Variable parameter welding system and method for rectangular tube fence based on double robot perception positioning

Abstract

The application discloses a kind of based on double robot perception positioning's square tube fence variable parameter welding system and method, belong to welding automation technical field, it includes center control system, conveying and variable position mechanism, double machine welding execution mechanism, three-dimensional perception unit and welding power supply.This application identifies weld gap in real time by perception positioning unit, and according to gap size automatically matches expert database to adjust welding current, speed and swing amplitude;With single vertical square tube as basic welding unit, control double robot opposite synchronous welding, build symmetric heat field.This application can effectively guide galvanized steam directional escape, significantly reduce porosity defects, automatically compensate the gap fluctuation caused by blanking error, while ensuring weld forming consistency, inhibit structural deformation, greatly improve the production efficiency and intelligent level of square tube fence.

Description

Technical Field

[0001] This invention belongs to the field of welding automation technology, specifically, it relates to a variable parameter welding system and method for rectangular tube fences based on dual-robot perception and positioning. Background Technology

[0002] Rectangular tube welded fences are widely used in building enclosures, equipment protection, and industrial isolation. Their structure typically consists of multiple vertical and horizontal rectangular tubes welded together to form a complex grid structure. In actual production, multiple L-shaped fillet welds are usually formed between a single vertical and horizontal rectangular tube. For larger fence components, the total number of welds can often reach hundreds, which places extremely high demands on welding efficiency and consistency of quality.

[0003] Currently, rectangular tube fences are mostly made of galvanized material. The galvanized layer is highly susceptible to vaporization at the high temperatures of welding, generating a large amount of zinc vapor. If the zinc vapor fails to escape before the molten pool solidifies, it will form porosity defects inside the weld. In existing automated welding processes, due to the dense distribution of welds and the often continuous welding sequence, severe heat accumulation can easily occur in localized areas. This heat concentration not only affects the solidification behavior of the molten pool but also worsens the environment for zinc vapor escape, leading to a concentration of porosity defects in localized areas and severely impacting structural strength.

[0004] Furthermore, during the mass production of rectangular tube fences, length errors are unavoidable in the vertical rectangular tubes. When the vertical rectangular tubes are too short, uneven weld gaps will form between their end faces and the horizontal rectangular tubes. This fluctuation in gaps causes significant differences in the actual contact state of the fillet weld at different locations, directly interfering with the stability of the weld pool and the effective distribution of heat input, thereby further exacerbating the problems of unsightly weld formation and frequent internal defects.

[0005] While existing welding robot systems have improved production efficiency to some extent, their technical focus is largely on parallel operation at workstations or simple path planning, typically assuming that the assembly state of the workpieces to be welded is perfectly ideal and consistent. For issues such as controlling zinc vapor escape caused by material properties and weld gap variations due to material cutting errors, current technologies often rely on human experience to adjust preset parameters, lacking effective sensing and adaptive adjustment capabilities. In high-paced automated production environments, existing systems struggle to dynamically correct welding trajectories, postures, and parameters in response to real-time changes in gap dimensions, leading to weld burn-through or poor weld formation at large gap locations. Therefore, developing a welding system capable of collaborative multi-robot operation, real-time weld condition sensing, and adaptive adjustment of process parameters is of significant practical importance for improving the welding quality of galvanized rectangular tube fences. Summary of the Invention

[0006] The purpose of this invention is to provide a variable parameter welding system and method for rectangular tube fences based on dual-robot perception and positioning. It is mainly used to solve the porosity defects caused by the difficulty of zinc vapor escape during the automated welding process of galvanized rectangular tubes, as well as the problem of inconsistent weld gaps caused by material cutting size errors.

