Servo rotating spot welding device

Abstract

This invention relates to welding apparatus for automotive parts, and more particularly to a servo rotary point-to-point welding apparatus, comprising: a clamping assembly for clamping sheet metal parts, the clamping assembly being mounted on a mold plate; a rotary drive assembly mounted on a frame and connected to the end of the mold plate via a connector; a rotary monitoring assembly mounted on the frame and connected to the end of the mold plate and opposite to the rotary drive assembly via a connector; and a welding assembly located on the side wall of the clamping assembly and mounted on a horizontal moving assembly. This invention not only improves the welding efficiency of large-area irregularly shaped curved sheet metal parts but also enhances welding quality.

Description

Technical Field

[0001] This invention relates to welding equipment for automotive parts, and more particularly to a servo rotary point welding device. Background Technology

[0002] In the automotive parts manufacturing industry, large-area irregular curved sheet metal parts are an indispensable and important component of the car body structure. These sheet metal parts usually have complex shapes and sizes, and their manufacturing process has a crucial impact on the overall performance and appearance quality of the car.

[0003] Currently, the manufacturing of large-area irregular curved sheet metal parts typically involves first forming them using a stamping process. Stamping can process metal sheets into the required shape, but the stamped sheet metal parts often require further assembly and connection to meet the integrity and functional requirements of automotive components. In existing technologies, welding of large-area sheet metal parts mainly relies on manual operation. Operators need to clamp and fix the sheet metal parts on specific tooling to ensure their stability during welding. Then, multiple sheet metal parts are welded into a single unit, or other components are welded onto the sheet metal parts, using manual welding. Because the welding process requires precise control of welding parameters and welding paths, and the workload of welding large-area sheet metal parts is substantial, the entire welding process is time-consuming. Furthermore, the stability and consistency of manual welding are difficult to guarantee, easily leading to welding defects such as uneven welds and incomplete welds. This not only affects the welding quality but may also increase subsequent rework costs.

[0004] Therefore, those skilled in the art are dedicated to developing a servo-driven rotary point welding device that not only improves the welding efficiency of large-area irregular curved sheet metal parts, but also enhances the welding quality. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to provide a servo rotary point welding device, which not only improves the welding efficiency of large-area irregular curved sheet metal parts, but also improves the welding quality.

[0006] The technical solution of the present invention to solve the above-mentioned technical problems is as follows: a servo rotary point welding device, comprising... A clamping assembly for clamping sheet metal parts, the clamping assembly being mounted on a mold plate; A rotary drive assembly, which is mounted on a frame and connected to the end of the mold plate via a connector; A rotation monitoring component is mounted on the frame and connected to the end of the mold plate and the side opposite to the rotation drive component via a connector. A welding assembly located on the side wall of the clamping assembly and mounted on the horizontal moving assembly.

[0007] The beneficial effects of adopting the above solution are: by driving the mold plate to rotate the sheet metal parts through the rotary drive component, and cooperating with the lateral movement of the welding component on the horizontal moving component, the welding gun can be automatically aligned with the normal direction of any point on the curved surface, avoiding the limitation of the welding gun posture caused by the curved surface occlusion in the fixed station welding, and improving the welding efficiency of the curved surface. The rotation monitoring component and the drive component are located at both ends of the mold plate, respectively, to realize real-time closed-loop monitoring of the rotation angle, ensuring precise control of the deflection angle of the sheet metal parts during the welding process, thereby ensuring the consistency of electrode pressure and welding parameters at each weld point.

[0008] Based on the above technical solution, the present invention can be further improved as follows.

[0009] Furthermore, the rotary drive assembly includes a drive shaft, one end of which is connected to a connector. The drive shaft is mounted on the frame via a shaft mounting base. A worm gear is connected to the end of the drive shaft. The worm gear is installed inside the housing. A servo power assembly is also installed on the side wall of the housing. The output end of the servo power assembly is connected to a worm that cooperates with the worm gear.

[0010] The beneficial effects of adopting the above-mentioned further solution are: the worm gear can effectively resist the mold plate deflection caused by the welding reaction force during the welding process, ensuring the welding position is locked, and at the same time, the closed-loop control characteristics of the servo motor can achieve high-precision control of the rotation angle, meeting the requirements of repeated positioning accuracy for dense point welding.

[0011] Furthermore, the rotation monitoring component includes a passive shaft, the end of which is connected to a connector. The passive shaft is mounted on the frame via a shaft mounting base. An encoder disk is connected to the end of the passive shaft. The encoder disk is mounted on a mounting housing. An encoder reading head that mates with the encoder disk is also installed inside the mounting housing.

[0012] The beneficial effect of adopting the above-mentioned further solution is that, through the cooperation of the encoder disk and the reading head, a fully closed-loop position monitoring system is formed, which can provide real-time feedback on the absolute rotation angle of the mold plate and provide accurate position information for the control system.

[0013] Furthermore, a limiting plate is connected to the side wall of the mold plate, and the limiting plate has multiple spaced limiting cavities. A limiting cylinder is installed on the frame, and a limiting rod that cooperates with the limiting cavity is connected to the output end of the limiting cylinder. The limiting rod is installed on the frame through a guide seat. A balance block is installed on the side wall of the mold plate opposite to the clamping assembly.

