A multi-robot cooperation welding trajectory planning method and system based on 3D vision

Abstract

The application is suitable for the field of industrial automatic welding technology, and provides a welding trajectory planning method and system for multi-robot cooperation based on 3D vision, which comprises the following steps: obtaining historical welding data of a preset number of historical workpieces that have been welded before target batch workpieces in a double-robot handover welding area; extracting laser projection patterns in the handover welding area of the front and rear robots corresponding to each historical workpiece from the historical welding data, respectively, and determining the maximum change area of each laser projection pattern. By introducing the maximum change area of the laser projection pattern and the 3D vision analysis technology, and combining the heat accumulation parameters of the rear robot, an angle correction mechanism based on the deviation trend slope is constructed, and dynamic compensation for the welding deviation evolution trend with time in the double-robot handover welding scene is realized for the first time.

Description

Technical Field

[0001] This invention belongs to the field of industrial automation welding technology, and in particular relates to a welding trajectory planning method and system based on 3D vision and multi-robot collaboration. Background Technology

[0002] In modern industrial manufacturing, multi-robot collaborative welding has become a mainstream technology for improving welding efficiency and consistency. To achieve efficient welding scheduling across multiple workstations, several workpieces with identical structures are typically grouped together, and multiple welding robots process their respective designated welding areas in parallel, cooperating sequentially within overlapping welding areas. In such dual-robot collaborative welding scenarios, to ensure weld continuity and quality consistency, the subsequent robot needs to fine-tune its angle based on the weld feature information left by the preceding robot in the welding area, in order to address the impact of weld geometry changes, thermal deformation, and other factors on trajectory planning.

[0003] To support the aforementioned angle fine-tuning operations, existing technologies widely employ laser projection devices combined with visual sensing modules to optically scan the weld surface. By identifying local changes in the laser pattern, they assist in determining features such as weld boundaries, curvature variations, and height abrupt changes. Based on these identification results, welding angle correction values ​​are generated, guiding the subsequent robot to correct the welding path within the junction area. This type of technical solution has been applied in many industrial welding lines, capable of correcting local trajectory changes caused by workpiece precision errors or thermal deformation to a certain extent, thereby improving the processing accuracy of weld joints.

[0004] However, existing technical solutions generally suffer from the following technical defects: First, the system assumes that the rear robot can stably execute the angle correction value calculated based on the weld characteristics of the front robot, without considering the decrease in motion accuracy of the rear robot body due to heat accumulation during long-term high-intensity welding; second, when the number of workpieces in the target batch is large, the rear robot experiences mechanical structure lag and increased angle execution error due to heat accumulation, ultimately causing the actual deviation between its welding effect and the set correction angle to gradually widen. Existing technologies do not dynamically identify and correct this deviation trend, which easily leads to the deterioration of the handover weld quality as the batch progresses, making it difficult to meet the requirements of high-consistency welding quality. Therefore, it is urgent to construct a path planning method for dual-robot handover scenarios that can sense welding deviation trends and dynamically correct angle compensation values. Summary of the Invention

[0005] The purpose of this invention is to provide a welding trajectory planning method based on 3D vision and multi-robot collaboration, which aims to solve the problems mentioned in the background art.

[0006] This invention is implemented as follows: a welding trajectory planning method based on 3D vision for multi-robot collaboration, the method comprising:

[0007] Obtain historical welding data of a preset number of historical workpieces that have been welded before the target batch of workpieces, located in the dual-robot handover welding area;

[0008] The laser projection patterns of the welding areas of the front and rear robots corresponding to each historical workpiece are extracted from the historical welding data, and the maximum change area of ​​each laser projection pattern is determined.

[0009] The deviation of the area with the greatest change in the laser projection pattern of the robot before and after is quantified, and the thermal accumulation parameters of the robot after a preset number of historical workpieces over time are obtained.

[0010] Analyze a pre-set number of historical workpieces to determine if the following characteristic pattern exists: as time goes on, the deviation trend of the maximum change area of ​​the laser projection pattern of the front and back robots and the change trend of the thermal accumulation parameters of the back robot satisfy a pre-set correlation.

[0011] If the above-mentioned feature pattern is confirmed, the initial angle correction value set by the robot for the welding area is obtained, and a correction factor is generated based on the trend of the deviation of the maximum change area of ​​the laser projection pattern of the front and rear robots, and the initial angle correction value is dynamically adjusted.

[0012] As a further limitation of the technical solution of the present invention, the laser projection pattern refers to a specific spot or pattern projected onto the weld surface by the laser equipment carried by the robot during the welding process, which is used to assist in identifying the weld contour, weld geometric changes and local deformation. The pattern will produce identifiable structural changes during the welding process due to the thermal deformation of the workpiece or changes in the surface reflection characteristics.

[0013] As a further limitation of the technical solution of this invention, the step of extracting the laser projection pattern in the handover welding area of ​​the front and rear robots corresponding to each historical workpiece from historical welding data, and determining the maximum change area of ​​each laser projection pattern includes:

[0014] Extract the laser projection patterns of the front and rear robots corresponding to each historical workpiece from the historical welding data and project them onto the junction welding area;

[0015] 3D vision technology was used to reconstruct the structure and analyze the curvature of the extracted laser projection pattern, identify the region in the pattern that changes most significantly with the welding process, and determine this region as the corresponding region of maximum change.

[0016] As a further limitation of the technical solution of this invention, the step of quantifying the deviation of the area of ​​maximum change in the laser projection pattern of the robots before and after the transition, and obtaining the thermal accumulation parameters of the robots of the pre-set number of historical workpieces over time, includes:

[0017] Based on 3D vision technology, the spatial coordinates of the area with the greatest change in the laser projection pattern between the front robot and the rear robot in each historical workpiece are extracted.