[0007] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0008] A variable-parameter welding system for rectangular tube fences based on dual-robot perception and positioning includes a central control system, a conveying and positioning mechanism, a dual-robot welding execution mechanism, a three-dimensional perception unit, and a welding power source. The conveying and positioning mechanism is equipped with a positioning fixture for carrying and driving the rectangular tube fence to move along its length. The dual-robot welding execution mechanism includes a first welding robot and a second welding robot symmetrically arranged on both sides of the conveying and positioning mechanism. The first and second welding robots are respectively equipped with welding torches and connected to the welding power source. The three-dimensional perception unit is installed on the end effector of the first or second welding robot to acquire three-dimensional point cloud data at the intersection of the vertical and horizontal rectangular tubes. The central control system establishes bidirectional communication links with the conveying and positioning mechanism, the first welding robot, the second welding robot, and the three-dimensional perception unit via an industrial Ethernet network.

[0009] Furthermore, in this invention, the central control system integrates an image processing module, which is used to extract features from the original point cloud acquired by the three-dimensional perception unit and calculate the axial gap value d between the vertical rectangular tube end face and the horizontal rectangular tube surface.

[0010] Furthermore, in this invention, the central control system integrates a process planning module, which has a preset gap parameter mapping expert library. The expert library stores welding current I, welding speed V, welding amplitude A, and oscillation frequency F corresponding to different axial gap value d intervals.

[0011] Furthermore, in this invention, when the central control system performs the welding task, it controls the first welding robot and the second welding robot to simultaneously perform hot symmetrical welding on the L-shaped fillet welds on both sides of a single vertical rectangular tube, so that the arc starting point, running trajectory and arc ending action of the two robots are consistent on the time axis.

[0012] Furthermore, in this invention, when the axial clearance value d satisfies d > 0.5 mm, the process planning module reduces the welding speed V and simultaneously increases the welding swing amplitude A according to the linear interpolation algorithm.

[0013] Furthermore, in this invention, the welding speed V is adjusted according to the formula V = V0. kv, where kv is the speed adjustment coefficient, with a value ranging from 0.6 to 0.9; the adjustment of the welding swing amplitude A follows the formula A = A0 ka, where ka is the swing amplitude adjustment coefficient, with a value ranging from 1.2 to 1.6.

[0014] Furthermore, in this invention, the process planning module has an automatic switching function for multi-layer and multi-pass welding. When the axial gap value d > 2.0 mm, the central control system controls the robot to switch from single-pass welding mode to multi-layer and multi-pass path planning.

[0015] Based on the above system, the present invention also provides a method for variable parameter welding of rectangular tube fences based on dual-robot perception and positioning, comprising the following steps:

[0016] S1, Unit Positioning Stage: The central control system sends instructions to the conveying and positioning mechanism to drive the rectangular tube fence to move a preset distance, so that the target vertical rectangular tube unit enters the welding envelope of the first welding robot and the second welding robot.

[0017] S2. Online sensing stage: The three-dimensional sensing unit performs structured light scanning on the area to be welded, and the central control system uses a random sampling consistency algorithm to fit the plane equation of the rectangular tube surface to identify the weld position coordinates and axial gap value d.

[0018] S3, Parameter Dynamic Mapping Stage: The central control system inputs the axial clearance value d to the process planning module, matches the corresponding process compensation coefficient from the expert database, and issues process parameter instructions to the first welding robot and the second welding robot respectively.

[0019] S4. Collaborative Welding Stage: After receiving the synchronous arc ignition signal, the first and second welding robots simultaneously ignite the arc and perform symmetrical welding along the preset L-shaped trajectory.

[0020] S5. Cyclic Operation Stage: After all the welds of a single vertical rectangular tube are completed, repeat steps S1 to S4 until the entire rectangular tube fence is welded.

[0021] In step S2, the central control system performs a secondary scan before welding using a three-dimensional sensing unit, compares the extracted weld centerline coordinates with the taught trajectory, calculates the spatial displacement deviations ΔX, ΔY, and ΔZ, and then superimposes these deviations into the robot's motion control algorithm in real time for trajectory compensation.

