A method of welding a cruciform corner joint

Abstract

The application provides a cross plate corner joint welding method, which comprises the following steps: performing the execution track according to a preset welding sequence, completing multi-layer and multi-pass welding of a main joint and corner welding, and sending a posture switching signal to a positioner controller; the positioner controller drives the cross plate workpiece to rotate or overturn to a next welding joint welding posture after receiving the posture switching signal, and feeds back a posture in-place signal; after receiving the posture in-place signal, an independent track program of a stiffened plate corner joint is started, and multi-layer and multi-pass corner welding of the stiffened plate periphery and corner welding are performed; a continuous path motion control mode is adopted to drive a welding gun to move along a correction track in a welding stage, and a point-to-point motion control mode is adopted to drive the welding gun to move to a next starting point in a non-welding displacement stage.

Description

Technical Field

[0001] This invention relates to the field of information technology, and in particular to a method for welding a corner joint of a cross plate. Background Technology

[0002] In the field of large steel structure manufacturing, especially in the production of wind turbine towers, power transmission towers, and support structures for heavy industrial equipment, the cross plate section, as a critical connection area bearing extreme loads, directly determines the overall stability and service life of the entire structure through its welding quality. The cross plate is typically assembled by welding main leg plates and multiple stiffening plates in a cross-shaped manner, resulting in a highly complex spatial distribution of weld seams.

[0003] Current welding methods primarily rely on pre-programmed fixed trajectories or simple teach-and-play reproduction. However, in actual production, due to dimensional tolerances in the machining of cross-plate workpiece components, unavoidable cumulative deformation during assembly, and the dispersed arrangement of multiple stiffening plates in three-dimensional space, significant deviations occur between the actual positions of the main seam and the corner welds of the stiffening plates and the designed preset trajectories. This deviation is not a single-direction overall offset, but rather a complex deviation resulting from multiple superimposed factors, such as local distortion of the main seam, random distribution of the start and end points of the stiffening plate welds, and changes in the included angle between adjacent welds. Existing welding equipment struggles to perceive these real-time spatial changes and can only forcibly execute the original program, leading to defects such as welding torch deviation from the actual weld, incorrect molten pool position, uneven weld width, and even undercut and incomplete fusion.

[0004] The direct consequences of these positional deviations are that the cross-plate welding process must be frequently interrupted for manual grinding and repair welding. When the main seam is filled with multiple layers and passes, incomplete fusion between layers is easily caused by trajectory drift. Furthermore, the stiffening plate corner welds require extensive rework due to the difficulty of re-aligning after posture changes. More importantly, the main seam and stiffening plate welds intersect and are closely adjacent in space. Once a defect occurs in the main seam weld, it directly affects the forming quality of the subsequent stiffening plate welds. Conversely, the welding accuracy of the stiffening plate welds decisively restricts the reinforcement effect in the corner area of ​​the main seam. This mutual constraint and quality transfer relationship between the main seam and stiffening plate welds due to positional deviations means that the deviation of a single weld can gradually amplify into overall welding control failure in the entire cross-plate intersection area.

[0005] Therefore, how to accurately obtain the actual coordinate positions of the main seam and the corner joint welds of each stiffening plate in three-dimensional space in real time during the cross plate welding process, and use this as a basis to dynamically correct the preset trajectory so that the welding path completely matches the actual weld, while ensuring the sequential coordination and quality consistency of the multi-layer welding of the main seam and the layered corner joint welding of the stiffening plate under the position correlation conditions, has become a key issue in ensuring the welding accuracy and reliability of complex cross plate structures. Summary of the Invention

[0006] This invention provides a method for welding a corner joint of a cross plate, mainly comprising:

[0007] The position of the main seam and stiffening plate corner weld on the cross-plate workpiece is detected by a positioning sensor to obtain real-time spatial coordinate data of the weld. The preset welding trajectory program is corrected based on the real-time spatial coordinate data to generate an execution trajectory that matches the actual weld position. The execution trajectory is executed according to the preset welding sequence to complete the multi-layer multi-pass welding of the main seam and the corner welding, and a posture switching signal is sent to the positioner controller. After receiving the posture switching signal, the positioner controller drives the cross-plate workpiece to rotate or flip to the next weld welding posture and feeds back the posture arrival signal. After receiving the posture arrival signal, the independent trajectory program of the stiffening plate corner weld is started to execute multi-layer multi-pass corner welding and corner welding around the stiffening plate. During the welding stage, the continuous path motion control mode is used to drive the welding torch to move along the corrected trajectory. During the non-welding displacement stage, the point-to-point motion control mode is used to drive the welding torch to move to the next starting point.

[0008] Furthermore, the step of detecting the actual position of the main seam and stiffening plate fillet weld on the cross-plate workpiece using a positioning sensor includes: the positioning sensor identifying the starting position and edge position of the main seam and stiffening plate fillet weld, and obtaining real-time spatial coordinate data of the main seam and stiffening plate fillet weld; comparing the real-time spatial coordinate data with the coordinates of the trajectory points in the preset welding trajectory program to determine the coordinate offset; and adjusting all trajectory points of the preset welding trajectory program as a whole according to the coordinate offset to generate the execution trajectory that matches the actual weld position.

[0009] Furthermore, the execution of the execution trajectory according to the preset welding sequence includes: calling the stored main seam trajectory program, completing the multi-layer multi-pass welding and the corner welding in the order from the corner weld between the sub-plates to the corner weld between the sub-plate and the bottom plate; after all the main seam welding is completed, generating the attitude switching signal and sending it to the positioner controller.