[0018] Calculate the spatial deviation of the area of ​​maximum change in the laser projection pattern of the robot before and after, and use this deviation as the degree of deviation of the corresponding historical workpiece;

[0019] The robot acquires post-weld heat accumulation parameters for each historical workpiece at the time of welding completion. These parameters include the cumulative operation time during the welding process, the temperature change value of key parts, and their derived functions.

[0020] As a further limitation of the technical solution of the present invention, the preset correlation refers to the following: in the previous preset number of historical workpieces, there is a positive correlation between the changing trend of the subsequent robot thermal accumulation parameters and the changing trend of the deviation of the maximum change area of ​​the laser projection pattern of the previous and subsequent robots, and the positive correlation satisfies the following conditions: the slopes of the changing trends of the two within the preset time interval are respectively extracted as a first slope value and a second slope value, and the ratio of the first slope value to the second slope value is within the allowable range expressed by the preset function, or satisfies the preset multiple relationship between the second slope value and the first slope value, thereby judging that the trend of the influence of the subsequent robot thermal accumulation on the welding deviation of the joint welding area conforms to the regularity requirement.

[0021] As a further limitation of the technical solution of this invention embodiment, if the above-mentioned feature pattern is determined to exist, the steps of obtaining the initial angle correction value set by the robot for the joint welding area, and generating a correction factor based on the changing trend of the deviation of the maximum change area of ​​the laser projection pattern of the front and rear robots, and dynamically adjusting the initial angle correction value include:

[0022] After confirming the existence of the above-mentioned feature patterns, the initial angle correction value set by the robot based on the joint welding area is obtained;

[0023] The slope of the trend of the deviation of the area with the greatest change in the laser projection pattern of the robot before and after the acquisition within a preset time interval: the second slope value;

[0024] The second slope value is set as the correction factor, and the initial angle correction value is corrected by the correction factor to obtain the corrected angle correction value.

[0025] The corrected angle correction value is applied to the welding trajectory planning of the robot in the handover welding area of ​​other workpieces to be welded in the target batch of workpieces, so as to achieve compensatory optimization of the initial angle correction value.

[0026] A welding trajectory planning system based on 3D vision and multi-robot collaboration, the system comprising: a data acquisition module, a maximum change area determination module, a deviation quantification module, a feature pattern judgment module, and an angle correction value adjustment module, wherein:

[0027] The data acquisition module is used to acquire historical welding data of a preset number of historical workpieces that have been welded before the target batch of workpieces, located in the welding area of ​​the dual robots.

[0028] The maximum change area determination module is used to extract the laser projection pattern in the welding area of ​​the front and rear robots corresponding to each historical workpiece from historical welding data, and to determine the maximum change area of ​​each laser projection pattern. The laser projection pattern refers to the specific light spot or pattern projected onto the weld surface by the laser equipment carried by the robot during the welding process. It is used to help identify the weld contour, weld geometric changes and local deformation. The pattern will produce identifiable structural changes during the welding process due to the thermal deformation of the workpiece or the change of surface reflection characteristics.

[0029] The deviation quantification module is used to quantify the deviation of the area with the greatest change in the laser projection pattern of the front and rear robots in the junction area, and to obtain the thermal accumulation parameters of the rear robot as a preset number of historical workpieces extend over time.

[0030] The feature pattern judgment module is used to analyze a preset number of historical workpieces and determine whether the following feature pattern exists: as time goes on, the deviation trend of the maximum change area of ​​the laser projection pattern of the front and back robots and the change trend of the thermal accumulation parameter of the back robot satisfy a preset correlation relationship.

[0031] The preset correlation refers to the following: in the previous preset number of historical workpieces, there is a positive correlation between the changing trend of the subsequent robot thermal accumulation parameters and the changing trend of the deviation of the maximum change area of ​​the laser projection pattern of the previous and subsequent robots, and the positive correlation satisfies the following conditions: the slopes of the changing trends of the two within the preset time interval are extracted as a first slope value and a second slope value, respectively, and the ratio of the first slope value to the second slope value is within the allowable range expressed by the preset function, or satisfies the preset multiple relationship between the second slope value and the first slope value, thereby judging that the trend of the influence of the subsequent robot thermal accumulation on the welding deviation of the joint welding area conforms to the regularity requirements;

[0032] The angle correction value adjustment module is used to obtain the initial angle correction value set by the robot for the joint welding area if the above feature pattern is determined to exist, and to generate a correction factor based on the trend of the deviation of the maximum change area of ​​the laser projection pattern of the front and rear robots, and dynamically adjust the initial angle correction value.

[0033] As a further limitation of the technical solution of this embodiment of the invention, the maximum change region determination module specifically includes:

[0034] The pattern extraction unit is used to extract the laser projection patterns projected by the front and rear robots onto the junction welding area corresponding to each historical workpiece from historical welding data.

[0035] The pattern change analysis unit is used to perform structural reconstruction and curvature analysis on the extracted laser projection pattern using 3D vision technology, identify the region in the pattern that changes most significantly with the welding process, and determine that region as the corresponding region of maximum change.

[0036] As a further limitation of the technical solution of this embodiment of the invention, the deviation quantification module specifically includes:

[0037] The coordinate extraction unit is used to extract the spatial coordinates of the area with the greatest change in the laser projection pattern between the front robot and the rear robot in each historical workpiece based on 3D vision technology.

[0038] The deviation calculation unit is used to calculate the deviation of the area of ​​maximum change in the laser projection pattern of the robot before and after in terms of spatial position, and to use the deviation as the degree of deviation of the corresponding historical workpiece.

[0039] The thermal accumulation parameter acquisition unit is used to acquire the post-robot thermal accumulation parameters when the welding of each historical workpiece is completed. The thermal accumulation parameters include the cumulative operation time during the welding process, the temperature change value of key parts, and their derived functions.

[0040] As a further limitation of the technical solution of this embodiment of the invention, the angle correction value adjustment module specifically includes:

[0041] The initial value acquisition unit is used to acquire the initial angle correction value set by the robot based on the joint welding area after determining that the above feature pattern exists;

[0042] The slope value acquisition unit is used to acquire the trend of the deviation of the maximum change area of ​​the laser projection pattern of the robot before and after the change within a preset time interval: the slope value of the second slope value.

