Adaptive weld line tracking system for welding robots
The adaptive welding line tracking system addresses the lack of adaptive control in welding robots by using an FPGA-based system with camera and line laser modules to automatically adjust the robot speed and alignment, ensuring high-quality and error-free welding.
Patent Information
- Application Number
- PCT/TR2024/051441
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2024-11-29
- Publication Date
- 2025-06-05
AI Technical Summary
Current welding robots lack adaptive control mechanisms to respond to errors such as material distortion and arc blowing during the welding process, leading to potential quality issues and increased labor, time, and financial losses.
An adaptive welding line tracking system utilizing an FPGA-based embedded system, which includes a camera and line laser module, to automatically determine the welding path geometry, adjust the robot speed in real-time, and align the welding torch on the path, thereby providing an adaptive control structure.
The system ensures high-quality, error-free welding by automatically adjusting the robot speed and alignment based on real-time data from the welding path, reducing manual programming errors and enhancing productivity.
Smart Images

Figure 00000009_0000 
Figure 00000009_0001 
Figure 00000010_0000
Abstract
Description
[0001] ADAPTIVE WELD LINE TRACKING SYSTEM FOR WELDING ROBOTS
[0002] Technical Field
[0003] The invention relates to welding robots.
[0004] The invention particularly relates to an adaptive welding line tracking system for welding robots.
[0005] State of the Art
[0006] Today, robots have begun to be preferred for jobs that are dangerous and tiring for humans in the industry due to the development of robot technologies and systems. The most common of these are welding robots, which have applications in areas such as automotive, household appliances, defence industry, medical and aviation. Welding robots automate welding applications and provide speed, time and security to the manufacturer. They can continue heavy duty cycles in an unchanging and uninterrupted structure. However, since the automation process is not under human control of the welding application, it creates an uncontrolled process against possible errors. In addition, manual programming of the welding line by an operator may cause the same precision not to be achieved in welding processes.
[0007] Current welding robots are programmed by an operator and the welding process is performed according to the parameters determined by the operator. During welding, these systems are unresponsive to known welding problems such as material distortion and arc blowing. The fact that the parameters are pre-loaded into the system and there is no control structure makes welding robots uncontrolled against possible errors during welding. This situation causes labour, time, productivity and financial losses. Different quality measurement and evaluation systems after welding have been proposed for welding operations performed with welding robots. However, in these systems, the measurement of weld quality after the welding process was provided, but an adaptive control system against errors was not developed. However, a real-time control mechanism has not been proposed on an FPGA-based embedded system for adaptive determination of the welding line.
[0008] In the state of the art, patent applications have been filed for welding robots and a few of these applications are given below. The patent document numbered EP3812105B1 in the state of the art comprises obtaining the working data of the welding robot during welding and obtaining the image data of a welded workpiece. The image data is processed to determine the parameters of the welded workpiece and these parameters are processed to determine the quality of the welded workpiece by rating the parameters using a scoring function. Only the camera is used, and the parameter setting is extracted from the images taken after the welding process. In addition, there is no control mechanism for whether the robot leaves the welding path. On the other hand, in the adaptive tracking system developed with the present invention, parameters are extracted depending on the welding path geometry obtained from the images taken before the welding process and the robot speed is adjusted depending on these parameters.
[0009] The patent document numbered CN104057202B in the state of the art relates to a system and method for remotely monitoring automatic welding of mobile robot based on FPGA. The welding robot is used to clamp a welding gun, carry the sensor system and move the welding gun in the X, Y and Z directions. By using the camera and line laser module, the welding torch is aligned on the welding path and remotely monitored using the FPGA. The mechanism in the invention is a Cartesian system and its movement capability is only in two axes. Therefore, the term welding robot speed cannot be used for this mechanism. The welding speed will be constant. On the other hand, the adaptive control system developed with the present invention is designed to control the speed of multi-axis welding robots. The speed and alignment parameters of the robot are calculated on the FPGA according to the welding path data received from the camera before the welding process.
[0010] The patent document numbered CN108127241 A in the state of the art relates to an intelligent control system of welding robots, comprising a welding robot, a CAD module, a sensor module and a control system, where the welding robot, CAD module and sensor module are connected to the control system, respectively; and the control system is used to control the welding robot to perform the welding process. FPGA is used for communication. On the other hand, in the present invention, FPGA is used for calculation processes. In this way, the speed of the robot can be adjusted in real time in an adaptive manner against instantaneous changes in the welding path. In addition, being compatible with multi-axis robots allows for high-quality welding on materials with complex geometry. The ability to adjust the speed parameter of the welding robot will ensure high-quality and error-free welding depending on the variable welding path geometry.
