cooling and water cooling circulation. The fan forces external cold air through multi-layer filter plates to remove dust and then into the mounting pipe. The airflow blows directly onto the camera module through specially designed heat dissipation holes on the mounting plate, achieving forced convection cooling and effectively controlling the surface temperature rise. At the same time, the system's built-in water tank and water pump drive the coolant to circulate through the spiral heat dissipation pipe, absorbing the heat generated by the electric actuator and the equipment's internal cavity. The flexible corrugated pipe connection structure adapts to the needs of the robotic arm's high-frequency and complex movements, ensuring that the cooling circuit is not affected by movement deformation.

[0019] 3. Through the efficient synergy of the aforementioned air-cooling and water-cooling systems, the camera always operates within a reasonable temperature range, avoiding drift, distortion, or even hardware damage caused by overheating. This system significantly improves the environmental adaptability and durability of the visual recognition components under harsh conditions such as high temperature and dust, ensuring that the robot can operate continuously and stably for a long time without interrupting cooling, thereby extending the overall service life of the equipment and supporting the unmanned and high-speed production needs of modern workshops. Attached Figure Description

[0020] Figure 1 This is a schematic diagram of the overall structure of the present invention;

[0021] Figure 2 This is a schematic diagram of the internal structure of the present invention;

[0022] Figure 3 This is a cross-sectional view of the connection structure of the second movable arm of the present invention;

[0023] Figure 4 This is a cross-sectional view of the connection structure of the transparent protective cover of the present invention.

[0024] In the diagram: 1. Base plate; 2. Mounting base; 3. First movable arm; 4. Motor 1; 5. Mounting hole; 6. First rotating shaft; 7. Second movable arm; 8. Motor 2; 9. Second rotating shaft; 10. Motor 3; 11. Mounting plate; 12. Welding head; 13. Inner cavity; 14. Electric push rod; 15. Connecting frame; 16. Slide rail; 17. Mounting plate; 18. Camera; 19. Transparent protective cover; 20. Heat dissipation hole; 21. Mounting pipe; 22. Fan; 23. Filter plate; 24. Spiral heat dissipation pipe; 25. Fixing block; 26. Corrugated pipe; 27. Water tank; 28. Control console. Detailed Implementation

[0025] 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.

[0026] like Figures 1-4As shown, the automatic welding device for turbine blades using a visual recognition robot of the present invention includes a base plate 1. A control console 28 is connected to one side of the upper end of the base plate 1. The control console 28 integrates a vision processing module, a motion control module, and a human-machine interface, and has functions of weld recognition, path planning, motion control, and fault diagnosis. It is the intelligent core of the entire system. A mounting base 2 is connected to one side of the control console 28 at the upper end of the base plate 1. The mounting base 2 has a hollow center, and several mounting holes 5 are evenly distributed on the outer side wall of the mounting base 2. A motor 4 is connected to the upper end of the base plate 1 inside the mounting base 2. The upper center of the mounting base 2 is... A first movable arm 3 is rotatably connected to a first movable arm 3. The output end of motor 4 is connected to the center of the lower end of the first movable arm 3. A first rotating shaft 6 is rotatably connected to the upper end of the first movable arm 3. A second movable arm 7 is fixedly sleeved on the first rotating shaft 6. One end of the first rotating shaft 6 extends through the outside of the first movable arm 3, and a second motor 8 is connected to the other end of the first rotating shaft 6. Motor 8 is connected to the side wall of the first movable arm 3. A welding assembly is connected to the end of the second movable arm 7 away from the first rotating shaft 6. An identification component is provided on the side wall of the second movable arm 7. A heat dissipation component is provided on both the second movable arm 7 and the first movable arm 3. The welding assembly includes... The second movable arm 7 includes a motor 10 and a mounting plate 11. A second rotating shaft 9 is rotatably connected to the upper end of the second movable arm 7. One end of the second rotating shaft 9 is connected to the center of the mounting plate 11. A welding head 12 is installed on the edge of the mounting plate 11 away from the second rotating shaft 9. The other end of the second rotating shaft 9 is connected to the motor 10, and the motor 10 is connected to the side wall of the second movable arm 7. The identification component includes an electric actuator 14 and a connecting frame 15. An inner cavity 13 is formed inside the second movable arm 7. The electric actuator 14 is connected to the center of one bottom end of the inner cavity 13. The electric actuator 14 is an electric linear actuator used for precise control of the camera 18. The extension and retraction action has high repeatability and load capacity, adapts to frequent lifting and lowering needs, and requires CNC control. The output end of the electric actuator 14 is connected to the connecting frame 15. The upper end of the second movable arm 7 on the side of the welding head 12 is provided with a slide groove 16. The connecting frame 15 extends to the outside of the second movable arm 7 through the slide groove 16, and the connecting frame 15 is connected to the mounting plate 17. The side of the mounting plate 17 away from the connecting frame 15 is connected to the camera 18, which is an existing industrial camera. The camera 18 is threaded with a transparent protective cover 19. The side of the mounting plate 17 near the slide groove 16 is provided with several heat dissipation holes 20.