[0014] The beneficial effects of adopting the above-mentioned further solution are: by cooperating with the limit rod driven by the limit cylinder and the different limit cavities on the limit plate, the sheet metal parts can be mechanically and rigidly locked at a specific welding angle, and can still maintain the current posture when the machine tool is powered off or the servo system fails, thereby improving the safety redundancy of the equipment. At the same time, the limiting rod and the limiting cavity work together to lock the mold plate, which is convenient for placing sheet metal automotive parts or picking up and removing welded parts.

[0015] Furthermore, the clamping assembly includes a support block and a clamping cylinder, the upper end of the support block abuts against the lower end face of the sheet metal part, and the support block has an air cavity; The support block is connected to an air inlet pipe and an air outlet pipe, both of which are connected to the air cavity. A solenoid valve is installed on the air inlet pipe, and the air outlet pipe is also connected to the gas output end of the clamping cylinder. A clamping block is installed on the output end of the clamping cylinder.

[0016] The beneficial effects of adopting the above-mentioned further solution are: the upper end of the support abuts against the lower end face of the sheet metal part; after compressed air enters the air cavity and the cylinder air intake from the air inlet pipe, it pushes the clamping block to clamp the sheet metal part and then the solenoid valve closes; the support is made of a high thermal conductivity material; during the welding process, the heat generated by welding is conducted through the support and heats the compressed gas in the air cavity, which increases the pressure in the air cavity, making the clamping cylinder more powerful and the clamping force of the clamping block greater, thus reducing the deformation of the sheet metal part during the welding process; The greater the regional welding strength, the more heat is generated, and the greater the clamping force generated by the regional clamping block. By adopting adaptive active intervention in the deformation process, a stronger and reverse correction force is applied to correct the shape in real time under hot conditions, ensuring the final cold-state dimensions, stabilizing the butt joint gap, ensuring the penetration depth and weld formation, and optimizing the distribution of residual stress through intelligent restraint. After welding, the rotary drive component rotates at high speed, which not only brushes off the welding slag produced by welding on the mold plate to reduce its adhesion to the mold, but also enhances air convection to cool the sheet metal parts. This ensures that the relative speed between every point on the circumference of the workpiece and the air is the same, providing a uniform and forced convection cooling environment for the entire workpiece. The heat in the weld area is carried away more evenly and symmetrically, the temperature gradient of the entire workpiece becomes smooth, and the cooling shrinkage is more uniform, thereby effectively reducing the residual stress peak, significantly preventing the cylinder from deforming such as ellipticity, local depressions or bending, accelerating the diffusion of heat into the environment, and preventing secondary overheating in local areas. High-speed rotation allows the weld to always be in the most favorable flat or spherical welding position to complete solidification and initial cooling, which not only achieves the best appearance but also physically ensures the stability of the weld shape.

[0017] Furthermore, the welding assembly includes a welding robot mounted on the horizontal moving assembly, and the end effector of the welding robot is also integrated with a vision inspection component; The visual inspection component is used to extract weld feature point sets based on cloud data. The weld trajectory curve L(t) is fitted using the least squares method, and the thickness of the sheet metal part is calculated based on the point cloud density distribution. ; Where δ is the estimated thickness of the sheet metal part, in millimeters; n is the total number of feature points in the point cloud data, which is a positive integer greater than 1; p i Let p be the three-dimensional coordinate vector of the i-th feature point. i =(x i ,y i ,z i ); L(t i ) represents the weld trajectory curve in parameter t i The coordinates of the point; t i For parametric coordinates, the value range is [0,1], representing the relative position of a point on the curve; Given the Euclidean norm, calculate the straight-line distance between two points; ∑ is the summation symbol, which sums up all terms from i to n; The visual inspection component calculates the welding current I and welding speed V based on the thickness of the sheet metal part. ; ; in, This is the welding current output value, in amperes. The ratio of current to thickness; δ is the estimated thickness of the sheet metal part, in millimeters; For current reference parameters; Welding speed, measured in millimeters per second; This is the proportional coefficient for speed adjustment; These are the speed reference parameters.

[0018] The beneficial effects of adopting the above-mentioned further solution are: by integrating visual inspection components and using point cloud data processing algorithms, the local thickness of irregular curved sheet metal parts can be accurately estimated online, and the optimal welding current and welding speed can be calculated in real time based on the thickness, thereby reducing the phenomenon of burn-through or insufficient penetration caused by uneven thickness due to curved surface stretching when using fixed parameters for welding. By using the thickness estimate as a feedforward control parameter, the welding current is proportional to the workpiece thickness and the welding speed is inversely proportional to the thickness, ensuring that a consistent weld nugget diameter and penetration depth can be obtained in different thickness regions, significantly reducing the welding defect rate caused by differences in human experience.

[0019] Furthermore, the vision inspection component is electrically connected to the control component, and the control component integrates a deformation compensation module. The deformation compensation module compares the deviation Δd = ||q(t) - q0(t)|| in real time with the actual weld trajectory q(t). When Δd > 0.5 mm, the following compensation mechanism is triggered: The deflection angle of the mold plate is adjusted by the rotary drive component: θ = arcsin(Δd / R); Adjust the end effector posture of the welding robot to keep the welding torch perpendicular to the tangent of the weld seam at all times; Based on the heat deformation prediction model, the welding path is corrected in advance, and the formula for calculating the heat deformation is:

[0020] Where R is the radius of rotation; This refers to the temperature change caused by the welding thermal cycle. The coefficient of thermal expansion of the material; This refers to the residence time of the electric arc per unit length. The density of the material; c represents the specific heat capacity of the material.