[0022] Compared with the prior art, the present invention has the following beneficial effects:

[0023] (1) The present invention adopts a dual-robot architecture for synchronous symmetrical welding on both sides of a single vertical rectangular tube, forming a centrally symmetrical thermal stress field in the weld area. This can not only offset the welding stress caused by the heat input on one side, but also control the overall deformation of the rectangular tube fence to within 0.5 mm / m. At the same time, the stable pressure gradient formed by the symmetrical thermal field can guide the zinc vapor generated by the gasification of the galvanized layer to escape directionally along the flow direction of the molten pool, avoiding the vapor being wrapped by the solidified molten pool to form internal pores, thus reducing the weld porosity defect rate.

[0024] (2) The present invention identifies the weld gap between the vertical and horizontal rectangular tubes in real time through a three-dimensional sensing unit. With the preset gap-process parameter mapping expert library, it can automatically match the welding current, speed and swing parameters according to different gap ranges such as within 0.5mm, 0.5-1.0mm, 1.0-2.0mm and above 2.0mm. In the case of large gap, it can automatically switch the multi-layer and multi-pass welding mode. It can compensate for the material size error within ±2mm without manual intervention. The weld penetration fluctuation range is controlled within ±0.2mm, which greatly reduces the probability of defects such as weld penetration, incomplete penetration and uneven forming.

[0025] (3) The present invention uses a single vertical rectangular tube as an independent welding unit. Through the intermittent displacement of the conveying and displacement mechanism and the coordinated operation of the two robots, the welding efficiency of the whole rectangular tube fence is more than 4 times higher than that of traditional manual welding and 1.8 times higher than that of single-machine automated welding. At the same time, the system’s built-in trajectory dynamic correction function and dual-machine interlock protection function can adapt to complex production scenarios such as material feeding error and clamping deviation, and the first-pass qualification rate of weld is improved. There is no need for subsequent manual welding, which greatly reduces the labor intensity and production cost.

[0026] (4) The central control system of the present invention supports access to the industrial Internet platform, which can form a closed loop iteration of gap distribution data, welding parameters and quality inspection data, and continuously optimize the parameter mapping relationship of the expert database through machine learning algorithms. It can adapt to the welding needs of rectangular tube materials with different thicknesses and different zinc coating contents. After subsequent expansion, it can also support the automated welding operation of other types of steel components, and has strong scene adaptability and technical scalability. Attached Figure Description

[0027] Figure 1 This is a simplified structural diagram of the system of the present invention.

[0028] Figure 2 This is a flowchart of the method of the present invention.

[0029] The names corresponding to the reference numerals in the attached figures are as follows:

[0030] 1. Conveying and positioning mechanism; 2. Positioning fixture; 3. Rectangular tube fence; 4. Welding torch; 5. First welding robot; 6. Second welding robot; 7. Vertical rectangular tube; 8. Horizontal rectangular tube. Detailed Implementation

[0031] The present invention will be further described below with reference to the accompanying drawings and embodiments. The embodiments of the present invention include, but are not limited to, the following embodiments.

[0032] like Figure 1 As shown, this invention discloses a variable-parameter welding system for rectangular tube fences based on dual-robot perception and positioning. The system includes a central control system, a first welding robot 5, a second welding robot 6, a conveying and positioning mechanism 1, a positioning fixture 2, a three-dimensional perception unit, and a welding power source. The central control system uses an industrial-grade control computer or a high-performance PLC as the logic control hub and data processing center for the entire system. The first welding robot 5 and the second welding robot 6 are symmetrically arranged on both sides of the conveying and positioning mechanism 1, and both robots' end effectors are equipped with welding torches 4. The conveying and positioning mechanism 1 controls the positioning fixture 2 to perform long-distance, rhythmic displacement on a horizontal guide rail via a servo drive system, conveying the rectangular tube fence 3 to be welded to the common working area of ​​the first welding robot 5 and the second welding robot 6.

[0033] At the hardware connectivity level, the central control system establishes real-time communication with the controllers of the first welding robot 5 and the second welding robot 6 via a high-speed industrial Ethernet, ensuring that the time delay of control command issuance is controlled within milliseconds. The welding power supply is electrically connected to the welding torches 4 of both robots, providing stable welding current and voltage. In this embodiment, the 3D perception unit employs a high-precision laser contour scanner, which is installed at the end of the first welding robot 5 or fixed above the welding station, used to perform 3D spatial modeling of the junction between the vertical rectangular tube 7 and the horizontal rectangular tube 8 before welding.