[0010] Furthermore, after receiving the attitude switching signal, the positioner controller drives the cross plate workpiece to rotate or flip to the next weld welding attitude, including: the positioner controller controls the servo system to drive the cross plate workpiece to flip or rotate to a preset angle according to the attitude switching signal, so that the stiffening plate corner weld is in the reachable position of the welding torch; the positioner controller generates the attitude arrival signal after monitoring the workpiece attitude; after receiving the attitude arrival signal and confirming that the workpiece attitude is stable, the independent trajectory program of the stiffening plate corner weld is started; the independent trajectory program includes multiple layers and multiple paths and corner wrapping paths for the stiffening plate periphery corner weld, and adjacent paths are connected by a smooth transition.

[0011] Furthermore, the step of starting an independent trajectory program for the stiffening plate corner weld after receiving the attitude positioning signal includes: manually completing the stiffening plate point splicing and positioning, loading a pre-stored stiffening plate trajectory program; and applying the previously acquired real-time spatial coordinate data to correct the stiffening plate trajectory program.

[0012] Furthermore, the step of using a continuous path motion control mode to drive the welding torch to move along the correction trajectory during the welding stage, and using a point-to-point motion control mode to drive the welding torch to move to the next starting point during the non-welding displacement stage, includes: the continuous path motion control mode generating a straight line or arc trajectory to ensure that the welding torch moves continuously along the correction trajectory to complete the corner weld; and the point-to-point motion control mode determining the coordinates of the next starting point and driving the welding torch to move rapidly during the non-welding displacement stage.

[0013] Furthermore, the method of using continuous path motion control mode to drive the welding torch to move along the correction trajectory during the welding stage includes: establishing a workpiece coordinate system by teaching at least three reference points on the cross-plate workpiece; converting all the execution trajectories into coordinates under the workpiece coordinate system; defining the tool center point before trajectory execution, calculating the tool center point value by teaching multiple reference points and applying it to the trajectory coordinate transformation; and compensating for the tool center point deviation through online correction instructions during trajectory execution.

[0014] Furthermore, the correction of the preset welding trajectory program includes: editing the trajectory point coordinates and path sequence of the preset welding trajectory program through a programming interface, and smoothly optimizing the connection of adjacent weld trajectories.

[0015] The technical solutions provided by the embodiments of the present invention may include the following beneficial effects:

[0016] This invention discloses a welding method for corner joints of cross-plate workpieces, addressing the unique business scenario where the actual position of the main seam and stiffening plate corner joint welds deviates significantly from the preset trajectory due to machining errors, assembly deformation, and the spatial dispersion of multiple stiffening plates. This results in difficulty in ensuring welding accuracy, easy generation of defects, and low efficiency. The invention utilizes a positioning sensor to detect the three-dimensional spatial coordinates of the main seam and stiffening plate welds in real time, performing precise correction of the preset trajectory through a combination of overall and segmented methods to generate an execution trajectory that perfectly matches the actual weld. Subsequently, a motion control mode combining continuous path welding based on the workpiece coordinate system and point-to-point rapid displacement is adopted. First, multi-layer, multi-pass welding of the main seam and corner welding are completed. Then, a positioner precisely switches the posture, and finally, layered corner joint and corner reinforcement welding are performed on the stiffening plates using an independent trajectory program. The entire process achieves dynamic online compensation for deviations and path smoothing optimization. This invention significantly improves the high-precision matching capability of weld positions in complex cross-plate workpieces, effectively avoids welding defects caused by positional deviations, ensures structural strength and fatigue performance, and greatly improves welding efficiency and production consistency. Attached Figure Description

[0017] Figure 1 This is a flowchart of a cross-plate corner weld method according to the present invention. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0019] like Figure 1 The specific welding method for a cross-plate fillet weld in this embodiment may include:

[0020] This invention provides a method for welding fillet welds on cross-plate workpieces, aiming to achieve high-precision welding control of welds on complex cross-plate structures, ensuring a high degree of matching between the welding trajectory and the actual weld position, while optimizing welding efficiency and quality. The technical solution of this invention is described in detail below with reference to specific embodiments. Step S1: The robot control system uses a positioning sensor to detect the actual position of the main seam and stiffening plate fillet welds on the cross-plate workpiece, acquiring real-time spatial coordinate data of the welds, and correcting the preset welding trajectory program based on this data to generate an execution trajectory that matches the actual weld position. Specifically, the positioning sensor can be a laser sensor or a contact probe, which acquires the actual spatial position information of the welds by scanning or contacting the workpiece surface. This information is stored in three-dimensional coordinates, such as the transverse, longitudinal, and height coordinates of each key weld point. The robot control system compares this real-time data with the pre-programmed welding trajectory. Upon detecting a positional deviation, the system automatically adjusts the trajectory parameters to ensure that subsequent welding operations accurately cover the actual weld position. This method is particularly suitable for cross-plate welding scenarios, effectively avoiding welding defects caused by deviations. Furthermore, in step S1, the detection process of the positioning sensor can be divided into multiple sub-steps to ensure the comprehensiveness and accuracy of data acquisition.

