A laser vision-guided mobile steel pipe bundle welding device and an arc striking control method

Abstract

The application relates to a laser vision-guided mobile steel pipe bundle welding device, which comprises a laser vision sensor, an industrial PC, a PLC controller, a motion controller and an automatic welding machine installed on the other end of a driving component and used for performing a welding task; a two-dimensional sliding table; the industrial PC is in bidirectional communication with the laser vision sensor and the PLC controller respectively, and the PLC controller is in bidirectional communication with the motion controller and the automatic welding machine respectively. The application further discloses an arc striking control method of the laser vision-guided mobile steel pipe bundle welding device. The application can accurately identify the starting point of a weld by judging the connectivity of a laser stripe in an image collected by the laser vision sensor; after the starting point of the weld is identified, the deviation between a welding gun and the weld is eliminated in time in the process of controlling the speed of the sliding table to the right to match the steel pipe bundle, and then the welding gun is controlled to strike an arc, so that the open-loop steel pipe bundle workpiece can be welded from the beginning to the end.

Description

Technical Field

[0001] This invention relates to the field of intelligent welding technology, and in particular to a laser vision-guided mobile steel pipe bundle welding device and an arc ignition control method. Background Technology

[0002] Welding is one of the most important material processing technologies in modern manufacturing, widely used in aerospace, marine, and energy fields. Currently, most production lines employ manual welding and teach-and-playback welding robots. Manual welding is costly, skilled workers are difficult to train, welding quality depends heavily on the experience of the technicians, and workers must endure long hours in harsh environments. The fumes and spatter generated during welding harm workers' health. These issues result in high labor costs, low efficiency, and inconsistent weld quality. Teach-and-playback welding robots are limited to a single welding object; changing the workpiece requires on-site recalibration, leading to a significant workload. Furthermore, teach-and-playback welding robots require multiple training sessions for complex welding tasks and struggle to adapt to changing working conditions, lacking environmental awareness.

[0003] Using sensors to perceive the environment, enabling robots to autonomously identify weld seams and perform welding, can significantly improve the efficiency and quality of pipe welding. Among these, laser vision sensors, due to their non-contact nature, high measurement accuracy, and fast measurement speed, are currently the most promising weld seam tracking sensors. Linear laser vision sensors, through steps such as acquiring laser line images, extracting weld seam features, and extracting weld seam feature points, convert the coordinates of the weld seam feature points in the image into three-dimensional spatial coordinates under the actuator. Based on the positional deviation between the weld seam feature points and the welding torch, the sensor controls the movement of the welding torch tip to the weld seam position, eliminating the deviation. For open-loop welds, it is necessary to accurately locate the starting position of the workpiece welding. If a positional deviation occurs, directly initiating the arc will result in weld misalignment. Therefore, accurately identifying the starting position of the weld seam using a laser vision sensor is crucial.

[0004] Currently, the identification of weld start points is done through algorithm design. In current welding devices, the workpiece remains stationary, and a laser vision sensor is installed at the end of the robot to identify the weld and guide the robot to weld. The drawback is that the welding at the end of the steel pipe bundle is incomplete and requires subsequent repair welding. Summary of the Invention

[0005] The primary objective of this invention is to provide a laser vision-guided mobile steel tube bundle welding device that ensures the complete welding of open-loop steel tube bundle workpieces from start to finish, thus solving the problem of complete welding of open-loop workpieces.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: a laser vision-guided mobile steel tube bundle welding device, comprising:

[0007] A laser vision sensor is mounted on one end of the drive component to capture images of the weld seam of the steel tube bundle as the drive component moves.

[0008] An industrial PC is used to process the acquired weld seam images. The laser stripe extraction process is used to obtain clear laser stripes from the weld seam images, and then the weld seam feature points are extracted based on the weld seam characteristics.

[0009] PLC controllers are used to control the working status of motion controllers and automated welding machines.

[0010] A motion controller is used to receive instructions from a PLC controller and control the motion state of the driven components.

[0011] An automated welding machine, mounted on the other end of the drive unit, is used to perform welding tasks;

[0012] A two-dimensional slide table is provided with a front sliding rail and a rear sliding rail along the direction of the steel tube bundle. An upper sliding rail and a lower sliding rail are provided along the direction perpendicular to the steel tube bundle. The driving component is placed inside the sliding rail and moves along the direction of the sliding rail under the control of the motion controller.