[0034] The rectangular tube fence 3 to be processed consists of one or two horizontal rectangular tubes 8 and several vertical rectangular tubes 7 arranged at equal intervals. Due to the length tolerance of the vertical rectangular tubes 7 during the initial cutting process, when multiple vertical rectangular tubes 7 are clamped in the positioning fixture 2, weld gaps of varying sizes will be formed between their end faces and the side surfaces of the horizontal rectangular tubes 8. This system defines a single vertical rectangular tube 7 as an independent welding unit, and completes all L-shaped fillet welds involved in this unit through the collaborative operation of two robots.

[0035] like Figure 2 As shown, the specific operating principle and process of this invention are as follows:

[0036] S1. Unit Positioning and Displacement Stage. The central control system first controls the conveying and displacement mechanism 1 to move the positioning fixture 2, which holds the workpiece, along the length direction, so that the first vertical rectangular tube 7 to be welded enters the welding station. At this time, the positioner rotates according to the preset process angle, so that the fillet weld formed by the vertical rectangular tube 7 and the horizontal rectangular tube 8 is in a flat fillet weld posture that is conducive to the formation of the molten pool. The positioning fixture 2 ensures that the vertical rectangular tube 7 does not shift or deviate during the movement by means of pneumatic or mechanical clamping.

[0037] S2. Online Sensing and Gap Recognition Stage. The 3D sensing unit is activated under the command of the central control system, performing point cloud scanning on the intersection area of ​​the four L-shaped weld seams in the current welding unit. The obtained 3D point cloud data is transmitted in real time to the image processing module of the central control system. The central control system uses a point cloud segmentation algorithm to extract the end face features of the vertical rectangular tube 7 and the surface features of the horizontal rectangular tube 8. By calculating the Euclidean distance between the two sets of point cloud features, the measured gap value d at each weld seam in the welding unit is accurately obtained.

[0038] S3. Dynamic Mapping Stage of Process Parameters. The central control system matches the optimal welding parameters from its built-in expert process database based on the measured gap value d. This expert database divides the gap into four intervals for differentiated control, as shown in Tables 1 and 2 below.

[0039] Table 1 Weld Size Division Table

[0040]

[0041] Table 2 Weld Parameter Classification Table

[0042]

[0043] When the measured gap value d ≤ 0.5 mm, the system determines it to be in a tight fit state. At this time, the central control system issues the reference current I0, the reference welding speed V0, and the reference swing amplitude A0, and adopts a low heat input mode to prevent the thin-walled rectangular tube from burning through. When the measured gap value d is between 0.5 mm and 1.0 mm, the system executes the first-level process compensation, adjusting the welding speed to V0. 0.9, and moderately increase the welding current, while simultaneously increasing the oscillation amplitude of welding torch 4 to A0. 1.2, By increasing the amount of weld deposited to fill the gap. When the measured gap value d is between 1.0 mm and 2.0 mm, the system executes the second-level process compensation, and the welding speed is significantly reduced to V0. 0.8, the swing amplitude expands to A0 1.4, to ensure the arc remains sufficiently on both sides of the gap, promoting fusion of the molten pool. When the measured gap value d When the thickness is 2.0mm, the central control system automatically triggers a multi-layer, multi-pass welding mode, first controlling the robot to perform root pass welding to close the gap, and then performing cover pass welding.

[0044] S4. Dual-machine collaborative thermo-symmetrical welding stage. This is the core physical process of the invention. The central control system synchronously starts the first welding robot 5 and the second welding robot 6. The welding torches 4 of the two robots are respectively pointed to the weld seams on opposite sides of the vertical rectangular tube 7. After receiving the synchronous arc ignition command, the two welding torches 4 simultaneously ignite the arc and advance from one end of the vertical rectangular tube 7 to the other in a completely symmetrical trajectory. During the welding process, since the two molten pools are generated synchronously, the resulting thermal field exhibits a centrally symmetrical distribution inside the vertical rectangular tube 7. The thermal stress fields generated by this thermo-symmetrical distribution cancel each other out, greatly reducing the welding deformation of the workpiece. More importantly, the symmetrical molten pool thermal field provides a stable pressure gradient to guide the zinc vapor generated by the vaporization of the galvanized layer, forcing the zinc vapor to be discharged along the direction of the molten pool fluid movement, thereby eliminating the technical defect of steam being trapped inside the molten pool and forming pores due to the accumulation of heat on one side.