[0043] Angle correction value adjustment unit is used to set the second slope value as a correction factor, and correct the initial angle correction value through the correction factor to obtain the corrected angle correction value;

[0044] The correction value application unit is used to apply the corrected angle correction value to the welding trajectory planning of the robot in the handover welding area of ​​other workpieces to be welded in the target batch of workpieces, so as to achieve compensatory optimization of the initial angle correction value.

[0045] Compared with the prior art, the present invention has the following beneficial effects:

[0046] This invention introduces the identification of the maximum change area of ​​the laser projection pattern and 3D visual analysis technology, combined with the thermal accumulation parameters of the subsequent robot, to construct an angle correction mechanism based on the deviation trend slope. This achieves, for the first time, dynamic compensation for the evolution of welding deviation over time in a dual-robot welding scenario. Compared to existing technologies that only set fixed angle correction values ​​based on the weld characteristics of the preceding robot, this invention can identify and correct correction failures caused by thermal accumulation in the subsequent robot, ensuring that the welding angle adjustment always adapts to the current working conditions. This significantly improves the welding consistency and weld quality in the junction area, demonstrating outstanding intelligence, adaptability, and engineering applicability. Attached Figure Description

[0047] Figure 1 A flowchart of the method provided in the embodiments of the present invention;

[0048] Figure 2 This is a flowchart illustrating the method for determining the region of maximum change in a laser projection pattern in an embodiment of the present invention.

[0049] Figure 3 This is a flowchart illustrating the calculation of the deviation degree corresponding to historical workpieces in the method provided in this embodiment of the invention;

[0050] Figure 4 This is a flowchart illustrating the correction of the initial angle correction value in the method provided in this embodiment of the invention;

[0051] Figure 5 Application architecture diagram of the system provided in the embodiments of the present invention;

[0052] Figure 6 This is a structural block diagram of the module for determining the maximum changing region in the system provided in this embodiment of the invention;

[0053] Figure 7 This is a structural block diagram of the deviation quantification module in the system provided in the embodiments of the present invention;

[0054] Figure 8 This is a structural block diagram of the angle correction value adjustment module in the system provided in the embodiment of the present invention. Detailed Implementation

[0055] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0056] Figure 1 A flowchart of the method provided by an embodiment of the present invention is shown.

[0057] Specifically, a welding trajectory planning method based on 3D vision and multi-robot collaboration includes the following steps:

[0058] Step S100: Obtain historical welding data of a preset number of historical workpieces that have been welded before the target batch of workpieces, located in the dual-robot handover welding area.

[0059] In this embodiment of the invention, the target batch of workpieces refers to a group of workpieces to be welded that have completely identical structural shape, material composition and welding process parameters, usually multiple duplicates of the same model in a production batch.

[0060] The transition welding area refers to the area where the welding work of the preceding robot is completed and the welding work of the following robot begins. It is usually located in the space between the end and the beginning of the respective working paths of the preceding and following robots. Within this area, the welding tasks of the two robots are sequential, and the continuity and geometric consistency of the weld directly affect the welding quality. It is the most critical part in welding trajectory planning that is most prone to deviation.

[0061] In this invention, historical welding data specifically includes: welding image data of completed welded workpieces, especially the laser projection pattern sequence in the handover welding area; execution information such as trajectory control parameters, start and end coordinates, movement speed and angle correction values ​​recorded by the front robot and the rear robot during the welding task on the corresponding workpiece; and heat accumulation state parameters formed by the rear robot during continuous welding, such as cumulative operation time, temperature change curve of key parts and its derived heat load index.

[0062] The laser projection pattern refers to a specific spot or pattern projected onto the weld surface by the laser equipment carried by the robot during the welding process. It is used to help identify the weld contour, weld geometry changes and local deformation. The pattern will produce identifiable structural changes during the welding process due to the thermal deformation of the workpiece or changes in the surface reflection characteristics.

[0063] The laser projection pattern refers to a specific spot or pattern projected onto the weld surface by the laser equipment mounted on the robot during the welding process. This pattern assists in identifying the weld contour, weld geometry, and local deformation. The pattern undergoes identifiable structural changes during welding due to workpiece thermal deformation or variations in surface reflectivity. This type of laser pattern recognition is a relatively mature weld-aiding identification method in existing technology. It is typically used in conjunction with 3D vision technology to improve recognition accuracy and responsiveness to minute weld deformations. In practical applications, both the front and rear robots should be equipped with such laser projection devices to ensure a continuous and symmetrical geometric recognition basis in the weld junction area.

[0064] In existing technologies, the aforementioned laser pattern and 3D vision technology are typically used to identify the surface condition of the weld. The system then automatically analyzes the geometric changes in the weld after the preceding robot completes the welding, particularly morphological information such as linear deflections, curvature changes, or discontinuities in the weld trajectory. Based on this, an angle correction value for the subsequent robot in the weld seam area is generated to adjust the initial trajectory planning of all subsequent workpieces in the target batch, ensuring welding continuity and joint accuracy. However, a key drawback of existing technologies is that this angle correction value is calculated only based on the weld output of the preceding robot, without fully considering the heat accumulation effect of the subsequent robot during long-term continuous welding. As the subsequent robot continues to operate among the batch of workpieces, its internal structure, welding torch mechanism, or end effector are affected by factors such as thermal expansion, stiffness changes, or dynamic response lag. This can lead to a gradual deviation between the actual welding effect and the geometric trend of the weld seam reserved by the preceding robot, even when the same angle correction value is used. This deviation trend is not recognized by the original correction model, resulting in a gradual increase in weld seam deviation, affecting welding quality and handover stability.