[0027] Specifically, when using this turbine blade vision recognition robot automatic welding device, the device is fixed to the working area through the mounting holes 5 on the base plate 1. The power and air supply are connected, and the correct connection of the control console 28 with each motor and sensor is ensured. The turbine blade to be welded is fixed on a special fixture and placed within the welding working range. The vision system is started through the control console 28, and the camera 18 initially acquires the blade image for coarse weld positioning. The camera 18 slides along the slide groove 16 under the push of the electric push rod 14 to scan the blade weld area at close range. The control console 28 processes the image data, generates a welding path, and dynamically compensates for the positional deviation caused by the change in the blade surface curvature. The control console 28 drives motor 4, motor 8, and motor 10 to adjust the welding head 12 to the predetermined posture and position. The welding power is started, and the welding head 12 automatically welds according to the planned path, ensuring the clarity of image acquisition. Combined with the vision algorithm of the control console 28, the welding positioning accuracy of complex curved blades is significantly improved.

[0028] Furthermore, the heat dissipation assembly includes a mounting pipe 21 and a fan 22. The mounting pipe 21 is located on the end of the second movable arm 7 near the first rotating shaft 6. The fan 22 is connected inside the mounting pipe 21 and provides continuous airflow. The filter plate 23 is a detachable filter structure used to filter dust and particulate matter in the air, ensuring the quality of the incoming air. The filter plate 23 is connected to the end of the mounting pipe 21 away from the second movable arm 7. A spiral heat dissipation pipe 24 is sleeved on the outside of the electric push rod 14. The spiral heat dissipation pipe 24 is coiled around the outside of the electric push rod 14 to increase the heat dissipation area and improve heat exchange efficiency. The corrugated pipe 26 has the characteristics of flexibility and high temperature resistance, which can keep the coolant flow uninterrupted when the robotic arm moves. The two ends of the spiral heat dissipation pipe 24 are respectively connected to the fixing blocks 25, and the two fixing blocks 25 are respectively connected to the inner wall of the inner cavity 13. A slot is opened on one side of the first movable arm 3, and a water tank 27 is installed in the slot. Both ends of the spiral heat dissipation pipe 24 are connected to the corrugated pipe 26. Both corrugated pipes 26 are metal corrugated hoses. Both corrugated pipes 26 are connected to the inside of the water tank 27. The water tank 27 has a built-in water pump, and the water outlet of the water pump is connected to one of the corrugated pipes 26.

[0029] Furthermore, during the welding process, the fan 22 continuously supplies air to cool the camera 18, and the water pump in the water tank 27 drives the coolant to circulate, dissipating heat from the electric push rod 14 and the inner cavity 13 through the spiral heat dissipation pipe 24, ensuring the stable operation of the vision system. After welding, the welding head 12 retracts to a safe position, the camera 18 retracts, and all power and air sources are turned off, completing the welding operation. The dual heat dissipation system controls the operating temperature of the camera 18 at a reasonable operating temperature, extending the equipment life and ensuring continuous trouble-free operation.

[0030] The method of using the device of the present invention is as follows: When using the automatic welding device of the turbine blade vision recognition robot, the device is fixed to the working area through the mounting hole 5 on the base plate 1, the power supply and air supply are connected, and the connection between the control console 28 and each motor and sensor line is correct. The turbine blade to be welded is fixed on the special fixture and placed within the welding working range. The vision system is started through the control console 28, and the camera 18 initially acquires the blade image for rough positioning of the weld. The camera 18 slides along the slide groove 16 under the push of the electric push rod 14 to scan the blade weld area at close range. The control console 28 processes the image data, generates the welding path, and dynamically compensates for the position deviation caused by the change of the blade surface. The control console 28 drives motor 4, motor 8 and motor 10 to adjust the welding head 12 to the predetermined posture and position. The welding power supply is turned on, and the welding head 12 automatically welds according to the planned path. During the welding process, the fan 22 continuously blows air to cool the camera 18, and the water pump in the water tank 27 drives the coolant to circulate. The coolant is cooled through the spiral heat dissipation pipe 24 to the electric push rod 14 and the inner cavity 13, ensuring the stable operation of the vision system. After the welding is completed, the welding head 12 returns to the safe position, the camera 18 retracts, and all power and air sources are turned off to complete the welding operation.