[0045] S5. Dynamic Trajectory Correction and Cyclic Operation. The central control system monitors the feedback value of the welding voltage in real time during the welding process. If arc length fluctuations due to thermal deformation are detected, the system adds a correction amount to the robot's motion trajectory in real time through an arc tracking algorithm. After a single vertical rectangular tube 7 is welded, the first welding robot 5 and the second welding robot 6 return to their initial positions. The conveying and positioning mechanism 1 drives the positioning fixture 2 to step to the center position of the next vertical rectangular tube 7, repeating steps S1 to S4 until the entire rectangular tube fence 3 is processed.

[0046] In the above process, the adaptive control logic implemented by the central control system exhibits extremely high robustness. For the special material of galvanized rectangular tubes, the system controls the intensity of molten pool stirring by adjusting the welding oscillation amplitude A and oscillation frequency F. When the gap is large, increasing the oscillation amplitude not only fills the gap but also assists in the rupture and escape of bubbles within the molten pool through mechanical stirring. The central control system, through precise step-wise adjustment of the welding speed V, ensures that the heat input per unit length of weld remains within a reasonable range under different gaps, thereby guaranteeing the consistency of weld penetration.

[0047] When the 3D sensing unit reports a gap d of 2.5mm, the central control system logically determines to enter multi-layer welding mode. At this point, the first welding robot 5 executes a straight-line root pass trajectory, switching the welding power supply to pulse current mode. The high peak current of the pulse is used to achieve spray transition, quickly forming a thin root pass weld. Subsequently, the robot automatically returns to its origin, and the central control system recalculates the spatial coordinates of the capping layer based on the surface height after the root pass, controlling the robot to complete the capping operation with large swing amplitude and high current parameters. This step-by-step processing scheme replaces traditional manual welding, achieving full-process automation.

[0048] The motion control of the conveying and positioning mechanism 1 is also highly coupled with the welding cycle. The central control system obtains the absolute coordinates of the rectangular tube fence 3 in space in real time by reading the absolute position encoder data of the positioner servo motor. When the rectangular tube fence 3 has a slight curvature due to its own weight or clamping stress, the central control system uses multiple feature points obtained by the three-dimensional sensing unit to perform spatial plane fitting, automatically corrects the robot's base coordinate system, and ensures that the posture of the welding torch 4 is always perpendicular to the surface of the rectangular tube, thereby ensuring the aesthetics of the weld formation.

[0049] The advantages of this system in practical industrial scenarios lie in its high adaptability to the production environment. On traditional rectangular tube fence 3 production lines, due to tool wear from the cutting machine or human error, the length of the vertical rectangular tubes 7 often exhibits a normal distribution deviation. This invention, through sensing positioning and variable parameter coordination, transforms this uncontrollable assembly error into calculable process parameters. When processing data, the central control system considers not only the current static gap but also the robot's dynamic response under high-speed motion. Using a pre-established kinematic model, the system automatically executes a deceleration smoothing algorithm and simultaneously reduces the welding current when the welding torch 4 approaches the corner of the rectangular tube, preventing collapse or burn-through due to excessive heat accumulation at the corner.

[0050] To address the synchronization requirements of dual-robot collaboration, the central control system employs master-slave synchronous control logic. The first welding robot, 5, acts as the master, sending a synchronization clock signal, while the second welding robot, 6, acts as the slave, following the master's movement phase in real time. When one robot stops due to wire blockage or insufficient shielding gas pressure triggering an alarm, the central control system instantly cuts off the arc-starting signal of the other robot and stops its movement, preventing irreversible distortion of the workpiece caused by unilateral welding. This interlocking mechanism ensures that the conditions for symmetrical welding are always met.