[0013] The industrial PC communicates bidirectionally with the laser vision sensor and the PLC controller, and the PLC controller communicates bidirectionally with the motion controller and the automated welding machine.

[0014] The two-dimensional slide is fixed to the base by a mounting bracket. The welding torch of the automated welding machine is aligned with the steel pipe bundle and perpendicular to the plane where the steel pipe bundle is located. The driving component is any one of a drive motor or a cylinder.

[0015] The laser vision sensor includes a line structure laser emitter and a CCD camera. The line structure laser emitter emits a line structure laser to illuminate the surface of the workpiece, and then the CCD camera images the weld. The acquired weld images are uploaded to a PC for processing via TCP / IP protocol. The front end of the CCD camera is equipped with a light-reducing filter to reduce the amount of light entering the camera and a filter to select the light of the required radiation band. The lens of the CCD camera is aligned with the steel tube bundle and is perpendicular to the plane where the steel tube bundle is located.

[0016] The PLC controller communicates with the motion controller, industrial PC, and automated welding machine using any one of the following methods: ProFiNet communication, TCP / IP communication, or DeviceNet communication.

[0017] Another object of the present invention is to provide an arc initiation control method for a laser vision-guided mobile steel tube bundle welding device, the method comprising the following sequential steps:

[0018] (1) Turn on the laser vision sensor. At this time, the steel pipe bundle is moving to the right. The laser vision sensor captures the weld image and uploads it to the industrial PC.

[0019] (2) The industrial PC processes the weld seam image and detects the connectivity of the laser stripes in the weld seam image in real time. If the laser stripes appear completely in the weld seam image, it is determined that the laser vision sensor has detected the weld seam starting point.

[0020] (3) After the weld start point is detected, the PLC controller sends information to the motion controller to control the welding torch to move up and down to the same horizontal line as the weld, eliminating the height error. At the same time, the component is driven to move to the right to match the speed of the steel pipe bundle. Meanwhile, the industrial PC continues to process the weld image to obtain the coordinates of the weld feature points. The pixel coordinates are converted into coordinates in the welding torch coordinate system through the calibration results of the laser vision sensor.

[0021] (4) After the height error is eliminated and the welding torch moving speed is equal to the steel pipe bundle moving speed, a signal is sent to the automated welding machine through the PLC controller to control the welding torch to start the arc from the weld seam starting point.

[0022] (5) After the welding torch starts to ignite, the drive component stops moving and the welding torch starts to weld. The welding torch moves up and down according to the weld feature point information to track the position of the weld until the weld of the entire steel tube bundle is completed, ensuring that the ends of the steel tube bundle are welded completely.

[0023] In step (3), the process of further processing the weld image on the industrial PC to obtain the coordinates of the weld feature points, and converting the coordinates of the weld feature points into coordinates in the welding torch coordinate system using the calibration results of the laser vision sensor, specifically includes the following steps:

[0024] (3a) Binarize the weld image;

[0025] (3b) The 8-neighborhood method is used to mark the connected regions in the weld image. The number of marked connected regions in the weld image is determined. If the number of connected regions is 1, it means that the laser stripe is completely displayed in the weld image. At this time, the weld image is the image of the weld starting point.

[0026] (3c) The center line of the laser stripe was extracted using the Steger algorithm;

[0027] (3d) The Hough transform is used to perform line detection on the obtained laser stripe centerline image. The pixel coordinates of the intersection of the lines in the laser stripe centerline image are the coordinates of the weld feature points.

[0028] (3e) Convert the coordinates of the two-dimensional weld feature points into three-dimensional coordinates in the welding torch coordinate system.

[0029] In step (3), eliminating height error means that the industrial PC sends the obtained weld coordinates to the motion controller through the PLC controller. The motion controller controls the welding torch to move up and down according to the deviation relationship between the weld and the welding torch to eliminate height error, so that the welding torch and the weld are on the same horizontal line.