[0031] In this invention, the control console 28 is fixedly installed on the upper side of the base plate and serves as the intelligent control hub of the entire welding device. The control console integrates a high-performance industrial computing unit and a variety of dedicated functional modules, specifically including a vision processing module, a motion control module, and a human-machine interface.

[0032] The vision processing module is equipped with a high-performance image processing chip and a dedicated weld seam recognition algorithm. It can receive and process image data from the camera in real time, effectively filtering out environmental interference such as welding arc light and spatter, accurately identifying the weld seam features and position information of the complex curved surface of the blade, and generating high-precision three-dimensional coordinate data. The motion control module, based on the coordinate information provided by the vision processing module, performs adaptive path planning and multi-axis motion coordination calculations, and generates control commands. This module connects to actuators such as motor one, motor two, motor three, and electric actuators through a bus system, achieving precise closed-loop control of the lifting and lowering movements of the first and second movable arms, the welding head, and the camera, thereby ensuring the welding torch operates stably along a predetermined trajectory. The human-machine interface uses a high-brightness industrial touchscreen, providing a visual operation window. Users can monitor the welding process, equipment status, and image recognition results in real time, and support parameter settings, program selection, and manual intervention. The interface also integrates a fault diagnosis system, which can judge, alarm, and log abnormal states such as visual deviation, execution errors, and communication interruptions in real time, greatly improving the maintainability and ease of operation of the system.

[0033] Through the collaborative work of the aforementioned modules, the control console integrates visual perception, decision-making and planning, and motion execution, forming an intelligent control system capable of adapting to complex working conditions and possessing high-precision tracking and anti-interference capabilities. This is the core guarantee for the high-precision and high-stability operation of the entire automated welding device. It should also be noted that in this invention, the weld seam recognition algorithm is a core software component of the control console's visual processing module. Its design aims to cope with the harsh conditions such as the complex curved surface of turbine blades, high reflectivity, and strong interference at the welding site, achieving real-time, accurate, and robust extraction and positioning of weld seam features. This algorithm adopts an existing mature multi-stage processing flow, with the specific steps as follows:

[0034] First, in the image preprocessing stage, the algorithm performs noise reduction and enhancement on the raw images captured by the camera. Adaptive median filtering combined with Gaussian filtering is used to effectively suppress interference from welding spatter particles, random noise, and arc flicker. Simultaneously, an image enhancement method based on the Retinex model is applied to overcome uneven illumination and local overexposure / underexposure problems, improving the overall image contrast and feature separability.

[0035] Secondly, in the weld feature extraction stage, the algorithm employs a strategy combining edge detection and feature semantic segmentation. On one hand, an improved Canny edge detection operator is used, whose high and low thresholds are dynamically and adaptively determined based on the image gradient histogram to accurately capture subtle gradient changes in the blade bevel edge and weld line. On the other hand, a lightweight deep learning segmentation model is introduced. This model has been trained on a large number of blade weld images and can effectively distinguish weld areas, base material, spatter, and background, and perform pixel-level precise segmentation of the weld centerline, greatly enhancing the algorithm's generalization ability under complex textures and variable working conditions.

[0036] Subsequently, in the feature point matching and 3D localization stage, the algorithm reconstructs the 3D features from the 2D image. Using pre-calibrated camera intrinsic and extrinsic parameters, the weld feature points extracted from the 2D image are transformed into 3D spatial coordinates in the robot's base coordinate system. During this process, the algorithm employs stereo vision or structured light projection principles, combined with the blade CAD model, to perform feature point matching and verification, effectively solving problems such as feature occlusion and missing features, ultimately outputting a series of continuous and accurate 3D point coordinates on the weld trajectory.

[0037] Finally, in the trajectory optimization and fault tolerance stage, the algorithm incorporates a process knowledge base and a real-time feedback mechanism. Mathematical methods such as B-spline curves are used to smoothly fit discrete feature points, generating a smooth welding path that meets process requirements. Simultaneously, this module monitors the recognition confidence level in real time. If abnormal offsets or recognition failures are detected, a re-recognition process is immediately triggered or a pause command is sent to the motion control module, and fault information is recorded to the human-machine interface, thereby ensuring the continuity and reliability of the system operation.