[0021] The beneficial effects of adopting the above-mentioned further solution are: by comparing the deviation between the actual weld trajectory and the theoretical trajectory in real time through the deformation compensation module, when the deviation exceeds the threshold, the rotation drive component actively adjusts the deflection angle of the mold plate and simultaneously adjusts the robot end posture to realize online correction of welding deformation and avoid subsequent weld point misalignment caused by cumulative deformation. Based on the thermal deformation prediction model, the temperature change caused by the welding thermal cycle is estimated in advance. The pre-correction amount is introduced in the welding path planning stage to realize a dual compensation mechanism, which has a significant inhibitory effect on the problem of large thin sheet metal parts being prone to heat warping.

[0022] Furthermore, the visual inspection component also integrates a molten pool state recognition module and a weld tracking module. The molten pool state recognition module uses a five-dimensional joint iterative spatial segmentation method to extract features of the arc region, the main molten pool region, and the tail region of the molten pool in one go based on the molten pool image acquired by the visual inspection component. The molten pool state recognition module establishes a mapping relationship between welding parameters and molten penetration state based on a CNN-LSTM hybrid neural network. When an incomplete molten penetration state is detected, the system increases the welding current at a rate of 3A / s to 7A / s and decreases the welding speed at a rate of 0.003m / s² to 0.007m / s². The weld seam tracking module uses a robust controller setting, and its weighting function is selected as follows: ; ; in, Let s be the performance weight transfer function, and s be the complex frequency variable; For complex variables in the Laplace transform; The peak value of the sensitivity function is typically set to 1.5-2.0. For bandwidth frequency; This is the steady-state error; To control the weight transfer function; To control bandwidth frequency.

[0023] The beneficial effects of adopting the above-mentioned further scheme are: by using a robust controller and performance weights and control weights, the uncertainty caused by workpiece surface reflection, spatter interference, etc. is effectively suppressed, ensuring the stability and control accuracy of weld seam tracking under complex working conditions, and improving the anti-interference ability of the system.

[0024] Furthermore, the control component integrates a sensor data fusion module, which performs spatiotemporal registration of the molten pool image, weld point cloud data, current, and voltage signals, and calculates the confidence assignment function for each sensor.

[0025] in, The basic probability assignment value for proposition A represents the degree of confidence in A; B and C represent different evidentiary propositions or assumptions; B∩C=A means that the intersection of propositions B and C equals A; For the first sensor, assign a basic probability to proposition B; For the second sensor, assign a basic probability to proposition C; It is a conflict factor.

[0026] The beneficial effects of adopting the above-mentioned further scheme are: based on the joint probability allocation of multi-source evidence, the system can automatically reduce the weight of a certain sensor signal and rely on other sensors to make decisions when the signal of a certain sensor is abnormal, thereby avoiding welding quality accidents caused by single-point failures and significantly improving the reliability of the equipment and continuous production capacity. Attached Figure Description

[0027] Figure 1 This is a schematic diagram of the structure of a servo-rotating point welding device according to a specific embodiment of the present invention; Figure 2 This is a schematic diagram of the limiting plate and limiting rod structure according to a specific embodiment of the present invention; Figure 3 This is a schematic diagram of the clamping component structure according to a specific embodiment of the present invention.

[0028] The attached diagram lists the components represented by each number as follows: 1. Clamping assembly; 2. Mold plate; 3. Rotary drive assembly; 4. Frame; 5. Rotary monitoring assembly; 6. Welding assembly; 7. Horizontal movement assembly; 8. Drive shaft; 9. Shaft mounting base; 10. Housing; 11. Servo power assembly; 12. Passive shaft; 13. Limiting plate; 14. Limiting cavity; 15. Limiting rod; 16. Limiting cylinder; 17. Guide seat; 18. Balance block; 19. Support block; 20. Clamping cylinder; 21. Air cavity; 22. Air inlet pipe; 23. Air outlet pipe; 24. Solenoid valve; 25. Welding robot; 26. Vision inspection assembly. Detailed Implementation

[0029] The principles and features of the present invention are described below with reference to the accompanying drawings. The examples given are only for explaining the present invention and are not intended to limit the scope of the present invention.

[0030] In the description of this invention, it should be understood that the terms "center," "length," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "inner," "outer," "circumferential," and "circumferential" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing this invention and simplifying the description, and are not intended to indicate or imply that the system or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0031] In the description of this invention, "a plurality of" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0032] In this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," "linking," and "fixing," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components, unless otherwise explicitly limited. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0033] like Figure 1 , Figure 2 and Figure 3 The servo rotary point welding device shown includes Clamping assembly 1 is used to clamp the sheet metal parts to ensure that the sheet metal parts do not shift or deform during the welding process. Clamping assembly 1 is installed on mold plate 2. Rotary drive assembly 3 is mounted on frame 4 and connected to the end of mold plate 2 via connector. The precise torque output by rotary drive assembly 3 drives mold plate 2 and sheet metal parts to perform controllable rotational motion. Rotation monitoring component 5 is mounted on frame 4 and is connected to the end of mold plate 2 and the opposite side of rotation drive component 3 via connector. It is coaxially symmetrically arranged with rotation drive component 3 to realize real-time feedback of the rotation angle of mold plate 2, forming a closed-loop control loop to ensure rotation accuracy. Welding component 6 is located on the side wall of clamping component 1 and mounted on horizontal moving component 7. Horizontal moving component 7 adopts a high-precision linear module structure, which can drive welding component 6 to make precise lateral movement. In conjunction with the rotation of mold plate 2, it can achieve full coverage welding of all points of irregular curved sheet metal parts.