[0030] As can be seen from the above technical solution, the beneficial effects of the present invention are as follows: First, the present invention installs a laser vision sensor on the left end of the driving component and a welding torch on the right end of the driving component. When the steel pipe bundle moves from left to right, the laser vision sensor captures the weld seam image. The captured weld seam image is transmitted to an industrial PC. After image processing, the weld seam feature points can be obtained. Then, the pixel coordinates of the weld seam feature points are converted into three-dimensional coordinates in the welding torch coordinate system through the calibration results of the laser vision sensor. The coordinates are sent to the motion controller through the PLC controller. Based on the positional deviation between the weld seam and the welding torch, the driving component is first controlled to move up and down to eliminate the height error. Then, the driving component is controlled to move left and right to the end of the steel pipe bundle. At this time, the welding torch is controlled to start arc welding, which can ensure that the open-loop steel pipe bundle workpiece can be welded from beginning to end, thus solving the problem of complete welding of open-loop workpieces. Second, by judging the connectivity of the laser stripes in the image collected by the laser vision sensor, the starting point of the weld seam can be accurately identified. Attached Figure Description

[0031] Figure 1 , 2 All of these are schematic diagrams of the device.

[0032] Figure 3 This is the circuit block diagram of this device;

[0033] Figure 4 This is a schematic diagram of the detection of the weld start point in this invention;

[0034] Figure 5 This is a flowchart of the method of the present invention. Detailed Implementation

[0035] like Figure 1 , 2 As shown, a laser vision-guided mobile steel pipe bundle 7 welding device includes:

[0036] A laser vision sensor 6 is mounted on one end of the drive component and is used to capture images of the weld seam of the steel pipe bundle 7 as the drive component moves.

[0037] An industrial PC is used to process the acquired weld seam images. The laser stripe extraction process is used to obtain clear laser stripes from the weld seam images, and then the weld seam feature points are extracted based on the weld seam characteristics.

[0038] PLC controllers are used to control the working status of motion controllers and automated welding machines.

[0039] A motion controller is used to receive instructions from a PLC controller and control the motion state of the driven components.

[0040] An automated welding machine, mounted on the other end of the drive unit, is used to perform welding tasks;

[0041] A two-dimensional slide table includes a vertical motion slide table 2 and a front-back motion slide table 1. The front-back motion slide table 1 is provided with front and rear sliding rails 5 along the direction of the steel tube bundle 7. The vertical motion slide table 2 is provided with upper and lower sliding rails 4 along the direction perpendicular to the steel tube bundle 7. The driving component is placed inside the sliding rails and moves along the direction of the sliding rails under the control of the motion controller.

[0042] like Figure 3 As shown, the industrial PC communicates bidirectionally with the laser vision sensor 6 and the PLC controller, and the PLC controller communicates bidirectionally with the motion controller and the automated welding machine.

[0043] The two-dimensional slide is fixed on the base 9 by the mounting bracket 8. The welding gun 3 of the automated welding machine is aligned with the steel tube bundle 7 and is perpendicular to the plane where the steel tube bundle 7 is located. The driving component is any one of a drive motor or a cylinder.

[0044] The laser vision sensor 6 includes a line structure laser emitter 6a and a CCD camera 6B. The line structure laser emitter 6a emits a line structure laser to illuminate the surface of the workpiece, and then the CCD camera 6B images the weld seam. The acquired weld seam image is uploaded to a PC for processing via TCP / IP protocol. The front end of the CCD camera 6B is equipped with a light-reducing filter 6d to reduce the amount of light entering the camera and a filter 6c to select the light of the required radiation band so that the acquired image appears as bright stripes against a dark background. The lens of the CCD camera 6B is aligned with the steel tube bundle 7 and is perpendicular to the plane where the steel tube bundle 7 is located.

[0045] The PLC controller communicates with the motion controller, industrial PC, and automated welding machine using any one of the following methods: ProFiNet, TCP / IP, or DeviceNet. The industrial PC receives images from the laser vision sensor 6 and can also adjust its parameters, such as camera exposure and trigger time, via software. The industrial PC contains the calibration results of the laser vision sensor 6, allowing it to convert the two-dimensional pixel coordinates of weld feature points in the image into three-dimensional coordinates in the welding torch 3 coordinate system. After obtaining the weld feature points, the industrial PC sends them to the PLC controller. The PLC controller is the central control core of the welding system, responsible for sending the weld coordinates detected by the industrial PC to the motion controller and controlling the start and stop of the drive components. It also controls the working status of the automated welding machine, controlling the arc ignition and extinguishing of the welding torch 3 to ensure timely welding. The laser vision sensor 6 and welding torch 3 are installed at the left and right ends of the drive components, respectively, and serve as the actuators of the entire system. The motion controller controls the movement of the drive components and has a built-in coordinate system that adjusts the movement of the drive components based on the positional deviation between the welding torch 3 and the weld to eliminate the deviation. Automated welding machines are used to complete welding tasks, and their arc initiation and extinguishing are controlled by a PLC controller.