[0011] The patent document numbered CN1 12621030A in the state of the art provides a method for creating welding trajectories for energy transmission tower nodal points. Using this solution, the welding trajectory is automatically created before the welding process by a software program, and the welding robot arm transfers the welding accuracy and the geometry of the welding material to a software on the computer with the help of an adaptable 3D scanner. The operator manually marks the path to be welded on the transferred 3D image. When the operator starts the process, the welding robot performs the welding process by following the points marked in the software. On the other hand, in the present invention, the geometric structure of the welding path is automatically defined with the help of a camera. In addition, with a line laser module, it is ensured that the welding robot's torch is aligned on the welding path. The FPGA used allows the operations to be performed in real time.
[0012] The Aim of the Invention
[0013] In order to eliminate the disadvantages in the state of the art, one aim of the invention is to provide a real-time device that has been designed on an FPGA-based embedded system to prevent errors that may affect the welding quality in welding works performed by welding robots before they occur, to automate the manual programming process and to provide an adaptive control structure. The designed system has been implemented on Field Programmable Gate Arrays (FPGA), which have advantages such as parallel processing against other processor types and instant response to multiple tasks, so that data calculation can be performed in real time with image processing and artificial intelligence techniques.
[0014] Another aim of the invention is to ensure that the system determines the speed and welding current of the welding robot according to the width of the welding line with artificial intelligence techniques, thus achieving optimum control and accordingly ensuring that the robots perform error-free, high-quality welding and automating the robot programming process.
[0015] Another aim of the invention is to shorten the programming processes of welding robots with the automation process and to create a system that is controlled against errors. The designed system aims to eliminate the lack of adaptive control mechanisms in the sector for welding robots. Another aim of the invention is to prevent errors that may affect the welding quality in welding operations performed by welding robots before they occur, to automate the manual programming process and to provide an adaptive control structure for welding robots.
[0016] Another aim of the invention is to eliminate the current deficiency in welding robot systems by the present system that has been developed as a real-time embedded system design that allows the welding line parameters to be automatically determined in advance and the welding parameters to be changed adaptively.
[0017] Another aim of the invention is to make it possible to eliminate measurement errors made in manual processes and to instantly re-adjust the system against assembly errors and errors that occur later due to external factors by determining the welding line before the welding process via sensors.
[0018] Another aim of the invention is to eliminate the fixed parameter structure in existing systems and to provide a feedback system to the robot in factors such as welding gap change thanks to the adaptive parameter structure. Thus, since the parameters of the welding process will change instantly in factors such as welding gap change and welding line deviation, the quality and strength of the weld can be kept at a constant rate.
[0019] Another aim of the invention is to ensure that the welding line is determined using camera and laser sensors and appropriate artificial intelligence techniques, unlike the studies in the literature.
[0020] Another aim of the invention is to ensure that the processes can be done in real time, the methods are implemented on the FPGA embedded processor, which has parallel processing capability and has a higher operating speed than other processors.
[0021] Another aim of the invention is to eliminate the lack of adaptive structure in existing welding robots, and to increase the time, cost and efficiency gains of people in production, especially in industrial applications where precision is important and requires high technology.
[0022] In order to provide the above advantages, the invention provides an adaptive welding line tracking system for a welding robot that is fixed to the ground comprising a lower arm on a body and an upper arm that moves articulated with this arm, preferably a torch for welding at the end of the upper arm. Description of Drawings
[0023] Figure 1 is the general side view of the welding robot system that is the subject of the invention,
[0024] Figure 2 is the general perspective view of the welding robot system that is the subject of the invention,
[0025] Figure 3 is the general perspective side views of the sensors and control units of the system that is the subject of the invention.
[0026] Description of Reference Numbers
[0027] I .Robot
[0028] I I .Gbvde
[0029] 12. Lower arm
[0030] 12. Upper arm
[0031] 2. Control Unit
[0032] 21 .Connection cable
[0033] 2. Sensor Unit
[0034] 31 .Line laser module
[0035] 32. Camera module
[0036] 33. Filter slider
[0037] 34. Torch mounting slot
[0038] 4. Torch
[0039] Detailed Description of the Invention
[0040] The invention is an adaptive welding line tracking system for welding robots (1 ). Figure 1 shows the side view and Figure 2 shows the perspective view of the welding robot (1 ) system which is the subject of the invention. In the figures, a welding robot (1 ) fixed to the ground is seen. The welding robot (1 ) comprises a lower arm (12) on a body (11 ) and an upper arm (13) that moves articulated with said arm (12).