[0034] like Figure 1As shown, in some embodiments, the rotary drive assembly 3 includes a drive shaft 8, which is made of high-strength alloy material and has sufficient torsional strength to resist welding reaction force. One end of the drive shaft 8 is connected to a connector, and the drive shaft 8 is mounted on the frame 4 via a shaft mounting base 9. A worm gear is connected to the end of the drive shaft 8, and the worm gear and the drive shaft 8 are connected by a flat key to achieve torque transmission. The worm gear is installed inside the housing 10, which can effectively block welding spatter, dust and other impurities from entering the transmission pair, avoiding affecting the meshing accuracy and service life of the worm gear. A servo power assembly 11 is also installed on the side wall of the housing 10. The servo power assembly 11 is an integrated structure of a servo motor and a reducer. The reducer can reduce the output speed and increase the output torque to meet the power requirements of welding operations. The output end of the servo power assembly 11 is connected to a worm gear that cooperates with the worm gear. The worm gear transmission pair has a self-locking characteristic, which can achieve position locking during welding to prevent the mold plate 2 from deflecting due to welding reaction force. At the same time, the servo motor has a closed-loop control function, which can accurately adjust the rotation angle according to the control command to achieve high-precision control of rotational motion.

[0035] The rotation monitoring component 5 includes a passive shaft 12, which is coaxially arranged with the active shaft 8 to ensure the accuracy of rotation angle monitoring. The end of the passive shaft 12 is connected to a connector and is mounted on the frame 4 via a shaft mounting base. The shaft mounting base has the same structure as the corresponding shaft mounting base of the active shaft 8 to ensure the uniformity of the installation reference. An encoder disk is connected to the end of the passive shaft 12. The encoder disk rotates synchronously with the passive shaft 12 and can convert mechanical rotational motion into optical signals. The encoder disk is mounted on a mounting housing with a sealing and protective function to prevent external environment from interfering with the disk and reading components. An encoder reading head that cooperates with the encoder disk is also installed inside the mounting housing. The encoder reading head maintains a preset gap with the disk and can collect the rotation signal of the disk in real time and convert it into an electrical signal to be transmitted to the control component, providing accurate feedback for the angle adjustment of the rotation drive component 3 and realizing full closed-loop monitoring and control of the rotation angle.

[0036] like Figure 1 and Figure 2As shown, in another embodiment, a limiting plate 13 is connected to the side wall of the mold plate 2. The limiting plate 13 is disc-shaped and has multiple spaced limiting cavities 14. A limiting cylinder 16 is installed on the frame 4. The output end of the limiting cylinder 16 is connected to a limiting rod 15 that cooperates with the limiting cavity 14, which can achieve a tight fit with the limiting cavity 14 and improve locking stability. The limiting rod 15 is installed on the frame 4 through a guide seat 17. The guide seat 17 can ensure that the limiting rod 15 moves in a straight line and avoids its deviation from affecting the fitting accuracy. A balance block 18 is installed on the side wall of the mold plate 2 and on the side opposite to the clamping assembly 1. The weight and installation position of the balance block 18 can be adjusted according to the center of gravity distribution of the sheet metal part to be welded. It is used to balance the overall center of gravity of the mold plate 2 and the sheet metal part, avoid the off-center load caused by the center of gravity shift during rotation, which would lead to increased wear of the shaft system and a decrease in rotation accuracy, and ensure the smoothness of the rotation of the mold plate 2.

[0037] like Figure 1 and Figure 3 As shown, in this embodiment, the clamping assembly 1 includes a support block 19 and a clamping cylinder 20. The support block 19 is made of copper alloy material with high thermal conductivity, which can quickly conduct the heat generated during the welding process. The upper end of the support block 19 abuts against the lower end face of the sheet metal part, and its upper end face structure is adapted to the curved contour of the sheet metal part to achieve surface contact support to improve support stability. The support block 19 has an air cavity 21, which is a sealed structure that can accommodate compressed air and realize the function of thermal expansion and pressurization. The support block 19 is connected to an air inlet pipe 22 and an air outlet pipe 23. The air inlet pipe 22 is used to connect to an industrial compressed air source, and the air outlet pipe 23 connects the air chamber 21 and the clamping cylinder 20. Both the air inlet pipe 22 and the air outlet pipe 23 are connected to the air chamber 21. A solenoid valve 24 is installed on the air inlet pipe 22. The solenoid valve 24 is controlled by the control component to achieve precise control of the air path. The air outlet pipe 23 is also connected to the gas output end of the clamping cylinder 20 to form a linkage air path structure. A clamping block is installed on the output end of the clamping cylinder 20. The lower end face of the clamping block is adapted to the upper end face of the sheet metal part to achieve the clamping and fixing of the sheet metal part. During operation, the solenoid valve 24 is opened, and compressed air enters the air chamber 21 through the air inlet pipe 22 and simultaneously enters the clamping cylinder 20 through the air outlet pipe 23, pushing the clamping block downward to press the sheet metal part onto the support block 19. Subsequently, the solenoid valve 24 closes the sealed air passage. During the welding process, the solenoid valve is closed, and heat is conducted through the support block 19 to the compressed air in the air chamber 21, causing the air to expand due to heat and the air pressure to increase, thereby increasing the output force of the clamping cylinder 20, realizing adaptive adjustment of the clamping force, and effectively suppressing the welding deformation of the sheet metal part.