[0021] Step S11: The positioning sensor first scans the main seam, identifying its start and end points, as well as key turning points along the way. The main seam is typically the weld connecting the main components in a cross-plate workpiece, such as a butt weld or corner weld between two main plates. The sensor records the three-dimensional coordinate data of each key point through point-by-point scanning, forming a spatial distribution map of the main seam. Next, the sensor moves to detect the corner weld of the stiffening plate, identifying the weld edge position at the connection between the stiffening plate and the main plate or base plate. As a component that enhances structural strength, the stiffening plate's weld positions are often distributed on multiple sides of the workpiece; therefore, the detection must cover all relevant areas to ensure no omissions. Step S12: After acquiring the real-time spatial coordinate data of the main seam and the corner weld of the stiffening plate, the robot control system preprocesses and analyzes this data. The system first performs noise reduction on the acquired coordinate points, eliminating abnormal data points caused by sensor jitter or uneven workpiece surfaces. Subsequently, the system matches the processed coordinate data with the standard trajectory points in the preset welding trajectory program, calculating the offset of each key point. The offset includes three dimensions: lateral offset, longitudinal offset, and height offset. The system uses these offsets to determine the degree of difference between the actual weld position and the designed position. For example, if the actual position of a main weld segment is offset 5 mm to the left compared to the designed position, the system will record this deviation value and apply it to subsequent trajectory correction. In step S13, based on the above offsets, the robot control system adjusts the preset welding trajectory program as a whole to generate a new execution trajectory. The preset welding trajectory program is usually pre-programmed based on workpiece design drawings or standard models, and includes the movement path of the welding torch in space and welding parameters. However, due to potential processing errors or assembly deformations in the actual workpiece, simply relying on the preset trajectory often cannot meet the accuracy requirements. Therefore, the system will correct the coordinates of each trajectory point in the trajectory program based on the real-time detected offsets. For example, if the main weld is offset 3 mm to the right, the system will uniformly increase the lateral coordinates of all relevant trajectory points by 3 mm, while maintaining the shape of the trajectory and the welding sequence unchanged. The adjusted execution trajectory can perfectly match the actual weld position, providing a reliable basis for subsequent welding operations.

[0022] For example, in one possible implementation, assuming the workpiece is a cross plate of a tower crane, the main seam is the fillet weld at the connection between the main leg plate and the base plate, and stiffening plates are distributed on the sides of the main leg plate to enhance the structural stability of the cross plate. The positioning sensor first scans along the connection between the main leg plate and the base plate, recording the main seam's starting point as the lower end of the connection and its ending point as the upper end, recording the 3D coordinates of a key point every 10 centimeters along the way. Then, the sensor moves to the connection between the stiffening plate and the side of the main leg plate, identifying the edge positions of the welds around the stiffening plate, forming a complete weld position dataset. The robot control system compares this data with a preset trajectory and finds that the main seam has shifted downwards by 4 millimeters, while the stiffening plate weld positions have local distortions, with some areas shifting by as much as 6 millimeters. Based on these offsets, the system recalculates the trajectory point coordinates, generating an execution trajectory that perfectly matches the actual weld. This method can significantly improve the welding accuracy of the cross plate, reduce weld defects caused by positional deviations, and ensure that the connection strength of the cross plate meets design requirements.

[0023] In another embodiment, the detection method and trajectory correction strategy of the positioning sensor can be adjusted for different types of cross-plate workpieces. For example, for the cross-plate of a power transmission tower, the main seam may have a complex zigzag shape, while the stiffening plates are numerous and densely distributed. During detection, the positioning sensor can employ a high-density scanning mode, increasing the acquisition frequency of key points, for example, recording a coordinate point every 5 centimeters to capture the shape changes of the zigzag weld. When correcting the trajectory, the robot control system, considering the characteristics of the zigzag weld, adopts a segmented fitting method, decomposing the entire weld trajectory into multiple small segments, calculating the offset for each segment, and adjusting the coordinates of the trajectory points. This segmented correction method can better adapt to the complex shape of the cross-plate weld, ensuring that the welding torch is accurately positioned in each area and avoiding uneven welding caused by trajectory errors.

[0024] It should be noted that the selection and configuration of the positioning sensor also affect the detection accuracy and trajectory correction effect. In one embodiment, if there is a lot of rust or oil on the workpiece surface, the reflected signal of the laser sensor may be unstable. In this case, a contact probe can be used as an auxiliary detection tool. The contact probe obtains accurate position data by directly contacting the edge of the weld. Although the detection speed is slower, it has higher reliability in complex environments. After receiving the probe data, the robot control system also performs noise reduction and offset calculation to generate the execution trajectory. This method is suitable for scenarios with extremely high accuracy requirements, such as the welding of key components in nuclear power equipment manufacturing. It can ensure the accuracy of weld position detection and avoid errors caused by environmental interference. In step S2, the robot control system executes the above-generated execution trajectory according to the preset welding sequence. First, it completes the multi-layer, multi-pass welding of the main seam and the corner welding, and then sends an attitude switching signal to the positioner controller. Specifically, the preset welding sequence is determined in advance according to the structural characteristics of the workpiece and the welding process requirements. Usually, the main seam welding is given priority to ensure the strength and stability of the main connection parts. After the main seam welding is completed, the system automatically generates a posture switching signal to notify the positioner controller to adjust the workpiece posture in preparation for the subsequent welding of the stiffening plate fillet weld. This staged and sequential welding method effectively avoids the impact of workpiece deformation on the position of subsequent welds during the welding process, while improving overall welding efficiency. Furthermore, in step S2, the main seam welding process can be broken down into multiple sub-steps to ensure welding quality and process consistency.