[0038] In summary, this weld seam recognition algorithm significantly improves the recognition accuracy, anti-interference ability, and system stability of complex curved surface weld seams in harsh welding environments by integrating multimodal image processing, dynamic adaptive segmentation, high-precision 3D reconstruction, and intelligent decision-making. It provides a crucial perceptual foundation for subsequent high-precision tracking and welding.

[0039] Example 1:

[0040] An automatic welding device for the last-stage blades of a large power plant steam turbine is disclosed. The device's control console incorporates a vision processing module equipped with the weld seam recognition algorithm described in this specification. Operators select "planar weld seam mode" and start the system via a human-machine interface. The vision processing module controls the camera to descend to a predetermined height and acquires images of the blade's welding area. After adaptive filtering and image enhancement, an improved Canny operator is used to accurately extract the weld seam edge, and a lightweight U-Net model is used for semantic segmentation to calculate the three-dimensional coordinate point set of the weld seam trajectory. The motion control module plans the welding path and controls the coordinated movement of the first and second motors, enabling the first and second movable arms to precisely position the welding head to the weld seam starting point. During welding, a fan continuously operates, delivering cooling air filtered by a dust-removing plate through a mounting pipe to the heat dissipation holes of the mounting plate for active air cooling of the camera. Simultaneously, a water pump in the water tank drives cooling water through a corrugated pipe into a spiral heat dissipation pipe, circulating water to cool the electric actuator, ensuring the camera remains at a normal operating temperature below 45°C. The system successfully completed the welding of a 4.2-meter-long weld seam without any tracking interruption caused by arc light interference throughout the process. The weld seam was uniformly formed, and the equipment ran continuously for 6 hours without overheating alarms.

[0041] Example 2:

[0042] A welding device for repairing tortuous blades of heavy-duty gas turbines is disclosed. The blade profile is complex, and the area to be repaired is located in a narrow space at the blade root, resulting in high welding fume density. The operator selects the "complex surface repair mode" via the control panel. After system startup, the electric actuator adjusts the camera height three times according to a preset program. The vision processing module employs multi-angle image fusion technology to effectively overcome local occlusion problems and accurately identify the contour of the worn weld seam to be repaired. During welding, due to space constraints, the welding head needs to frequently change angles. A third motor drives the mounting plate to perform micro-rotation to adapt to curvature changes. At this time, welding spatter increases significantly, but the threaded, transparent protective cover of the camera effectively isolates contamination, ensuring clear and stable image acquisition. Simultaneously, the gradient-diameter heat dissipation holes on the mounting plate effectively direct cooling airflow to key heat-generating parts of the camera housing, improving heat dissipation efficiency. A spiral heat dissipation pipe is tightly wound around the outer wall of the electric actuator, and its temperature is controlled below 60℃ through circulating water cooling, avoiding positioning errors caused by thermal expansion. Ultimately, the system successfully completed the precise repair of the worn area of ​​the twisted blade, with the maximum deviation of weld tracking not exceeding 0.1 mm, demonstrating the system's high adaptability and reliability under harsh working conditions.

[0043] This invention presents an automated welding device for turbine blades using a vision-based robotic system. Addressing the core issues of insufficient welding precision, poor environmental adaptability, and difficulty in maintaining continuous and stable operation in existing technologies, this device offers a highly integrated intelligent solution. Through a control console equipped with advanced vision processing and motion control capabilities, the system achieves accurate identification and intelligent path planning for complex curved weld seams. Utilizing a multi-degree-of-freedom mechanical structure and servo drive, it achieves high-precision dynamic tracking and execution of the welding trajectory. The system innovatively integrates a dual heat dissipation mechanism of active air cooling and water cooling circulation, effectively ensuring the long-term operational stability and reliability of the vision system under harsh conditions of high temperature and dust. This device combines intelligent perception, decision-making, and robust execution capabilities, significantly improving the quality, efficiency, and automation level of turbine blade welding, and possesses broad prospects for industrial applications.