[0065] This invention addresses the aforementioned problems by constructing a correlation model between thermal accumulation parameters and the evolution trend of weld pattern deviation, thereby achieving dynamic compensation for the initial angle correction value. This enables the robot to stably achieve high-precision welding connections under different thermal conditions.

[0066] Furthermore, the 3D vision-based multi-robot collaborative welding trajectory planning method also includes the following steps:

[0067] Step S200: Extract the laser projection pattern in the welding area of ​​the front and rear robots corresponding to each historical workpiece from the historical welding data, and determine the maximum change area of ​​each laser projection pattern.

[0068] Specifically, Figure 2 A flowchart is shown to determine the region of maximum variation in the laser projection pattern.

[0069] The process of extracting the laser projection pattern from the welding area of ​​the front and rear robots corresponding to each historical workpiece from historical welding data, and determining the area of ​​maximum change for each laser projection pattern, specifically includes the following steps:

[0070] Step S201: Extract the laser projection patterns of the front and rear robots corresponding to each historical workpiece from the historical welding data and project them onto the joint welding area.

[0071] Step S202: Use 3D vision technology to perform structural reconstruction and curvature analysis on the extracted laser projection pattern, identify the region in the pattern that changes most significantly with the welding process, and determine that region as the corresponding region of maximum change.

[0072] In this embodiment of the invention, to extract the weld state change characteristics of each historical workpiece within the handover welding area, it is necessary to obtain the corresponding laser projection patterns of the front and rear robots within the handover welding area from the historical welding data, and then, using 3D vision analysis, identify the key areas in the patterns that best reflect the evolution of the welding state. This process includes the following two specific steps:

[0073] First, the system extracts the laser projection pattern of each historical workpiece in the handover welding area from historical welding data using image decoding and pose matching algorithms. Specifically, based on an image timestamp and robot task log matching strategy, the system extracts the corresponding laser pattern image frames in the handover welding area for each of the preceding and following robot's respective work stages. The extraction process combines the projected structured light encoding information with historical workpiece positioning information to achieve high-precision reconstruction and segmented extraction of the pattern within the handover area.

[0074] Subsequently, 3D vision technology was used to reconstruct and analyze the curvature of the laser-projected pattern. Specifically, spatial point cloud reconstruction technology was employed to restore the two-dimensional image sequence into a three-dimensional structural outline, and a regional curvature change detection algorithm was used to identify geometrically abrupt regions in the pattern caused by factors such as welding stress release, thermal deformation, or structural discontinuities. These regions typically appear as local segments or edge bands where the curvature significantly deviates from the continuous trend, reflecting the concentrated response area of ​​the weld morphology caused by external disturbances.

[0075] Identifying the region of greatest change in the pattern aims to obtain representative feature areas that most effectively characterize the trend of changes on the weld surface. The magnitude and stability of this region directly reflect the influence of complex factors such as heat-affected zones, stress disturbances, and changes in robot execution accuracy on the weld structure during welding. Therefore, it becomes a key data foundation for judging welding deviation trends and determining the basis for angle correction. Especially in multi-robot welding scenarios, the degree of offset of this region of greatest change not only reflects changes in the workpiece state but may also indirectly reveal subtle differences or coordination mismatches between the welding behaviors of two robots, serving as an important basis for subsequent trajectory correction judgments.

[0076] Furthermore, the 3D vision-based multi-robot collaborative welding trajectory planning method also includes the following steps:

[0077] Step S300: Quantify the degree of deviation of the area with the greatest change in the laser projection pattern between the front and rear robots in the junction area, and obtain the thermal accumulation parameters of the rear robot for a preset number of historical workpieces over time.

[0078] Specifically, Figure 3 A flowchart is shown to calculate the degree of deviation corresponding to historical artifacts.

[0079] The specific steps for obtaining the thermal accumulation parameters of the robot before and after the laser projection pattern in the junction area, which involves quantifying the deviation of the area with the greatest change in the laser projection pattern between the robots before and after the transition, and obtaining the thermal accumulation parameters of the robot over time for a predetermined number of historical workpieces, include the following:

[0080] Step S301: Based on 3D vision technology, extract the spatial coordinates of the area with the greatest change in the laser projection pattern between the front robot and the rear robot in each historical workpiece.

[0081] Step S302: Calculate the spatial deviation of the area of ​​maximum change in the laser projection pattern of the robot before and after, and use this deviation as the degree of deviation of the corresponding historical workpiece.

[0082] Step S303: Obtain the post-robot thermal accumulation parameters when the welding of each historical workpiece is completed. The thermal accumulation parameters include the cumulative operation time during the welding process, the temperature change value of key parts, and their derived functions.

[0083] In this embodiment of the invention, to achieve quantitative analysis of the differences in the surface state response of the weld seam between the front and rear robots in the junction area, the system uses 3D vision technology to spatially compare the area of ​​maximum change in the laser projection pattern, and combines this with the robot's workload evolution process after extracting thermal state parameters. The specific implementation process is as follows:

[0084] First, in step S301, the system extracts the spatial coordinates of the region with the greatest change in the laser projection pattern between the front and rear robots in the intersection area of ​​each historical workpiece based on spatial point cloud registration technology. Specifically, a multi-view image 3D reconstruction algorithm is used to perform point cloud modeling of the region with the greatest change in the laser pattern, and spatial registration is performed by combining laser encoding information and robot posture data to accurately restore the position coordinates of the region with the greatest change in the actual workpiece coordinate system.

[0085] Subsequently, in step S302, the system performs difference analysis on the spatial coordinates of the corresponding areas of maximum change between the front and rear robots based on the three-dimensional Euclidean distance calculation method, obtaining their offset in three-dimensional space. This offset is defined in this invention as the "degree of deviation" of the corresponding historical workpiece, used to measure the consistency and response error of the two robots in the critical area of ​​the weld seam during the welding process. The larger the value, the more significant the dynamic deviation between the actual welding behavior of the rear robot and the expected trajectory of the front robot.