[0046] like Figure 5 As shown, this method includes the following steps in sequence:

[0047] (1) Turn on the laser vision sensor 6. At this time, the steel pipe bundle 7 is moving to the right. The laser vision sensor 6 captures the weld image and uploads it to the industrial PC.

[0048] (2) The industrial PC processes the weld seam image and detects the connectivity of the laser stripes in the weld seam image in real time. If the laser stripes appear completely in the weld seam image, it is determined that the laser vision sensor 6 has detected the weld seam start point. Figure 4 As shown;

[0049] (3) After the weld start point is detected, the PLC controller sends information to the motion controller to control the welding torch 3 to move up and down to the same horizontal line as the weld to eliminate the height error. At the same time, the component is driven to move to the right to match the speed of the steel pipe bundle 7. Meanwhile, the industrial PC continues to process the weld image to obtain the coordinates of the weld feature points. The pixel coordinates are converted into coordinates in the coordinate system of the welding torch 3 through the calibration results of the laser vision sensor 6.

[0050] (4) After the height error is eliminated and the moving speed of the welding torch 3 is equal to the moving speed of the steel pipe bundle 7, a signal is sent to the automated welding machine through the PLC controller to control the welding torch 3 to start the arc from the starting point of the weld.

[0051] (5) After the welding torch 3 starts to ignite, the drive component stops moving and the welding torch 3 starts to weld. The welding torch 3 moves up and down to track the position of the weld according to the weld feature point information until the weld of the entire steel tube bundle 7 is completed, ensuring that the ends of the steel tube bundle 7 are welded completely.

[0052] In step (3), the process of further processing the weld image on the industrial PC to obtain the coordinates of the weld feature points, and converting the coordinates of the weld feature points into coordinates in the welding torch 3 coordinate system using the calibration results of the laser vision sensor 6, specifically includes the following steps:

[0053] (3a) Binarize the weld image;

[0054] (3b) The 8-domain method is used to mark the connected regions in the weld image. The number of marked connected regions in the weld image is determined. If the number of connected regions is 1, it means that the laser stripe is completely displayed in the weld image. At this time, the weld image is the image of the weld starting point. The laser vision sensor 6 emits a laser line structure. When the laser shines on the surface of the workpiece, different laser stripes will be produced for different bevel shapes. Since the workpiece is moving and the sensor is stationary, the beginning is equivalent to a laser stripe slowly and completely appearing in the image. The image mark corresponding to the moment of complete display is the weld starting point.

[0055] (3c) The center line of the laser stripe was extracted using the Steger algorithm;

[0056] (3d) The Hough transform is used to perform line detection on the obtained laser stripe centerline image. The pixel coordinates of the intersection of the lines in the laser stripe centerline image are the coordinates of the weld feature points.

[0057] (3e) Convert the coordinates of the two-dimensional weld feature points into three-dimensional coordinates in the welding torch 3-coordinate system through the coordinate system.

[0058] In step (3), eliminating height error means that the industrial PC sends the obtained weld coordinates to the motion controller through the PLC controller. The motion controller controls the welding torch 3 to move up and down according to the deviation relationship between the weld and the welding torch 3 to eliminate height error, so that the welding torch 3 and the weld are on the same horizontal line.

[0059] In summary, this invention accurately identifies the starting point of the weld by judging the connectivity of the laser stripes in the image acquired by the laser vision sensor 6; after identifying the starting point of the weld, the deviation between the welding torch 3 and the weld is eliminated in time by controlling the speed of the slide table to match the steel tube bundle 7 to the right, and then the welding torch 3 is controlled to start the arc. This process ensures that the open-loop steel tube bundle 7 workpiece can be welded from beginning to end.