[0025] Step S21: The robot control system first calls the stored main seam trajectory program and performs multi-layer, multi-pass welding on the main seam according to a preset sequence. Multi-layer, multi-pass welding refers to sequentially filling multiple weld passes at the same weld location. After each weld pass is completed, the welding torch will proceed along the same path to weld the next layer until the weld reaches the designed thickness. For example, for a thicker main seam, three weld passes may be required. The width and height of each weld pass must be strictly controlled to avoid stress concentration or weld defects. During the welding process, the welding torch moves precisely along the execution trajectory to ensure that the start and end positions of each weld pass are consistent, forming a uniform weld appearance. Step S22: After completing the multi-layer, multi-pass welding of the main seam, the robot control system starts the corner welding program to reinforce the corner areas at both ends of the main seam. Corner welding refers to additional welding at the junction of the weld and the workpiece edge to enhance the connection strength and fatigue resistance of the weld. For example, in the fillet weld of a beam-column joint, the fillet weld typically covers an area of ​​10 cm on each side of the weld. The welding torch moves along a small circular trajectory to form a continuous fillet weld bead. After the fillet weld is completed, the system performs a rapid inspection of the welded area. Once it confirms that there are no obvious defects, it records the completion status of the main weld, preparing for subsequent posture switching. In step S23, after all welding tasks of the main weld are completed, the robot control system generates a posture switching signal and sends it to the positioner controller via the communication interface. The posture switching signal contains specific parameters of the target posture, such as the angle of rotation or the direction of rotation of the workpiece. These parameters are preset according to the workpiece structure and the position of the stiffening plate weld, aiming to adjust the stiffening plate fillet weld to the optimal position reachable by the welding torch. After the signal is sent, the system enters a waiting state, ready to receive posture positioning information from the positioner controller to start the next stage of welding operation. For example, in one possible implementation, assume that the cross plate workpiece is a cross plate of a heavy industrial equipment support structure, and the main weld is located at the connection between the bottom plate of the cross plate and the main leg plate, requiring a high weld thickness. The robot control system first executes the main seam trajectory program, controlling the welding torch to complete the first layer of weld filling, with a weld width controlled at approximately 8 mm, covering the entire length of the main seam. Subsequently, the system adjusts welding parameters, increasing the welding current and wire feed speed to complete the second and third layers of weld filling, each layer approximately 3 mm high, ensuring the overall weld thickness meets design requirements. At both ends of the main seam, the welding torch moves according to a preset corner trajectory, forming a 12 mm wide corner weld to enhance the connection strength of the weld edges. After the main seam welding is completed, the system generates a posture switching signal, instructing the positioner controller to rotate the cross plate workpiece 90 degrees, so that the side stiffening plate weld faces upwards, facilitating subsequent welding operations. This layered, staged welding method effectively controls the welding deformation of the cross plate, ensuring the overall stability of the cross plate structure.

[0026] In another embodiment, the optimization of the main seam welding sequence can be adjusted according to the stress characteristics of the workpiece. For example, for the cross plate of a wind turbine tower, the main seam is the corner weld between the main leg plate and the base plate of the cross plate, which may bear a large vertical load and shear force. The welding sequence should prioritize completing the weld filling of the critical stress area. The robot control system first performs multi-layer welding on the core area in the middle of the main seam to ensure the strength of the core connection part of the cross plate, and then gradually expands to both ends to complete the filling of the remaining welds and corner welding. This welding sequence that expands from the center to both sides can effectively disperse the welding stress of the cross plate and avoid deformation of the cross plate structure caused by welding on one side. After welding is completed, the system generates a posture switching signal, instructing the positioner controller to flip the cross plate workpiece to a side posture, creating conditions for welding the stiffening plate weld.

[0027] It should be noted that during the main seam welding process, the robot control system can also incorporate real-time monitoring technology to further improve welding quality. For example, the system can determine whether the weld filling is uniform by monitoring fluctuations in welding current and voltage. If abnormal fluctuations are detected, the system will automatically adjust welding parameters or pause the operation, awaiting manual inspection. This real-time monitoring method can promptly detect potential defects, such as weld porosity or lack of fusion, thereby preventing subsequent processes from being affected. After the main seam welding is completed, the system will store the monitoring data in a log file, providing a basis for subsequent quality traceability, and simultaneously generate a posture switching signal to proceed to the next stage of operation. In step S3, after receiving the posture switching signal, the positioner controller drives the cross plate workpiece to rotate or flip to the next weld welding posture and feeds back the posture positioning signal to the robot control system. Specifically, the positioner controller is the core component responsible for workpiece posture adjustment. By receiving signals sent by the robot control system, it controls the servo motor or hydraulic system to drive the workpiece movement. The goal of workpiece posture adjustment is to adjust the stiffening plate corner joint weld to a position accessible to the welding torch, typically at a horizontal or near-horizontal angle, so that the welding torch can perform welding operations at the optimal angle. After the posture adjustment is completed, the positioner controller uses sensors to detect whether the workpiece is stable and generates a posture positioning signal, which is then sent to the robot control system to notify it that the next stage of the welding process can begin. Furthermore, in step S3, the workpiece posture adjustment process can be broken down into multiple sub-steps to ensure adjustment accuracy and safety.