[0038] In this embodiment, the welding assembly 6 includes a welding robot 25, which may be a six-axis welding robot. The welding robot 25 is mounted on a horizontal moving assembly 7, which is driven by a servo motor and can achieve high-precision lateral movement of the welding robot 25. It works in coordination with the rotational movement of the mold plate 2 to cover all welding points on the curved surface of the sheet metal part. The welding robot 25 also integrates a vision inspection assembly 26 at its end. The vision inspection assembly 26 is a 3D vision camera with high-precision point cloud data acquisition capability, which can quickly capture the three-dimensional features of the weld area. Visual inspection component 26 is used to extract weld feature point sets based on cloud data. The weld trajectory curve L(t) is fitted using the least squares method. The least squares method can effectively reduce error interference in point cloud data and improve trajectory fitting accuracy. Then, the thickness of the sheet metal part is calculated based on the point cloud density distribution. ; Where δ is the estimated thickness of the sheet metal part, in millimeters, and the numerical value can reflect the local thickness difference of the sheet metal part. n is the total number of feature points in the point cloud data, which is a positive integer greater than 1. The number of feature points is set according to the size of the weld area and the accuracy requirements. The more feature points there are, the higher the calculation accuracy. p i Let p be the three-dimensional coordinate vector of the i-th feature point. i =(x i ,y i ,z i The data is directly acquired by the visual inspection component 26. L(t i ) represents the weld trajectory curve in parameter t i The coordinates of the points at each location form a one-to-one correspondence with the feature points; t i For parametric coordinates, the value range is [0,1], representing the relative position of a point on the curve; Using the Euclidean norm, the straight-line distance between two points is calculated, and the mean of the distance is used to accurately estimate the thickness. ∑ is the summation symbol, which accumulates all terms from i to n, integrating the deviation information of all feature points; The vision inspection component calculates the welding current I and welding speed V based on the thickness of the sheet metal part; ; ; in, This is the welding current output value, in amperes. The ratio of current to thickness; δ is the estimated thickness of the sheet metal part, in millimeters; For current reference parameters; Welding speed, measured in millimeters per second; This is the proportional coefficient for speed adjustment; This parameter serves as a speed reference parameter to ensure the stability of the welding process. By adjusting the logic of this parameter, adaptive optimization of welding parameters in areas of different thicknesses can be achieved, avoiding problems such as burn-through or insufficient penetration.

[0039] In one embodiment, the vision inspection component 26 is electrically connected to the control component. The control component is an integrated structure of PLC and industrial computer, possessing data processing, command sending, and logic control functions. The control component integrates a deformation compensation module, which is used to address the thermal deformation of sheet metal parts during welding. The deformation compensation module compares the deviation Δd = ||q(t) - q0(t)|| in real time between the actual weld trajectory q(t) and the theoretical trajectory q0(t). When Δd > 0.5 mm, the deformation has affected the welding accuracy, triggering the following compensation mechanism: The deflection angle θ = arcsin(Δd / R) of the mold plate is adjusted by the rotation drive component. This is the distance from the rotation center of the mold plate 2 to the weld. This formula is derived based on geometric relationships and can achieve precise correction of deviations. Adjust the end effector posture of the welding robot to keep the welding torch perpendicular to the tangent of the weld seam, ensuring the rationality of the welding torch posture and avoiding weld defects caused by posture deviation. Based on the heat deformation prediction model, the welding path is corrected in advance, and the formula for calculating the heat deformation is:

[0040] Where R is the radius of rotation; This refers to the temperature change caused by the welding thermal cycle. The coefficient of thermal expansion of the material; This refers to the residence time of the electric arc per unit length. The density of the material; c represents the specific heat capacity of the material.

[0041] This formula allows for the estimation of thermal deformation in advance, enabling the introduction of pre-correction amounts in path planning, thus allowing for early intervention in deformation and ensuring welding accuracy.

[0042] In another embodiment, the visual inspection component 26 also integrates a molten pool state recognition module and a weld tracking module. The molten pool state recognition module is used to monitor the welding penetration quality in real time. Based on the molten pool image acquired by the visual inspection component, it uses a five-dimensional joint iterative spatial segmentation method to extract the features of the arc region, the main molten pool region, and the tail region of the molten pool in one go. The five-dimensional joint iterative spatial segmentation method can segment the image from five dimensions: brightness, color, texture, shape, and dynamic changes, effectively distinguishing different regions and improving the accuracy and efficiency of feature extraction. The molten pool state recognition module establishes a mapping relationship between welding parameters and penetration state based on a CNN-LSTM hybrid neural network. When an incomplete penetration state is detected, the system increases the welding current at a rate of 3A / s to 7A / s and decreases the welding speed at a rate of 0.003m / s² to 0.007m / s². The adjustment rate has been verified by the process and can quickly correct the penetration state while avoiding the aggravation of welding defects, thus ensuring weld quality. The weld seam tracking module uses a robust controller setting, and its weighting function is selected as follows: ; ; in, Let s be the performance weight transfer function, and s be the complex frequency variable; For complex variables in the Laplace transform; The peak value of the sensitivity function is typically set to 1.5-2.0. For bandwidth frequency; This is the steady-state error; To control the weight transfer function; To control bandwidth frequency.