Claims

1. A laser vision-guided mobile steel pipe bundle welding device, characterized in that: include: A laser vision sensor is mounted on one end of the drive component to capture images of the weld seam of the steel tube bundle as the drive component moves. An industrial PC is used to process the acquired weld seam images. The laser stripe extraction process is used to obtain clear laser stripes from the weld seam images, and then the weld seam feature points are extracted based on the weld seam characteristics. PLC controllers are used to control the working status of motion controllers and automated welding machines. A motion controller is used to receive instructions from a PLC controller and control the motion state of the driven components. An automated welding machine, mounted on the other end of the drive unit, is used to perform welding tasks; A two-dimensional slide table is provided with front and rear sliding rails along the direction of the steel tube bundle, and upper and lower sliding rails are provided along the direction perpendicular to the steel tube bundle. The driving component is placed inside the sliding rails and moves along the direction of the sliding rails under the control of the motion controller. The industrial PC communicates bidirectionally with the laser vision sensor and the PLC controller, and the PLC controller communicates bidirectionally with the motion controller and the automated welding machine. The two-dimensional slide is fixed on the base by a mounting bracket. The welding gun of the automated welding machine is aligned with the steel pipe bundle and perpendicular to the plane where the steel pipe bundle is located. The driving component is any one of a drive motor or a cylinder. The laser vision sensor includes a line structure laser emitter and a CCD camera. The line structure laser emitter emits a line structure laser to illuminate the surface of the workpiece, and then the CCD camera images the weld. The acquired weld images are uploaded to a PC for processing via TCP / IP protocol. The front end of the CCD camera is equipped with a light-reducing filter to reduce the amount of light entering the camera and a filter to select the light of the required radiation band. The lens of the CCD camera is aligned with the steel tube bundle and is perpendicular to the plane where the steel tube bundle is located. The PLC controller communicates with the motion controller, industrial PC, and automated welding machine using any one of the following methods: ProFiNet communication, TCP / IP communication, or DeviceNet communication.

2. The arc-starting control method for the laser vision-guided mobile steel tube bundle welding device according to claim 1, characterized in that: The method includes the following steps in sequence: (1) Turn on the laser vision sensor. At this time, the steel pipe bundle is moving to the right. The laser vision sensor captures the weld image and uploads it to the industrial PC. (2) The industrial PC processes the weld image and detects the connectivity of the laser stripes in the weld image in real time. If the laser stripes appear completely in the weld image, it is determined that the laser vision sensor has detected the weld start point. (3) After the weld start point is detected, the PLC controller sends information to the motion controller to control the welding torch to move up and down to the same horizontal line as the weld to eliminate the height error. At the same time, the component is driven to move to the right to match the speed of the steel pipe bundle. Meanwhile, the industrial PC continues to process the weld image to obtain the coordinates of the weld feature points. The pixel coordinates are converted into coordinates in the welding torch coordinate system through the calibration results of the laser vision sensor. (4) After the height error is eliminated and the welding torch moving speed is equal to the steel pipe bundle moving speed, the PLC controller sends a signal to the automated welding machine to control the welding torch to start the arc from the weld seam starting point; (5) After the welding torch starts to ignite, the drive component stops moving and the welding torch starts to weld. The welding torch moves up and down according to the weld feature point information to track the position of the weld until the weld of the entire steel tube bundle is completed, ensuring that the ends of the steel tube bundle are welded completely. In step (3), the industrial PC continues to process the weld image to obtain the coordinates of the weld feature points. The coordinates of the weld feature points are then converted into coordinates in the welding torch coordinate system based on the calibration results of the laser vision sensor. This process specifically includes the following steps: (3a) Binarize the weld image; (3b) The 8-neighborhood method is used to mark the connected regions in the weld image. The number of marked connected regions in the weld image is determined. If the number of connected regions is 1, it means that the laser stripe is completely displayed in the weld image. At this time, the weld image is the image of the weld starting point. (3c) The center line of the laser stripe is extracted using the Steger algorithm; (3d) The Hough transform is used to detect straight lines in the obtained laser stripe centerline image. The pixel coordinates of the intersection of straight lines in the laser stripe centerline image are the coordinates of the weld feature points. (3e) Convert the coordinates of the two-dimensional weld feature points into three-dimensional coordinates in the welding torch coordinate system; In step (3), eliminating height error means that the industrial PC sends the obtained weld coordinates to the motion controller through the PLC controller. The motion controller controls the welding torch to move up and down according to the deviation relationship between the weld and the welding torch to eliminate height error, so that the welding torch and the weld are on the same horizontal line.

Images