[0041] Again, as can be seen in Figures 1 and 2, there is a control unit (2) preferably mounted on the robot (1 ) body (1 1 ) and close to the ground. Again, a torch (4) is preferably mounted at the end of the upper arm (13) for the welding process. Optionally, a sensor unit (3) is connected at the end of said upper arm (13). The communication between said control unit (2) and said sensor unit (3) is provided by the connection cable (21 ). This connection cable (21 ) is optionally positioned inside the robot (1 ). Figure 3 shows the general perspective side views of the sensor (3) and control units
[0042] (2) of said system. According to the figure, the control unit (2) has a box form optionally. The connection cables (21 ) are positioned inside the control unit (2). The sensor unit
[0043] (3) is seen in Figure 3. A line laser module (31 ) and a filter slider (33) are positioned on the sensor unit (3), and a camera module (32) is positioned inside this slider (33). Again, a torch mounting slot (34) is formed on the sensor unit (3). This slot (34) provides the mounting with said torch (4).
[0044] With said system, the process steps of extracting parameters depending on the welding path geometry obtained from the images taken by means of the camera module (32) before the welding process, automatic determination of the welding process path and adaptive control of the speed of the robot (1 ), adjustment of the speed of the robot (1 ) with the FGPA (2) located in the control unit depending on these parameters, and alignment of the welding robot (1 ) torch (4) on the welding path with the line laser module (31 ) are carried out.
[0045] For a welding robot (1 ) that is the subject of the invention, the automatic determination of the welding process path and the adaptive control function of the robot (1 ) speed are realised with the FPGA (Field Programmable Gate Array) (2), camera module (32), line laser module (31 ), and fuzzy logic controller control unit (2) that are in the control unit. FPGA is a digital integrated circuit consisting of programmable logic blocks and the interconnections between these blocks and has a wide range of applications. It is manufactured for the purpose of realising the logic functions required by the designer. Before the welding process, the camera module (32) extracts parameters depending on the welding path geometry obtained from the images taken and the speed of the robot (1 ) is adjusted depending on these parameters.
[0046] The control system is designed to perform the speed control of multi-axis welding robots (1 ). The speed and alignment parameters of the robot are calculated on the FPGA according to the welding path data received from the camera (32) before the welding process.
[0047] The geometric structure of the welding path is automatically defined with the help of a camera (32). In addition, a line laser module (31 ) is used to align the welding robot (1 ) torch (4) on the welding path. The FPGA used allows the operations to be performed in real time In the control function, unlike the existing inventions, the welding line parameters are determined by the FPGA before the welding process is performed and the speed of the welding robot (1 ) is adjusted in real time according to these parameters. In order to ensure that the operations can be performed in real time, the application is run on FPGA-based embedded hardware. The sensors (3) placed on the upper arm (13) section of the welding torch (4) enable the welding torch (4) to be used effectively in narrower areas.
Claims
CLAIMS1 . An adaptive welding line tracking system for a welding robot (1 ) that is fixed to the ground, having a lower arm (12) on a body (11 ) and an upper arm (13) that moves articulated with said arm (12), preferably a torch (4) for welding at the end of the upper arm (13), comprising• an embedded fuzzy logic controller control unit (2) based on FGPA, preferably mounted close to the robot (1) body (11 ) and the ground part,• a sensor unit (3) preferably at the end of said upper arm (13) and in connection with said control unit (2),• a line laser module (31 ) that is positioned on said sensor unit (3), and enables the alignment of the welding robot (1 ) torch (4) on the welding path,• filter slider (33) with camera module (32) inside, and• a camera module (32) that is placed in the filter slider (33) and enables parameter extraction depending on the weld path geometry obtained from the images taken, automatic determination of the welding process path and adaptive control of the robot (1 ) speed, and adjustment of the robot (1 ) speed with the FGPA-based embedded system depending on these parameters.
2. A system according to Claim 1 , comprising a connection cable (21 ) to enable communication between the control unit (2) and the sensor unit (3).
Citation Information
Patent Citations
Autonomous mobile robot system for metal-inert-gas (MIG) / metal-active-gas (MAG) multilayer multipass welding of large thick plate
CN102689083A
System and method for remotely monitoring automatic welding of mobile robot based on FPGA
CN104057202A