[0086] In step S303, to identify the thermal accumulation factors affecting the evolution of the deviation degree, the system further extracts the thermal accumulation parameters of the robot after the welding of each historical workpiece is completed. The thermal accumulation parameters mainly include the following:

[0087] Cumulative working time: refers to the total continuous working time since the robot started welding, used to measure the total duration of the heat source's continuous effect.

[0088] Temperature change values ​​of key parts: The temperature change curve of the robot body in high heat load areas (such as welding ends, joint motor housings) during operation is obtained by embedded or external thermal sensors, reflecting the internal heat diffusion state.

[0089] Derived functions: To facilitate subsequent trend analysis, the system calculates the first derivative (temperature rise rate), local extreme point distribution, or thermal equilibrium recovery time based on the above temperature change curves, so as to represent the dynamic characteristics of the heat load in a structured way.

[0090] Based on the above parameters, the thermal accumulation characteristics of the robot not only reflect its current thermal load level.

[0091] Furthermore, the 3D vision-based multi-robot collaborative welding trajectory planning method also includes the following steps:

[0092] Step S400: Analyze the historical workpieces of the previous preset number and determine whether the following characteristic pattern exists: as time goes on, the deviation trend of the maximum change area of ​​the laser projection pattern of the front and back robots and the change trend of the thermal accumulation parameter of the back robot satisfy the preset correlation relationship.

[0093] The preset correlation refers to the following: in a preset number of historical workpieces, there is a positive correlation between the changing trend of the subsequent robot thermal accumulation parameters and the changing trend of the deviation of the maximum change area of ​​the laser projection pattern of the preceding and subsequent robots, and the positive correlation satisfies the following conditions: the slopes of the changing trends of the two within a preset time interval are extracted as a first slope value and a second slope value, respectively, and the ratio of the first slope value to the second slope value is within the allowable range expressed by the preset function, or satisfies the preset multiple relationship between the second slope value and the first slope value, thereby determining that the trend of the influence of the subsequent robot thermal accumulation on the welding deviation of the joint welding area conforms to the regularity requirement.

[0094] In this embodiment of the invention, the core of step S400 is to identify a deviation pattern with associated features. This pattern indicates that there is a highly coupled intrinsic relationship between the heat accumulation change trend and the laser projection pattern offset trend of the robot in the process of continuously completing the handover welding task, thereby providing a reliable judgment basis for the generation of subsequent correction strategies.

[0095] First, if, in a predetermined number of historical workpieces, it is observed that "the deviation of the maximum variation area of ​​the laser projection patterns of the front and rear robots shows a continuous upward trend over time," it can be preliminarily determined that as the workpieces progress sequentially, the spatial overlap between the laser pattern formed by the welding performed by the front robot and the laser pattern formed by the subsequent robot gradually decreases. This gradual increase in deviation indicates that the subsequent robot's responsiveness to existing trajectories or weld conditions during the welding process is decreasing, signifying a continuous deterioration in its welding stability and joint consistency.

[0096] Furthermore, if the evolution trend of such deviation is positively correlated with the change trend of the thermal accumulation parameters of the subsequent robot, that is, as time progresses, the thermal accumulation parameters (such as the temperature of key parts, operation time, etc.) increase accordingly and increase synchronously with the degree of deviation, it indicates that the thermal load accumulation generated by the subsequent robot due to continuous operation has significantly affected the mechanical rigidity, positioning accuracy or response rate of its welding head or body structure, thus causing a systematic deviation in its perception and execution of the weld state left by the previous robot.

[0097] To ensure the stability and applicability of this positive correlation, the system defines a preset time interval as a constraint range for trend slope extraction. This time interval is generally set to a typical stage where heat accumulation shows continuous change, avoiding interference data from the initial stage before thermal equilibrium is established or the later stage when the temperature tends to saturate. Specific criteria include statistical results of dynamic parameters such as the total number of historical welded workpieces, the average welding time per piece, and the rate of change of the temperature rise rate at the thermally sensitive point. The system can adaptively adjust this interval according to the actual welding process to ensure the representativeness of the trend extraction.

[0098] Furthermore, the positive correlation conditions expressed by the slope ratio or multiple relationship are not empirically set, but rather quantifiable criteria derived from statistical analysis of a large amount of historical welding sample data. The system identifies typical coupling patterns between deviation trends and heat accumulation trends through offline regression modeling and cluster analysis, and extracts multiple stable slope intervals from the training set. These are further abstracted into permissible range expression functions or multiple mapping rules, ensuring that this "preset correlation" has traceability and statistical support, avoiding reliance on subjective human judgment in the technical solution.

[0099] The ultimate goal of this step is to identify the recurring trend that "post-robot heat accumulation significantly promotes welding deviation in the joint welding area," providing a dynamic basis for determining whether the angle correction value needs adjustment. This recurring judgment mechanism not only avoids the problem of insufficient adaptability of static correction values ​​in continuous multi-workpiece operations, but also enables the system to intelligently control the welding strategy based on changes in thermal state. This constitutes the core technical highlight and practical innovation foundation of this invention in the dynamic optimization of multi-robot collaborative welding trajectories.

[0100] The "welding deviation" described in this invention refers to the positional or posture deviation of the subsequent robot during the actual welding operation in the welding area compared to the expected weld path or weld geometry. Specifically, when the subsequent robot connects to the starting end of the weld left by the preceding robot, due to sensing errors, thermal expansion and contraction, or inaccurate posture control, the vertical angle, swing direction, or lateral / longitudinal positioning of its welding torch relative to the weld centerline may deviate. This results in problems such as discontinuous weld formation, weld bead offset, and uneven heat input. This type of deviation not only affects the consistency of the weld appearance but may also cause internal weld defects (such as lack of fusion, slag inclusions, or undercut), thereby reducing the overall strength and reliability of the welded joint. It is a crucial issue requiring real-time control in multi-robot continuous collaborative welding.