[0051] The point cloud processing of the 3D sensing unit is also crucial for achieving precise welding in this system. Due to the reflective properties of the rectangular tube surface, the laser contour scanner generates noise during data acquisition. The central control system internally runs a denoising algorithm based on statistical filtering to remove outliers and uses the least squares method to fit the edge contour of the rectangular tube. By comparing the measured contour with the CAD theoretical model, the system not only calculates the gap d but also the tilt angle of the vertical rectangular tube 7 relative to the horizontal rectangular tube 8. Based on this, the central control system adjusts the pointing vector of the welding torch 4, ensuring precise distribution of arc energy at the weld root.

[0052] The welding power source and the central control system exchange parameters via analog or digital communication protocols. During the variable parameter execution phase, the current command issued by the central control system is converted into a high-frequency pulse waveform by the welding power source. For the galvanized layer, the system controls the preheating effect of the arc on the galvanized layer by adjusting the rise edge slope of the pulse and the base current. Before the arc contacts the molten pool, the galvanized layer in front of the welding trajectory is vaporized by the advance heat input and carried away by the fume extraction system, thereby further reducing the zinc content entering the molten pool and reducing the inducing factors for porosity formation at the source.

[0053] In summary, this embodiment organically integrates the first welding robot 5, the second welding robot 6, the positioner 4, and the three-dimensional perception unit through a central control system, constructing an intelligent welding unit with a closed loop of perception, decision-making, and execution. The system solves the problem of porosity in the galvanized layer by performing thermo-symmetrical collaborative welding on a single vertical rectangular tube 7, utilizing the physical thermal field distribution law. Simultaneously, based on a variable parameter mapping mechanism of the measured gap d, the welding process possesses adaptive compensation capabilities, enabling it to handle complex assembly errors. In practical applications, this system significantly improves the production efficiency of the rectangular tube fence 3, increasing the weld pass rate by more than 30% compared to traditional single-machine welding, and greatly reduces manual labor intensity, demonstrating significant technological advancement and practical value.

[0054] In future expansion implementations, the central control system can also connect to an industrial internet cloud platform to correlate and analyze the gap distribution data of each fence section with the final welding quality inspection data, continuously optimizing the mapping ratio in the expert database using machine learning algorithms. For example, it can fine-tune the heat input compensation coefficient based on the variation in galvanized layer thickness across different batches. This continuous evolution capability based on big data will enable the system to demonstrate greater versatility when facing welding tasks for a wider range of metal structural components. The combination of all the aforementioned hardware configurations and software logic constitutes a highly efficient, stable, and intelligent 3-parameter welding technology system for rectangular tube fences.

[0055] The above embodiments are merely one of the preferred embodiments of the present invention and should not be used to limit the scope of protection of the present invention. Any modifications or refinements made to the main design concept and spirit of the present invention that are not of substantial significance, but solve the same technical problem as the present invention, should be included within the scope of protection of the present invention.

Claims

1. A variable-parameter welding system for rectangular tube fences based on dual-robot perception and positioning, characterized in that, The system includes a central control system, a conveying and positioning mechanism (1), a dual-machine welding actuator, a three-dimensional sensing unit, and a welding power source. The conveying and positioning mechanism (1) is equipped with a positioning fixture (2) for carrying and driving the rectangular tube fence (3) to move along its length. The dual-machine welding actuator includes a first welding robot (5) and a second welding robot (6) symmetrically arranged on both sides of the conveying and positioning mechanism (1). The first welding robot (5) and the second welding robot (6) are respectively equipped with welding torches (4) and connected to the welding power source. The three-dimensional sensing unit is installed on the end effector of the first welding robot (5) or the second welding robot (6) for acquiring three-dimensional point cloud data at the junction of the vertical rectangular tube (7) and the horizontal rectangular tube (8). The central control system establishes bidirectional communication links with the conveying and positioning mechanism (1), the first welding robot (5), the second welding robot (6), and the three-dimensional sensing unit via industrial Ethernet.