[0028] Step S31: After receiving the attitude switching signal, the positioner controller first parses the signal content and extracts the target attitude parameters, such as a rotation angle of 90 degrees or a clockwise rotation direction. Then, based on these parameters, the controller calculates the drive commands for the servo system and controls the fixture or turntable on the positioner to move the workpiece. During workpiece movement, the controller monitors the rotation speed and angle changes in real time to ensure smooth movement and prevent workpiece swaying or fixture loosening due to excessive speed. Step S32: After the workpiece moves to the target attitude, the positioner controller detects whether the workpiece is accurately positioned using an angle sensor or limit switch. For example, if the target attitude is workpiece side-up, the sensor detects whether the angle between the workpiece side and the horizontal plane reaches a preset value. If the angle deviation is within the allowable range, such as less than 0.5 degrees, the attitude adjustment is considered complete. Next, the controller further checks whether the workpiece is stable, for example, by detecting slight swaying using a vibration sensor. If stability is confirmed, the controller generates an attitude positioning signal. This signal contains specific parameters of the workpiece's current attitude, such as the rotation angle and stability status, and is sent to the robot control system via the communication interface. In step S33, after receiving the posture positioning signal, the robot control system first verifies the validity of the signal, for example, by checking whether the posture parameters in the signal are consistent with the expectation. If the verification passes, the system records the current posture state of the workpiece and prepares to start the welding program for the stiffening plate fillet weld. It should be noted that if the signal verification fails, for example, if the posture parameters deviate too much from the target value, the system will send a readjustment command to the positioner controller, requiring it to readjust the workpiece posture again until the expected state is reached. This verification mechanism can effectively avoid welding position deviations caused by posture adjustment errors, ensuring the smooth progress of subsequent welding operations. For example, in one possible implementation, assuming the cross plate workpiece is a cross plate of a wind turbine tower, after receiving the posture switching signal, the positioner controller analyzes the signal content and finds that the target posture is for the cross plate workpiece to rotate 90 degrees, so that the stiffening plate weld is facing upwards. The controller then controls the servo motor to drive the turntable to rotate at a speed of 5 degrees per second, and the cross plate workpiece slowly moves to the target position under the drive of the turntable. During rotation, the controller monitors the workpiece's posture in real time using an angle sensor. When the rotation angle reaches 89.8 degrees, the drive stops. Subsequently, a vibration sensor checks the workpiece's stability, and after confirming no shaking, a posture positioning signal is generated. This signal contains parameters indicating a rotation angle of 89.8 degrees and a normal stability state, and is sent to the robot control system via a communication interface. Upon receiving the signal, the system verifies that the angle deviation is within the allowable range, records the cross-plate workpiece's posture state, and prepares to initiate the stiffening plate welding program. This precise posture adjustment method ensures that the welding torch approaches the stiffening plate weld at the optimal angle, improving the quality and efficiency of cross-plate welding. In another embodiment, for the posture adjustment of large cross-plate workpieces, the positioner controller can adopt a step-by-step adjustment strategy to reduce the risks during the adjustment process.For example, for heavy-duty cross-plate workpieces weighing over 10 tons, the controller first rotates the workpiece to an intermediate posture, such as a 45-degree angle, pauses briefly, and uses sensors to detect whether the workpiece's center of gravity is stable. Then, the controller continues to drive the workpiece to the target posture, such as a 90-degree angle, checks stability again, and generates a posture positioning signal. This step-by-step adjustment method effectively avoids workpiece instability caused by excessively large rotation angles at once, making it particularly suitable for welding ultra-heavy or ultra-large workpieces, ensuring the safety and reliability of the posture adjustment process. It should be noted that the drive method and adjustment accuracy of the positioner controller can be optimized according to the characteristics of the workpiece and welding requirements. For example, for smaller but structurally complex workpieces, a high-precision servo system can be used to drive the positioner, ensuring that the posture adjustment error is controlled within 0.2 degrees. For larger but structurally simple workpieces, a hydraulic system can be used to drive the positioner, prioritizing load-bearing capacity and stability. After the posture adjustment is completed, the controller generates detailed posture positioning signals based on the actual state of the workpiece, providing reliable feedback information to the robot control system and ensuring seamless transitions to subsequent welding operations. In step S4, after receiving the attitude positioning signal, the robot control system starts an independent trajectory program for the stiffening plate fillet weld, executing multi-layer, multi-pass fillet welding and corner welding around the stiffening plate. Specifically, the arrival of the attitude positioning signal indicates that the cross plate workpiece has been adjusted by the positioner to a stable spatial orientation suitable for stiffening plate welding. At this time, the accessibility of the welding torch relative to the stiffening plate weld and the welding angle have met the process requirements. The robot control system then calls the independent trajectory program specifically for the stiffening plate fillet weld from the storage unit. This program differs significantly from the trajectory program previously used for the main weld in path planning logic, layer allocation strategy, and transition connection method, to adapt to the closed, multi-segmented, and spatially dispersed characteristics typically exhibited by stiffening plate welds. Furthermore, in step S4, the initiation and execution of the stiffening plate independent trajectory program can be decomposed into several continuous actions.

[0029] Step S41: The robot control system first loads the pre-stored basic data of the stiffening plate corner weld trajectory. This data includes the theoretical path point sequence of the welds around the four edges of the stiffening plate, the layer planning for each edge weld, the suggested values ​​for the width and height of each weld layer, and the description of the reinforcement path in the corner area. After loading, the system immediately reads the real-time spatial coordinate data of the stiffening plate obtained by the positioning sensor in step S1 and applies this actual position information to the coordinate transformation of the basic trajectory. Step S42: Based on the real-time spatial coordinate data, the independent trajectory program of the stiffening plate is specifically corrected. Unlike the overall translation adjustment of the main seam trajectory, the stiffening plate weld is often composed of multiple relatively independent straight line segments or polygonal line segments, and there are angle changes between each segment. Therefore, the correction process adopts a combination of segmented independent offset and global attitude fine-tuning. Specifically, the system first identifies the reference points of the weld seams on the four main edges of the stiffening plate, then calculates the translation vector and rotational fine-tuning amount of each edge relative to its theoretical position. These local transformation parameters are then applied sequentially to all trajectory points of the corresponding path segment. Finally, smooth interpolation is performed on the connection points of adjacent path segments to ensure that the welding torch does not produce significant speed changes or path jumps at corners. For example, in one possible implementation, assuming the cross plate workpiece is a cross plate of a power transmission tower, and the stiffening plate is a 20mm thick trapezoidal reinforcing plate, the welding task requires three layers of multi-pass filling for the four corner weld seams formed by the stiffening plate, the cross plate base plate, and the main leg plate, and corner welding is performed at each inner corner position. After receiving the attitude positioning signal, the robot control system immediately loads an independent trajectory program for the cross plate stiffening plate. The original program plan includes four weld seam paths, each path planned as three layers of weld seams, with each layer of weld seam width set to 7mm to 9mm. The system read the real-time coordinates of the stiffening plate obtained from the initial positioning and found that due to accumulated assembly errors, the first edge weld was offset by 3.2 mm in the X direction, and the second edge weld had a slight twist of about 1.8 degrees. The system first applied a uniform 3.2 mm X-direction offset to all trajectory points on the first edge, then applied a coordinate rotation transformation to the second edge to eliminate the twist angle. Subsequently, cubic spline interpolation was used to smooth the trajectory points in the corner area at the intersection of the two edges. After these corrections, the welding torch could sequentially complete each layer of filling along the actual weld centerline, and the path transition was natural at corners, avoiding localized overheating or incomplete penetration of the weld.