[0044] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An automatic welding device for turbine blades using a visual recognition robot, comprising a base plate (1), characterized in that, A control console (28) is fixedly connected to one side of the upper end of the base plate (1). A mounting base (2) is fixedly connected to one side of the control console (28) at the upper end of the base plate (1). The mounting base (2) is hollow in the middle. Several mounting holes (5) are evenly opened on the outer side wall of the mounting base (2). A motor (4) is fixedly connected to the upper end of the base plate (1) on the inner side of the mounting base (2). A first movable arm (3) is rotatably connected to the center position of the upper end of the mounting base (2). The output end of the motor (4) is connected to the center position of the lower end of the first movable arm (3). The upper end of the boom (3) is rotatably connected to a first rotating shaft (6), and a second movable arm (7) is fixedly sleeved on the first rotating shaft (6). One end of the first rotating shaft (6) extends through to the outside of the first movable arm (3), and one end of the first rotating shaft (6) is connected to a second motor (8). The second motor (8) is fixedly connected to the side wall of the first movable arm (3). The end of the second movable arm (7) away from the first rotating shaft (6) is connected to a welding assembly. An identification assembly is provided on the side wall of the second movable arm (7). A heat dissipation assembly is provided on both the second movable arm (7) and the first movable arm (3).

2. The automatic welding device for turbine blade vision recognition robot according to claim 1, characterized in that, The welding assembly includes a motor (10) and a mounting plate (11). The upper end of the second movable arm (7) is rotatably connected to a second rotating shaft (9). One end of the second rotating shaft (9) is fixedly connected to the center of the mounting plate (11). A welding head (12) is fixedly installed on the edge of the mounting plate (11) away from the second rotating shaft (9). The other end of the second rotating shaft (9) is connected to the motor (10), and the motor (10) is fixedly connected to the side wall of the second movable arm (7).

3. The automatic welding device for turbine blade vision recognition robot according to claim 2, characterized in that, The identification component includes an electric actuator (14) and a connecting frame (15). The second movable arm (7) has an inner cavity (13) inside. The electric actuator (14) is fixedly connected to the center of one bottom end of the inner cavity (13). The output end of the electric actuator (14) is connected to the connecting frame (15). The upper end of the second movable arm (7) on one side of the welding head (12) has a sliding groove (16). The connecting frame (15) extends to the outside of the second movable arm (7) through the sliding groove (16), and the end of the connecting frame (15) is fixedly connected to a mounting plate (17).

4. The automatic welding device for turbine blade vision recognition robot according to claim 3, characterized in that, A camera (18) is fixedly connected to the side of the mounting plate (17) away from the connecting frame (15). A transparent protective cover (19) is threaded onto the outside of the camera (18). Several heat dissipation holes (20) are opened on the side of the mounting plate (17) near the slide (16).

5. The automatic welding device for turbine blade vision recognition robot according to claim 4, characterized in that, The heat dissipation holes (20) are evenly distributed on the mounting plate (17). The diameter of the heat dissipation holes (20) gradually increases from the side closer to the mounting tube (21) to the side farther away from the mounting tube (21), and the axis of each heat dissipation hole (20) is inclined towards the camera (18) housing.

6. The automatic welding device for turbine blade vision recognition robot according to claim 1, characterized in that, The heat dissipation assembly includes a mounting pipe (21) and a fan (22). The mounting pipe (21) is fixedly installed on the end of the second movable arm (7) near the first rotating shaft (6). The fan (22) is fixedly connected inside the mounting pipe (21). A filter plate (23) is detachably connected to the port of the mounting pipe (21) away from the second movable arm (7).

7. The automatic welding device for turbine blade vision recognition robot according to claim 6, characterized in that, The mounting pipe (21) is connected to the heat dissipation holes (20) on the mounting plate (17) through an air duct.

8. The automatic welding device for turbine blade vision recognition robot according to claim 3, characterized in that, The electric push rod (14) is fitted with a spiral heat dissipation tube (24) on its outer side. Fixing blocks (25) are inserted and connected to the two ends of the spiral heat dissipation tube (24). The two fixing blocks (25) are fixedly connected to the inner wall of the inner cavity (13).

9. The automatic welding device for turbine blade vision recognition robot according to claim 8, characterized in that, A slot is provided on one side of the first movable arm (3), and a water tank (27) is detachably installed in the slot. Both ends of the spiral heat dissipation pipe (24) are connected to corrugated pipes (26). Both corrugated pipes (26) are connected to the interior of the water tank (27). The water tank (27) is equipped with a water pump, and the water outlet of the water pump is connected to one of the corrugated pipes (26).

10. The automatic welding device for turbine blade vision recognition robot according to claim 9, characterized in that, Both of the bellows (26) are made of metal corrugated hoses.

Images