2. The rectangular tube fence variable parameter welding system based on dual-robot perception and positioning according to claim 1, characterized in that, The central control system integrates an image processing module, which is used to extract features from the original point cloud collected by the three-dimensional perception unit and calculate the axial gap value d between the end face of the vertical rectangular tube (7) and the surface of the horizontal rectangular tube (8).

3. The rectangular tube fence variable parameter welding system based on dual-robot perception and positioning according to claim 2, characterized in that, The central control system integrates a process planning module, which has a preset gap parameter mapping expert library. The expert library stores welding current I, welding speed V, welding amplitude A, and oscillation frequency F corresponding to different axial gap value d intervals.

4. The rectangular tube fence variable parameter welding system based on dual-robot perception and positioning according to claim 3, characterized in that, When performing welding tasks, the central control system controls the first welding robot (5) and the second welding robot (6) to perform hot symmetrical welding on the L-shaped fillet welds on both sides of a single vertical rectangular tube (7) simultaneously, so that the starting point, running trajectory and ending action of the two robots are consistent on the time axis.

5. The rectangular tube fence variable parameter welding system based on dual-robot perception and positioning according to claim 4, characterized in that, When the axial clearance value d satisfies d > 0.5 mm, the process planning module reduces the welding speed V and simultaneously increases the welding swing amplitude A according to the linear interpolation algorithm.

6. The rectangular tube fence variable parameter welding system based on dual-robot perception and positioning according to claim 5, characterized in that, The welding speed V is adjusted according to the formula V = V0. kv, where kv is the speed adjustment coefficient, with a value ranging from 0.6 to 0.9; the adjustment of the welding swing amplitude A follows the formula A = A0 ka, where ka is the swing amplitude adjustment coefficient, with a value ranging from 1.2 to 1.

6.

7. The rectangular tube fence variable parameter welding system based on dual-robot perception and positioning according to claim 6, characterized in that, The process planning module has an automatic switching function for multi-layer and multi-pass welding. When the axial gap value d > 2.0mm, the central control system controls the robot to switch from single-pass welding mode to multi-layer and multi-pass path planning.

8. A method for variable-parameter welding of rectangular tube fences based on dual-robot perception and localization, characterized in that, The application of the rectangular tube fence (3) variable parameter welding system based on dual robot perception and positioning as described in any one of claims 1 to 7 includes the following steps: S1, Unit positioning stage: The central control system sends an instruction to the conveying and positioning mechanism (1) to drive the rectangular tube fence (3) to move a preset distance so that the target vertical rectangular tube (7) unit enters the welding envelope of the first welding robot (5) and the second welding robot (6); S2. Online sensing stage: The three-dimensional sensing unit performs structured light scanning on the area to be welded, and the central control system uses a random sampling consistency algorithm to fit the plane equation of the rectangular tube surface to identify the weld position coordinates and axial gap value d. S3, Parameter Dynamic Mapping Stage: The central control system inputs the axial clearance value d to the process planning module, matches the corresponding process compensation coefficient from the expert database, and issues process parameter instructions to the first welding robot (5) and the second welding robot (6) respectively. S4, Collaborative Welding Stage: After receiving the synchronous arc ignition signal, the first welding robot (5) and the second welding robot (6) simultaneously ignite the arc and perform symmetrical welding along the preset L-shaped trajectory; S5. Cyclic Operation Stage: After all the welds of a single vertical rectangular tube (7) are completed, repeat steps S1 to S4 until the entire rectangular tube fence (3) is welded.

9. The method for variable-parameter welding of rectangular tube fences based on dual-robot perception and positioning according to claim 8, characterized in that, In step S2, the central control system performs a secondary scan before welding using a three-dimensional sensing unit, compares the extracted weld centerline coordinates with the taught trajectory, calculates the spatial displacement deviations ΔX, ΔY, and ΔZ, and then superimposes these deviations into the robot's motion control algorithm in real time for trajectory compensation.

Images