[0030] In another embodiment, when the cross plate contains multiple parallel stiffening plates, such as the multiple parallel stiffening plates inside the cross plate of a wind turbine tower, the independent trajectory program can be organized using a parameterized template method. The robot control system first automatically matches the appropriate template number based on the actual position of each stiffening plate returned by the positioning sensor, and then performs batch coordinate transformation on the path parameters in the template. After the transformation is completed, the system executes the corrected trajectory block by block according to the welding sequence from bottom to top and from inside to outside. This parameterized and batch correction method is particularly suitable for standardized cross plate components produced in batches, and can significantly shorten the programming and debugging cycle while ensuring accuracy. In step S43, during the multi-layer, multi-pass corner joint welding around the stiffening plate, the robot control system strictly follows the corrected trajectory to advance layer by layer. After each layer of weld is completed, the welding torch returns to the vicinity of the starting point, waits for the current and wire feeding system parameters to automatically switch to the next layer's set value before starting the next layer of welding. To prevent the performance degradation of the heat-affected zone due to excessively high interlayer temperatures, the system can insert a short natural cooling waiting time after each layer of weld is completed, or use an external air cooling device to force cooling of the weld area. After cooling, the welding torch is repositioned to the starting point of the next weld layer and continues to fill along the same path. In step S44, for the corner welds of the stiffening plate, a dedicated small-range closed path is pre-planned in the independent trajectory program. After reaching the weld end, the welding torch moves along an arc or near-arc trajectory according to the preset corner radius and angle range, typically covering an area of ​​8 mm to 15 mm on each side of the weld end, forming a reinforced transition weld. After the corner weld is completed, the system records the completion status of all layers of the edge weld and determines whether all surrounding welds have been completed. If not, the welding torch quickly moves to the starting point of the next unwelded edge weld and repeats the above layered welding and corner reinforcement process until all surrounding welds of the entire stiffening plate are completed. It should be noted that there are certain differences in the selection of motion control modes between the independent trajectory program for the stiffening plate corner weld and the main weld trajectory program. Because the stiffening plate weld path is short, has many turns, and the number of weld layers is relatively fixed, the program prioritizes accurate path reproduction and smooth, low-speed transitions at turns, rather than long-distance, high-speed movement. Therefore, numerous speed look-ahead and acceleration / deceleration control points are embedded during the path planning stage to ensure smooth attitude changes and speed transitions at each turn. In step S5, during the welding stage, the robot control system uses a continuous path motion control mode to drive the welding torch along the corrected trajectory. During the non-welding relocation stage, a point-to-point motion control mode is used to drive the welding torch to the next starting point. Specifically, the continuous path motion control mode means that when the welding torch is actually filling the weld, its trajectory consists of a large number of dense intermediate points. The system generates smooth straight lines or arc segments through an interpolation algorithm, ensuring that the welding torch speed, attitude, and welding process parameters are highly synchronized, thereby achieving a uniform weld formation.Point-to-point motion control mode is mainly used for idle movement between weld beads, weld seams, or layers. In this mode, the requirement for path smoothness is relatively low, and the system prioritizes rapid arrival and positioning accuracy to shorten non-production time. Furthermore, the specific implementation of continuous path motion control in the welding stage can be divided into the following steps.

[0031] Step S51: When the welding torch is near the starting point of a weld bead and the welding parameters are ready, the robot control system switches to continuous path mode. The system drives each joint axis in a coordinated manner according to the corrected trajectory point sequence, and the welding torch moves continuously along the weld centerline while maintaining its posture consistent with the weld normal direction. In straight sections, the system generates motion commands using linear interpolation; in areas with curvature changes, such as corners or bends, the system automatically inserts circular transition segments to ensure the welding torch does not produce sharp corner movements, thus avoiding undercut or incomplete fusion defects on the weld surface. Step S52: During continuous welding, the robot control system collects the deviation between the actual position of the welding torch and the theoretical trajectory points in real time. If the deviation exceeds a preset threshold of 0.8 mm obtained from experiments (e.g., a lateral deviation greater than 0.8 mm), the online correction function is triggered. Based on the current deviation direction and magnitude, the system dynamically superimposes a small compensation vector onto several subsequent trajectory points, gradually returning the welding torch to the correct path. This online correction mechanism continuously compensates for trajectory drift caused by thermal deformation, slight loosening of the fixture, or workpiece weight during welding, ensuring the overall forming quality of long or multi-layered welds. For example, in one possible implementation, suppose we are welding the bottom corner joint of a large stiffening plate on a cross-plate workpiece. The weld is approximately 1.8 meters long and requires four weld passes. During the first layer of welding, the robot control system drives the welding torch in a continuous path mode, starting from the left end of the weld and advancing along the correction trajectory at a constant speed of 8 mm / s, keeping the torch perpendicular to the base plate surface. When the welding reaches approximately 0.9 meters, due to localized thermal expansion of the cross-plate workpiece, the actual weld position shifts outward by about 1.1 mm. After detecting this deviation, the system gradually superimposes a reverse compensation vector at subsequent path points. After a transition distance of approximately 120 mm, the welding torch re-aligns with the weld centerline and continues welding the remaining section. The same control logic is repeated for subsequent welding layers two through four. Each interlayer shift switches to point-to-point mode, and the welding torch quickly returns to above the starting point at the maximum permissible speed. After cooling, it re-enters continuous path mode to begin filling the next layer. This control strategy, combining dynamic correction and mode switching, ensures a uniform weld appearance while keeping the welding cycle of a single weld within a reasonable range. In another embodiment, when the cross-plate workpiece contains multiple small stiffening plates and the weld is short, the point-to-point motion control during the non-welding shifting phase can be further optimized to improve the overall cycle time. After completing a short weld, the robot control system does not move directly to the starting point of the next weld in a straight line. Instead, it plans an obstacle-avoiding zigzag path to avoid welded areas, fixture protrusions, and other structural features that may cause interference. In point-to-point mode, the system allows the welding torch to operate at a high joint speed while gradually decelerating before approaching the target point, ensuring that the positioning repeatability accuracy is controlled within ±0.3 mm.This obstacle avoidance and polygonal displacement method is particularly effective in components with densely arranged stiffening plates, significantly reducing collision risks and shortening idle travel time. In step S6, throughout the entire welding cycle, the robot control system performs trajectory correction and motion control based on the workpiece coordinate system. Establishing the workpiece coordinate system is fundamental to achieving precise trajectory transformation and motion control.