[0043] This weighting function design enables the weld seam tracking module to maintain stable tracking accuracy even under complex working conditions such as workpiece surface reflection and welding spatter interference.

[0044] In this embodiment, the control component integrates a sensor data fusion module. This module integrates information from multiple sensor sources to improve the reliability of system decisions. It performs spatiotemporal registration on the molten pool image and weld point cloud data acquired by the visual inspection component, as well as the current and voltage signals acquired by the welding system. Spatiotemporal registration eliminates deviations in acquisition time and spatial location between different sensors, ensuring data consistency. Subsequently, the reliability allocation function for each sensor is calculated. The function formula is as follows:

[0045] in, The basic probability assignment value for proposition A represents the degree of confidence in A. Proposition A is the judgment result of the welding state, such as "normal welding" or "incomplete penetration". B and C represent different evidentiary propositions or hypotheses, corresponding to the detection results of different sensors. • B∩C=A means that the intersection of propositions B and C equals A; For the first sensor, assign a basic probability to proposition B; The basic probability assignment for proposition C by the second sensor is preset based on the sensor's detection accuracy and stability. The conflict factor is used to measure the degree of conflict between the detection results of different sensors. When the value of k is large, it indicates that there is a conflict in the sensor data. At this time, the system automatically reduces the sensor weight corresponding to the conflicting data and relies on the sensor data with higher reliability to make decisions, thereby avoiding welding quality accidents caused by single-point sensor failures and improving the continuous working capability and reliability of the device. Example

[0046] This embodiment is for automotive roof sheet metal parts (material is cold-rolled steel sheet DC01, thickness range 0.8 mm - 1.2 mm, curved surface is gentle arc, 32 welding points need to be completed).

[0047] Visual inspection component 26 extracts weld feature point set P={p1, p2, …, p} based on cloud data. 800 (n=800, determined based on the length and precision requirements of the roof weld), the weld trajectory curve L(t) is fitted using the least squares method to reduce the interference of point cloud data errors, and then the thickness of the sheet metal part is calculated. The formula is as follows: δ=1 / n∑_(i=1)^n‖p_i-L(t_i)‖, Where pi is the three-dimensional coordinate vector of the i-th feature point (e.g., p1=(100.3, 50.2, 20.1)mm); p2=(100.4, 50.3, 20.0)mm……), ti is the parametric coordinate (value range [0,1], evenly distributed according to the feature point order); L(ti) represents the coordinates of the weld trajectory curve at point ti; ‖pi-L(ti)‖ is the Euclidean distance between two points. The estimated thickness of the roof sheet metal in this area is calculated to be δ=1.0mm.

[0048] The vision inspection component calculates the welding current I and welding speed V based on δ, using the following formulas: I=k1*δ+b1、 V = k² / δ + b² Where k1 = 40A / mm (the proportionality coefficient of DC01 welding current to thickness of cold-rolled steel sheet, determined through process verification). b1=20A (current reference parameter); k2 = 3.0 mm² / s (speed adjustment ratio coefficient); b2 = 0.3 mm / s (velocity reference parameter); Substituting δ=1.0mm, we get: I=40×1.0+20=60A, V=3.0 / 1.0+0.3=3.3mm / s, thus achieving parameter optimization for thickness adaptation.

[0049] The vision inspection component 26 is electrically connected to the control component (Siemens S7-1500 PLC, paired with an industrial touch screen). The control component integrates a deformation compensation module to address the thermal deformation problem of large-area welding of the roof sheet metal parts. It compares the deviation Δd = ||q(t) - q0(t)|| between the actual weld trajectory q(t) and the theoretical trajectory q0(t) in real time. The theoretical trajectory q0(t) is generated based on the roof design model (e.g., the starting weld point coordinates are (100, 50, 20) mm, the ending weld point coordinates are [missing information]). The arc-shaped trajectory of the weld point (1700, 50, 20) mm is measured. The actual trajectory q(t) is acquired in real time by the vision inspection component 26 (sampling frequency 2kHz). When welding reaches the 16th weld point, Δd = 0.6 mm is detected (exceeding the 0.5 mm threshold), triggering the compensation mechanism: the mold plate deflection angle θ = arcsin(Δd / R) is adjusted by the rotation drive component, where R = 250 mm (rotation radius, i.e., the distance from the rotation center of the mold plate to the weld). Substituting these values ​​into the calculation yields... θ=arcsin(0.6 / 250)≈0.137°; The rotary drive assembly precisely adjusts the mold plate deflection by 0.137°. Simultaneously, the welding robot's end effector posture is adjusted to keep the welding torch perpendicular to the weld tangent. The welding path is corrected based on a thermal deformation prediction model; the thermal deformation formula is ∆Q=α*I²*τ / (ρ*c*δ). Where τ is the dwell time of the electric arc per unit length (τ=1 / V=1 / 3.3≈0.303s / mm). The coefficient of thermal expansion of cold-rolled steel sheet DC01 is α = 11.7 × 10⁻ 6 / ℃; Density ρ = 7850 kg / m³; Specific heat capacity c = 460 J / (kg·℃); Substitute I=60A, τ=0.303s / mm, δ=1.0mm; The calculated value is ∆Q = 11.7 × 10⁻ 6×60²×0.303 / (7850×460×1.0×10⁻³)=11.7×10⁻ 6 ×3600×0.303 / (3611)≈1.08℃. Based on this thermal deformation amount, the welding path is corrected in advance to compensate for the deformation deviation caused by temperature changes.