[0101] Furthermore, the 3D vision-based multi-robot collaborative welding trajectory planning method also includes the following steps:

[0102] Step S500: If the above-mentioned feature pattern is determined to exist, the initial angle correction value set by the robot for the joint welding area is obtained, and a correction factor is generated according to the trend of the deviation of the maximum change area of ​​the laser projection pattern of the front and rear robots, and the initial angle correction value is dynamically adjusted.

[0103] Specifically, Figure 4 A flowchart illustrating the correction of the initial angle correction value is shown.

[0104] If the aforementioned characteristic pattern is confirmed, the initial angle correction value set by the robot for the welding area is obtained, and a correction factor is generated based on the trend of the deviation of the maximum change area of ​​the laser projection pattern of the front and rear robots. The initial angle correction value is then dynamically adjusted, specifically including the following steps:

[0105] Step S501: After confirming the existence of the above-mentioned feature pattern, obtain the initial angle correction value set by the robot based on the joint welding area.

[0106] Step S502: Obtain the slope of the trend of the deviation of the maximum change area of ​​the laser projection pattern of the front and rear robots within a preset time interval: the second slope value;

[0107] Step S503: Set the second slope value as a correction factor, and correct the initial angle correction value using the correction factor to obtain the corrected angle correction value.

[0108] Step S504: The corrected angle correction value is applied to the welding trajectory planning of the robot in the handover welding area of ​​other workpieces to be welded in the target batch of workpieces, so as to achieve compensatory optimization of the initial angle correction value.

[0109] In this embodiment of the invention, the slope of the deviation trend of the region with the largest change in the laser projection pattern of the robot before and after (i.e., the second slope value) is selected as the correction factor because this slope can dynamically reflect the evolution rate of the difference between the laser patterns of the robot before and after the change under the influence of heat accumulation. Compared with static deviation values ​​or average values, the trend slope is more sensitive to capturing the evolution law of the gradual increase in welding error in continuous batches of workpieces, thereby realizing the forward-looking adjustment of the angle correction strategy. Using this slope as the correction factor has the following advantages: first, it can quantify the evolution intensity of the deviation; second, it can be correlated with time; and third, it can clearly capture the acceleration of error amplification, which helps to improve the adaptability and robustness of the correction strategy.

[0110] Besides using the trend slope as a correction factor, the following alternatives can be considered: First, use the standard deviation of the deviation within the sliding window as a measure of error volatility for correction; second, construct a nonlinear fitting function, analyze the fitting residuals between heat accumulation and deviation, and use the residual trend as a correction factor; third, use an empirical model or a small-sample learning model to predict error growth and generate a factor. However, compared to other methods, the trend slope still has advantages such as simple implementation, timely response, and low model dependence.

[0111] The calculation method for correcting the initial angle correction value can adopt the following linear amplification form:

[0112] ;

[0113] in, This refers to the corrected initial angle correction value. This refers to the initial angle correction value. This refers to the correction factor (i.e., the second slope value). This refers to a proportionality coefficient that is greater than 0. This method is simple to implement and facilitates rapid deployment within the system. Alternatively, an additive approach can also be used.

[0114] ;

[0115] Alternatively, a weighted fusion function expression can be constructed, but care must be taken to ensure that the corrected angle value remains within the physical boundaries under reasonable working conditions. During the correction process, directional constraints on the angle vector must be introduced to ensure that the angle adjustment direction is consistent with the actual weld offset trend, thereby avoiding over-correction or directional errors.

[0116] In step S504, the corrected angle correction value will serve as a key input parameter in the subsequent robot welding path generation module, specifically applied to the trajectory control logic of the subsequent robot in the handover welding area of ​​the target batch of subsequent workpieces. This trajectory planning generally includes control nodes such as welding torch posture adjustment, starting point offset, and transition section rotation angle setting. The correction value will directly affect the posture angle setting stage within the handover area, and through motion control commands, it will be sent to the subsequent robot end effector (such as a servo welding torch mechanism) for posture fine-tuning, ensuring that the weld trajectory more closely matches the structural features left by the end point of the previous robot's weld, thereby improving welding continuity and joint quality.

[0117] Furthermore, Figure 5 An application architecture diagram of the system provided in an embodiment of the present invention is shown.

[0118] In another preferred embodiment of the present invention, a welding trajectory planning system based on 3D vision and multi-robot collaboration includes:

[0119] The data acquisition module 100 is used to acquire historical welding data of a preset number of historical workpieces that have been welded before the target batch of workpieces, located in the welding area of ​​the dual robots.

[0120] Furthermore, the 3D vision-based multi-robot collaborative welding trajectory planning system also includes:

[0121] The maximum change area determination module 200 is used to extract the laser projection pattern in the handover welding area of ​​the front and rear robots corresponding to each historical workpiece from the historical welding data, and to determine the maximum change area of ​​each laser projection pattern. The laser projection pattern refers to the specific light spot or pattern projected onto the weld surface by the laser equipment carried by the robot during the welding process. It is used to help identify the weld contour, weld geometric changes and local deformation. The pattern will produce identifiable structural changes due to the thermal deformation of the workpiece or the change of surface reflection characteristics during the welding process.

[0122] Specifically, Figure 6 A structural block diagram of the maximum change region determination module 200 in the system provided by an embodiment of the present invention is shown.

[0123] In a preferred embodiment of the present invention, the maximum change region determination module 200 specifically includes:

[0124] The pattern extraction unit 201 is used to extract the laser projection patterns of the front and rear robots corresponding to each historical workpiece from historical welding data and projected onto the handover welding area.

[0125] The pattern change analysis unit 202 is used to perform structural reconstruction and curvature analysis on the extracted laser projection pattern using 3D vision technology, identify the region in the pattern that changes most significantly with the welding process, and determine that region as the corresponding region of maximum change.