[0032] Step S61: Before the welding task begins, the operator selects at least three non-collinear reference points on the cross-plate workpiece using a teach pendant. These are typically the centers of machined holes on the workpiece's base plate or the edges of precision-machined positioning surfaces. The robot control system records the actual posture and coordinate positions of the welding torch at these reference points and calculates the transformation matrix of the workpiece coordinate system relative to the robot's base coordinate system using a three-point coordinate system method. Subsequently, all real-time spatial coordinate data acquired from the positioning sensors and the path point coordinates in the preset trajectory program are first transformed to this workpiece coordinate system for processing and correction. Step S62: After the workpiece coordinate system is established, the system uniformly transforms all path points in the main seam trajectory program and the stiffening plate independent trajectory program to the workpiece coordinate system. During the transformation, the system simultaneously considers the actual positional offset of the tool center point. The precise definition of the tool center point, i.e., the spatial offset of the welding wire end relative to the welding torch flange center, is crucial for trajectory accuracy. In one implementation, the tool center point is calibrated by teaching the same point under multiple different postures: the welding torch is aligned with the same reference needle tip in different postures, and the system records the coordinate position of the flange center under each posture. The three-dimensional offset vector of the tool center point is calculated by least squares fitting. This offset vector is then applied to the forward and reverse transformations of all trajectory points to ensure that the welding wire tip can accurately reach the planned position. In step S63, if a large change in the welding torch posture or thermal expansion drift of the tool center point is found during trajectory execution, the robot control system can compensate for the deviation of the tool center point through online correction commands. For example, after completing several welds, the system automatically drives the welding torch to a fixed verification point, remeasures the actual deviation from the reference needle tip, and if the deviation exceeds a threshold, the tool center point compensation value is updated and superimposed on all subsequent trajectory points until the next verification. This periodic online verification and compensation mechanism can effectively suppress the cumulative error caused by long-term continuous welding and maintain the long-term stability of the weld position. It should be noted that the workpiece coordinate system-based trajectory management method allows the same welding program to adapt to different batches and clamping states of the same type of workpiece. As long as the three reference points are quickly taught after each workpiece is clamped, the system can automatically establish the current workpiece coordinate system and seamlessly map all pre-stored trajectory programs to the current actual position. This coordinate system independence design significantly reduces programming workload and improves the flexibility and consistency of mass production. Step S7, the robot control system's correction and optimization of the preset welding trajectory program also includes editing the trajectory point coordinates and path sequence through the programming interface, and smoothly optimizing the connection of adjacent weld trajectories.

[0033] Step S71: After trajectory correction is completed, the operator can view the generated execution trajectory through the graphical programming interface of the robot teach pendant, and manually fine-tune the spatial coordinates or attitude angles of individual key points if necessary. For example, if the transition of a certain corner path is not smooth enough, the operator can select the path segment on the interface, insert additional intermediate transition points, and adjust the speed and acceleration constraints of each point to make the movement of the welding torch in the corner area smoother. Step S72: For the connection path between the main seam and the stiffening plate weld, and between different stiffening plate welds, the system provides a path optimization function. The operator can specify the end point of several adjacent welds and the starting point of the next weld, and the system automatically plans a smooth idle path to minimize sharp turns and speed changes. The optimized idle path ensures safety and shortens non-welding time, thereby improving the overall welding cycle time. For example, in one possible implementation, when the cross plate workpiece contains multiple parallel stiffening plates, the distance between the end point and the starting point of the weld of two adjacent stiffening plates is relatively small, but there is a height difference. The operator selects the connection area between the two weld seams through the programming interface, sets the maximum allowable speed to 400 mm / s and the maximum acceleration to 180 mm / s², and the system then generates a transition path with an S-shaped acceleration and deceleration curve. After the welding torch completes the last weld seam of the first stiffening plate, it quickly rises and moves laterally along a smooth curve to the starting point of the first weld seam of the second stiffening plate. The entire movement process takes less than 4 seconds, and the posture change is continuous, avoiding mechanical impact and welding torch vibration. This smooth and optimized connection method can save a significant amount of non-productive time and improve equipment utilization in the mass welding production of multi-stiffening plate components. In another embodiment, if the workpiece structure allows, the operator can also predefine several shared transition path templates through the programming interface, such as a fixed rise-lateral movement-descent path from the main seam area back to the stiffening plate area. When the system reaches the end of the main seam, it can directly call this template path to quickly move to the starting area of ​​the stiffening plate welding, reducing the computational burden of replanning the idle path each time and further improving the production cycle.