[0050] The visual inspection component 26 also integrates a molten pool state recognition module and a weld seam tracking module. The molten pool state recognition module acquires molten pool images in real time through a visual camera and uses a five-dimensional joint iterative spatial segmentation method to extract features of the arc region, the main molten pool region, and the tail region of the molten pool in one go. Based on a CNN-LSTM hybrid neural network (trained with 1000 sets of cold-rolled steel plate welding data), a mapping relationship between parameters and penetration state is established. When an incomplete penetration state is detected, the system increases the welding current at a rate of 5A / s and decreases the welding speed at a rate of 0.005m / s² to quickly correct the penetration state. The weld seam tracking module uses a robust controller and a performance weight transfer function. W p (s)=(s / M+ω_B) / (s+ω_BA); The control weight transfer function is W_μ(s) = (S + ω_BC / M) / (A_s + ω_BC); Where M=1.8 (peak value constraint of sensitivity function); ωB = 50 rad / s (bandwidth frequency), A = 0.02 (steady-state error), ω_BC = 60 rad / s (control bandwidth frequency). Substituting these values ​​into the equation can effectively resist interference from reflections on the roof surface and welding spatter, ensuring the accuracy of weld seam tracking.

[0051] The control component integrates a sensor data fusion module, which performs spatiotemporal registration on the molten pool image and weld point cloud data acquired by the vision inspection component, and the current (sampling accuracy ±0.1A) and voltage (sampling accuracy ±0.01V) signals acquired by the welding system. This eliminates spatiotemporal data bias and is achieved through a formula. E(A)=(∑〖B∩C=A^(m1 (B) m2 (C)) 〗) / (1-k) calculates the confidence allocation of each sensor.

[0052] Assuming proposition A is "welding is normal", the basic probability assignment of sensor 1 (vision camera) to proposition B (molten pool is normal) is m1(B) = 0.85; The basic probability assignment of sensor 2 (current sensor) to proposition C (current stability) is m2(C) = 0.88; B∩C=A, conflict factor k=0.12 Substituting the values ​​into the equation, we get E(A) = (0.85 × 0.88) / (1 - 0.12) = 0.748 / 0.88 ≈ 0.85, which has a reliability of 85%. The system determines that the welding condition is normal.

[0053] If a sensor signal is abnormal (e.g., the vision camera is blocked by a splash, m1(B)=0.3), the system will automatically reduce its weight and rely on current and voltage sensor data to make decisions, thus avoiding welding defects caused by a single point of failure.

[0054] The welding process in this embodiment is as follows: First, adjust the position of the balance block 18 to balance the center of gravity of the roof sheet metal part and the mold plate 2. Place the roof sheet metal part on the support block 19. The control component opens the solenoid valve 24, the clamping cylinder 20 presses the sheet metal part, and the limit cylinder 16 drives the limit rod 15 to lock the mold plate. The rotation drive component drives the mold plate to rotate to the initial welding point angle, and the rotation monitoring component feeds back the angle signal to achieve precise positioning. The horizontal movement component drives the welding robot to the initial position, and the vision detection component collects point cloud data, calculates the thickness δ=1.0mm, and generates welding parameters I=60A, V=3.3mm / s. The welding operation is started, and the state of the molten pool and the weld trajectory are monitored in real time. When the deviation exceeds the threshold, deformation compensation is triggered. After a single welding point is completed, the mold plate rotates to the next angle, the horizontal movement component adjusts the robot position, and the welding process is repeated.

[0055] After all 32 welding points are completed, the rotary drive assembly drives the mold plate to rotate at high speed (angular velocity 5 rad / s), using centrifugal force to clean the welding slag. At the same time, the roof sheet metal is cooled by forced convection. After cooling to room temperature, the clamping cylinder resets, the limit rod locks the mold plate, and the welded roof sheet metal is taken out, completing the entire welding operation.

[0056] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0057] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A servo-driven rotary point welding device, characterized in that: include Clamping assembly (1), the clamping assembly (1) is used to clamp sheet metal parts, the clamping assembly (1) is mounted on mold plate (2); A rotary drive assembly (3) is mounted on a frame (4) and is connected to the end of the mold plate (2) via a connector; Rotation monitoring component (5), which is mounted on the frame (4) and is connected to the end of the mold plate (2) and the opposite side of the rotation drive component (3) via a connector; Welding assembly (6), which is located on the side wall of clamping assembly (1) and mounted on horizontal moving assembly (7).

2. The servo rotary point welding device according to claim 1, characterized in that: The rotary drive assembly (3) includes a drive shaft (8), one end of which is connected to a connector. The drive shaft (8) is mounted on the frame (4) via a shaft mounting base (9). A worm gear is connected to the end of the drive shaft (8). The worm gear is installed inside the housing (10). A servo power assembly (11) is also installed on the side wall of the housing (10). The output end of the servo power assembly (11) is connected to a worm that cooperates with the worm gear.