[0126] Furthermore, the 3D vision-based multi-robot collaborative welding trajectory planning system also includes:

[0127] The deviation quantification module 300 is used to quantify the deviation of the area of ​​maximum change in the laser projection pattern of the front and rear robots in the junction area, and to obtain the thermal accumulation parameters of the rear robot as a preset number of historical workpieces over time.

[0128] Specifically, Figure 7 The diagram shows a structural block diagram of the deviation quantification module 300 in the system provided in an embodiment of the present invention.

[0129] In a preferred embodiment provided by the present invention, the deviation quantification module 300 specifically includes:

[0130] The coordinate extraction unit 301 is used to extract the spatial coordinates of the area with the maximum change in the laser projection pattern between the front robot and the rear robot in each historical workpiece based on 3D vision technology.

[0131] The deviation calculation unit 302 is used to calculate the deviation of the area of ​​maximum change in the laser projection pattern of the robot before and after in spatial position, and to use the deviation as the degree of deviation of the corresponding historical workpiece.

[0132] The thermal accumulation parameter acquisition unit 303 is used to acquire the post-robot thermal accumulation parameters when the welding of each historical workpiece is completed. The thermal accumulation parameters include the cumulative operation time during the welding process, the temperature change value of key parts, and their derived functions.

[0133] Furthermore, the 3D vision-based multi-robot collaborative welding trajectory planning system also includes:

[0134] The feature pattern judgment module 400 is used to analyze a preset number of historical workpieces and determine whether the following feature pattern exists: as time goes on, the deviation trend of the maximum change area of ​​the laser projection pattern of the front and back robots and the change trend of the thermal accumulation parameter of the back robot satisfy a preset correlation relationship.

[0135] The preset correlation refers to the following: in a preset number of historical workpieces, there is a positive correlation between the changing trend of the subsequent robot thermal accumulation parameters and the changing trend of the deviation of the maximum change area of ​​the laser projection pattern of the preceding and subsequent robots, and the positive correlation satisfies the following conditions: the slopes of the changing trends of the two within a preset time interval are extracted as a first slope value and a second slope value, respectively, and the ratio of the first slope value to the second slope value is within the allowable range expressed by the preset function, or satisfies the preset multiple relationship between the second slope value and the first slope value, thereby determining that the trend of the influence of the subsequent robot thermal accumulation on the welding deviation of the joint welding area conforms to the regularity requirement.

[0136] Furthermore, the 3D vision-based multi-robot collaborative welding trajectory planning system also includes:

[0137] The angle correction value adjustment module 500 is used to obtain the initial angle correction value set by the robot for the joint welding area if the above-mentioned feature pattern is determined to exist, and to generate a correction factor based on the trend of the deviation of the maximum change area of ​​the laser projection pattern of the front and rear robots, and dynamically adjust the initial angle correction value.

[0138] Specifically, Figure 8 A structural block diagram of the angle correction value adjustment module 500 in the system provided by an embodiment of the present invention is shown.

[0139] In a preferred embodiment of the present invention, the angle correction value adjustment module 500 specifically includes:

[0140] The initial value acquisition unit 501 is used to acquire the initial angle correction value set by the robot based on the joint welding area after determining that the above-mentioned feature pattern exists.

[0141] The slope value acquisition unit 502 is used to acquire the trend of the deviation of the maximum change area of ​​the laser projection pattern of the front and rear robots within a preset time interval: the second slope value;

[0142] Angle correction value adjustment unit 503 is used to set the second slope value as a correction factor, and correct the initial angle correction value through the correction factor to obtain the corrected angle correction value.

[0143] The correction value application unit 504 is used to apply the corrected angle correction value to the welding trajectory planning of the robot in the handover welding area of ​​other workpieces to be welded in the target batch of workpieces, so as to achieve compensatory optimization of the initial angle correction value.

[0144] It should be understood that although the steps in the flowcharts of the various embodiments of the present invention are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the various embodiments may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least a portion of the sub-steps or stages of other steps.

[0145] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0146] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0147] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

[0148] 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, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A welding trajectory planning method based on 3D vision and multi-robot collaboration, characterized in that, The method includes: Obtain historical welding data of a preset number of historical workpieces that have been welded before the target batch of workpieces, located in the dual-robot handover welding area; The laser projection patterns of the welding areas of the front and rear robots corresponding to each historical workpiece are extracted from the historical welding data, and the maximum change area of ​​each laser projection pattern is determined. The deviation of the area with the greatest change in the laser projection pattern of the robot before and after is quantified, and the thermal accumulation parameters of the robot after a preset number of historical workpieces over time are obtained. The steps include: Based on 3D vision technology, the spatial coordinates of the area with the greatest change in the laser projection pattern between the front robot and the rear robot in each historical workpiece are extracted. Calculate the spatial deviation of the area of ​​maximum change in the robot laser projection pattern before and after, and use this deviation as the degree of deviation of the corresponding historical workpiece; The robot acquires post-weld heat accumulation parameters for each historical workpiece at the time of welding completion. These heat accumulation parameters include the cumulative operation time during the welding process, the temperature change value of key parts, and their derived functions. Analyze a pre-set number of historical workpieces to determine if the following characteristic pattern exists: as time goes on, the deviation trend of the maximum change area of ​​the laser projection pattern of the front and back robots and the change trend of the thermal accumulation parameters of the back robot satisfy a pre-set correlation. The preset correlation refers to the following: in the previous preset number of historical workpieces, there is a positive correlation between the changing trend of the subsequent robot thermal accumulation parameters and the changing trend of the deviation of the maximum change area of ​​the laser projection pattern of the previous and subsequent robots, and the positive correlation satisfies the following conditions: the slopes of the changing trends of the two within the preset time interval are extracted as a first slope value and a second slope value, respectively, and the ratio of the first slope value to the second slope value is within the allowable range expressed by the preset function, or satisfies the preset multiple relationship between the second slope value and the first slope value, thereby judging that the trend of the influence of the subsequent robot thermal accumulation on the welding deviation of the joint welding area conforms to the regularity requirements; If the above feature pattern is confirmed, the initial angle correction value set by the robot for the welding area is obtained, and a correction factor is generated based on the trend of the deviation of the maximum change area of ​​the laser projection pattern of the front and rear robots, and the initial angle correction value is dynamically adjusted. The steps include: After confirming the existence of the above-mentioned feature patterns, the initial angle correction value set by the robot based on the joint welding area is obtained; The slope of the trend of the deviation of the area with the greatest change in the laser projection pattern of the robot before and after the acquisition within a preset time interval: the second slope value; The second slope value is set as the correction factor, and the initial angle correction value is corrected by the correction factor to obtain the corrected angle correction value. The corrected angle correction value is applied to the welding trajectory planning of the robot in the handover welding area of ​​other workpieces to be welded in the target batch of workpieces, so as to achieve compensatory optimization of the initial angle correction value.