[0034] If the technical solution of this application involves the collection, processing, or application of personal information, the relevant products have, before implementing any personal information processing activities, fully and clearly informed individuals of the processing rules in accordance with the "Personal Information Protection Law of the People's Republic of China" and other current laws and regulations, and obtained their voluntary and explicit consent. If sensitive personal information is involved, the product has obtained the individual's separate consent before processing, and such consent is given in an explicit manner. For example, prominent signs are set up in the area where information collection devices such as cameras are located, clearly indicating "Entering is considered as consent to the collection of personal information"; or through pop-ups, checkboxes, user-initiated uploads, etc., under the premise of clearly listing the processor's identity, processing purpose, processing method, and information type, the user actively completes the authorization operation. The above mechanisms ensure that all personal information processing activities are based on legal authorization and fully comply with national compliance requirements regarding personal information protection.

[0035] The above-disclosed embodiments are merely preferred embodiments of the present invention and should not be construed as limiting the scope of the invention. Those skilled in the art will understand that implementing all or part of the above-described embodiments and making equivalent changes in accordance with the claims of the present invention are still within the scope of the invention.

Claims

1. A method for welding a corner joint of a cross plate, characterized in that, include: The position of the main seam and stiffening plate corner weld on the cross-plate workpiece is detected by a positioning sensor to obtain real-time spatial coordinate data of the weld. The preset welding trajectory program is corrected based on the real-time spatial coordinate data to generate an execution trajectory that matches the actual weld position. The execution trajectory is executed according to the preset welding sequence to complete the multi-layer multi-pass welding of the main seam and the corner welding, and a posture switching signal is sent to the positioner controller. After receiving the posture switching signal, the positioner controller drives the cross-plate workpiece to rotate or flip to the next weld welding posture and feeds back the posture arrival signal. After receiving the posture arrival signal, the independent trajectory program of the stiffening plate corner weld is started to execute multi-layer multi-pass corner welding and corner welding around the stiffening plate. During the welding stage, the continuous path motion control mode is used to drive the welding torch to move along the corrected trajectory. During the non-welding displacement stage, the point-to-point motion control mode is used to drive the welding torch to move to the next starting point.

2. The welding method for the corner joint of the cross plate as described in claim 1, characterized in that, The step of detecting the actual position of the main seam and stiffening plate corner weld on the cross-plate workpiece using a positioning sensor includes: the positioning sensor identifying the starting position and edge position of the main seam and stiffening plate corner weld, and obtaining real-time spatial coordinate data of the main seam and stiffening plate corner weld; comparing the real-time spatial coordinate data with the coordinates of the trajectory points in the preset welding trajectory program to determine the coordinate offset; and adjusting all trajectory points of the preset welding trajectory program as a whole according to the coordinate offset to generate the execution trajectory that matches the actual weld position.

3. The welding method for the corner joint of the cross plate as described in claim 1, characterized in that, The execution of the execution trajectory according to the preset welding sequence includes: calling the stored main seam trajectory program, completing the multi-layer multi-pass welding and the corner welding in the order of the corner weld between the sub-plates to the corner weld between the sub-plate and the bottom plate; after all the main seam welding is completed, generating the attitude switching signal and sending it to the positioner controller.

4. The welding method for the corner joint of the cross plate as described in claim 1, characterized in that, After receiving the attitude switching signal, the positioner controller drives the cross plate workpiece to rotate or flip to the next weld welding attitude, including: the positioner controller controls the servo system to drive the cross plate workpiece to flip or rotate to a preset angle according to the attitude switching signal, so that the stiffening plate corner weld is in the reachable position of the welding torch; the positioner controller generates the attitude arrival signal after monitoring the workpiece attitude; after receiving the attitude arrival signal and confirming that the workpiece attitude is stable, the independent trajectory program of the stiffening plate corner weld is started; the independent trajectory program includes multiple layers and multiple paths and corner wrapping paths for the corner welds around the stiffening plate, and adjacent paths are connected by a smooth transition.

5. The welding method for the corner joint of the cross plate as described in claim 1, characterized in that, The step of starting an independent trajectory program for the stiffening plate fillet weld after receiving the attitude positioning signal includes: manually completing the stiffening plate point splicing and positioning, loading the pre-stored stiffening plate trajectory program; and applying the previously acquired real-time spatial coordinate data to correct the stiffening plate trajectory program.

6. The welding method for the corner joint of a cross plate as described in claim 1, characterized in that, The method of using a continuous path motion control mode to drive the welding torch to move along the correction trajectory during the welding stage, and using a point-to-point motion control mode to drive the welding torch to move to the next starting point during the non-welding displacement stage, includes: the continuous path motion control mode generating a straight line or arc trajectory to ensure that the welding torch moves continuously along the correction trajectory to complete the corner weld; and the point-to-point motion control mode determining the coordinates of the next starting point and driving the welding torch to move rapidly during the non-welding displacement stage.

7. The welding method for the corner joint of a cross plate as described in claim 1, characterized in that, The method of using continuous path motion control mode to drive the welding torch to move along the correction trajectory during the welding stage includes: establishing a workpiece coordinate system by teaching at least three reference points on the cross-plate workpiece; converting all the execution trajectories into coordinates under the workpiece coordinate system; defining the tool center point before trajectory execution, calculating the tool center point value by teaching multiple reference points and applying it to the trajectory coordinate transformation; and compensating for the tool center point deviation through online correction instructions during trajectory execution.

8. The welding method for the corner joint of a cross plate as described in claim 1, characterized in that, The correction of the preset welding trajectory program includes: editing the trajectory point coordinates and path sequence of the preset welding trajectory program through a programming interface, and smoothly optimizing the connection of adjacent weld trajectories.

Images