3. The servo rotary point welding device according to claim 1, characterized in that: The rotation monitoring component (5) includes a passive shaft (12), the end of which is connected to a connector. The passive shaft (12) is mounted on the frame (4) via a shaft mounting base. An encoder disk is connected to the end of the passive shaft (12). The encoder disk is mounted on a mounting housing. An encoder reading head that cooperates with the encoder disk is also installed inside the mounting housing.

4. The servo rotary point welding device according to claim 1, characterized in that: The mold plate (2) has a limiting plate (13) connected to its side wall. The limiting plate (13) has multiple spaced limiting cavities (14). A limiting cylinder (16) is installed on the frame (4). The output end of the limiting cylinder (16) is connected to a limiting rod (15) that cooperates with the limiting cavity (14). The limiting rod (15) is installed on the frame (4) through a guide seat (17). A balance block (18) is installed on the side wall of the mold plate (2) and on the side opposite to the clamping assembly (1).

5. The servo rotary point welding device according to claim 1, characterized in that: The clamping assembly (1) includes a support block (19) and a clamping cylinder (20). The upper end of the support block (19) abuts against the lower end face of the sheet metal part. The support block (19) has an air cavity (21). The support block (19) is connected to an air inlet pipe (22) and an air outlet pipe (23). Both the air inlet pipe (22) and the air outlet pipe (23) are connected to the air cavity (21). A solenoid valve (24) is installed on the air inlet pipe (22). The air outlet pipe (23) is also connected to the gas output end of the clamping cylinder (20). A clamping block is installed on the output end of the clamping cylinder (20).

6. The servo rotary point welding device according to claim 1, characterized in that: The welding assembly (6) includes a welding robot (25), which is mounted on the horizontal moving assembly (7), and the welding robot (25) also integrates a vision inspection assembly (26) at its end. The visual inspection component (26) is used to extract weld feature point sets based on cloud data. The weld trajectory curve L(t) is fitted using the least squares method, and the thickness of the sheet metal part is calculated based on the point cloud density distribution. ; Where δ is the estimated thickness of the sheet metal part, in millimeters; n is the total number of feature points in the point cloud data, which is a positive integer greater than 1; p i Let p be the three-dimensional coordinate vector of the i-th feature point. i =(x i ,y i ,z i ); L(t i ) represents the weld trajectory curve in parameter t i The coordinates of the point; t i For parametric coordinates, the value range is [0,1], representing the relative position of a point on the curve; Given the Euclidean norm, calculate the straight-line distance between two points; ∑ is the summation symbol, which sums up all terms from i to n; The visual inspection component calculates the welding current I and welding speed V based on the thickness of the sheet metal part. ; ; in, This is the welding current output value, in amperes. The ratio of current to thickness; δ is the estimated thickness of the sheet metal part, in millimeters; For current reference parameters; Welding speed, measured in millimeters per second; This is the proportional coefficient for speed adjustment; These are the speed reference parameters.

7. The servo rotary point welding device according to claim 6, characterized in that: The visual inspection component (26) is electrically connected to the control component. The control component integrates a deformation compensation module. The deformation compensation module compares the deviation Δd = ||q(t) - q0(t)|| in real time with the actual weld trajectory q(t). When Δd > 0.5 mm, the following compensation mechanism is triggered: The deflection angle of the mold plate is adjusted by the rotary drive component: θ = arcsin(Δd / R); Adjust the end effector posture of the welding robot to keep the welding torch perpendicular to the tangent of the weld seam at all times; Based on the heat deformation prediction model, the welding path is corrected in advance, and the formula for calculating the heat deformation is: ; Where R is the radius of rotation; This refers to the temperature change caused by the welding thermal cycle. The coefficient of thermal expansion of the material; This refers to the residence time of the electric arc per unit length. The density of the material; c represents the specific heat capacity of the material.

8. The servo rotary point welding device according to claim 6, characterized in that: The visual inspection component (26) also integrates a molten pool state recognition module and a weld tracking module. The molten pool state recognition module uses a five-dimensional joint iterative spatial segmentation method to extract the features of the arc region, the main molten pool region and the tail region of the molten pool at one time based on the molten pool image collected by the visual inspection component. The molten pool state recognition module establishes a mapping relationship between welding parameters and molten penetration state based on a CNN-LSTM hybrid neural network. When an incomplete molten penetration state is detected, the system increases the welding current at a rate of 3A / s to 7A / s and decreases the welding speed at a rate of 0.003m / s² to 0.007m / s². The weld seam tracking module uses a robust controller setting, and its weighting function is selected as follows: ; ; in, Let s be the performance weight transfer function, and s be the complex frequency variable; For complex variables in the Laplace transform; The peak value of the sensitivity function is typically set to 1.5-2.

0. For bandwidth frequency; This is the steady-state error; To control the weight transfer function; To control bandwidth frequency.

9. The servo rotary point welding device according to claim 7, characterized in that: The control component integrates a sensor data fusion module, which performs spatiotemporal registration of molten pool images, weld seam point cloud data, current and voltage signals, and calculates the reliability allocation function for each sensor. ; in, The basic probability assignment value for proposition A represents the degree of confidence in A; B and C represent different evidentiary propositions or assumptions; B∩C=A means that the intersection of propositions B and C equals A; For the first sensor, assign a basic probability to proposition B; For the second sensor, assign a basic probability to proposition C; It is a conflict factor.

Images