2. The welding trajectory planning method based on 3D vision for multi-robot collaboration according to claim 1, characterized in that, The laser projection pattern refers to the pattern projected onto the weld surface by the laser equipment carried by the robot during the welding process. It is used to help identify the weld contour, weld geometry changes and local deformation. The pattern will produce identifiable structural changes during the welding process due to the thermal deformation of the workpiece or changes in the surface reflection characteristics.

3. The welding trajectory planning method based on 3D vision for multi-robot collaboration according to claim 2, characterized in that, The steps of extracting the laser projection pattern from the welding area of ​​the front and rear robots corresponding to each historical workpiece from historical welding data, and determining the area of ​​maximum change for each laser projection pattern, include: Extract the laser projection patterns of the front and rear robots corresponding to each historical workpiece from the historical welding data and project them onto the junction welding area; 3D vision technology was used to reconstruct the structure and analyze the curvature of the extracted laser projection pattern, identify the region in the pattern that changes most significantly with the welding process, and determine this region as the corresponding region of maximum change.

4. A welding trajectory planning system based on 3D vision and multi-robot collaboration, characterized in that, The system is used to execute the welding trajectory planning method based on 3D vision for multi-robot collaboration as described in any one of claims 1-3. The system includes: a data acquisition module, a maximum change area determination module, a deviation degree quantification module, a feature pattern judgment module, and an angle correction value adjustment module, wherein: The data acquisition module is used to acquire historical welding data of a preset number of historical workpieces that have been welded before the target batch of workpieces, located in the welding area of ​​the dual robots. The maximum change area determination module is used to extract the laser projection pattern in the welding area of ​​the front and rear robots corresponding to each historical workpiece from historical welding data, and to determine the maximum change area of ​​each laser projection pattern. The laser projection pattern refers to the pattern projected onto the weld surface by the laser equipment carried by the robot during the welding process. It is used to help identify the weld contour, weld geometric changes and local deformation. The pattern will produce identifiable structural changes during the welding process due to the thermal deformation of the workpiece or the change of surface reflection characteristics. The deviation quantification module is used to quantify the deviation of the area with the greatest change in the laser projection pattern of the front and rear robots in the junction area, and to obtain the thermal accumulation parameters of the rear robot as a preset number of historical workpieces extend over time. The feature pattern judgment module is used to analyze a preset number of historical workpieces and determine whether the following feature pattern exists: as time goes on, the deviation trend of the maximum change area of ​​the laser projection pattern of the front and back robots and the change trend of the thermal accumulation parameter of the back robot satisfy a preset correlation relationship. The preset correlation refers to the following: in the previous preset number of historical workpieces, there is a positive correlation between the changing trend of the subsequent robot thermal accumulation parameters and the changing trend of the deviation of the maximum change area of ​​the laser projection pattern of the previous and subsequent robots, and the positive correlation satisfies the following conditions: the slopes of the changing trends of the two within the preset time interval are extracted as a first slope value and a second slope value, respectively, and the ratio of the first slope value to the second slope value is within the allowable range expressed by the preset function, or satisfies the preset multiple relationship between the second slope value and the first slope value, thereby judging that the trend of the influence of the subsequent robot thermal accumulation on the welding deviation of the joint welding area conforms to the regularity requirements; The angle correction value adjustment module is used to obtain the initial angle correction value set by the robot for the joint welding area if the above feature pattern is determined to exist, and to generate a correction factor based on the trend of the deviation of the maximum change area of ​​the laser projection pattern of the front and rear robots, and dynamically adjust the initial angle correction value.

5. The welding trajectory planning system based on 3D vision for multi-robot collaboration according to claim 4, characterized in that, The module for determining the region of maximum change specifically includes: The pattern extraction unit is used to extract the laser projection patterns projected by the front and rear robots onto the junction welding area corresponding to each historical workpiece from historical welding data. The pattern change analysis unit is used to perform structural reconstruction and curvature analysis on the extracted laser projection pattern using 3D vision technology, identify the region in the pattern that changes most significantly with the welding process, and determine that region as the corresponding region of maximum change.

6. The welding trajectory planning system based on 3D vision for multi-robot collaboration according to claim 5, characterized in that, The deviation quantification module specifically includes: The coordinate extraction unit is used to extract the spatial coordinates of the area with the greatest change in the laser projection pattern between the front robot and the rear robot in each historical workpiece based on 3D vision technology. The deviation calculation unit is used to calculate the deviation of the area of ​​maximum change in the laser projection pattern of the robot before and after in terms of spatial position, and to use the deviation as the degree of deviation of the corresponding historical workpiece. The thermal accumulation parameter acquisition unit is used to acquire the post-robot thermal accumulation parameters when the welding of each historical workpiece is completed. The thermal accumulation parameters include the cumulative operation time during the welding process, the temperature change value of key parts, and their derived functions.

7. The welding trajectory planning system based on 3D vision for multi-robot collaboration according to claim 6, characterized in that, The angle correction value adjustment module specifically includes: The initial value acquisition unit is used to acquire the initial angle correction value set by the robot based on the handover welding area after determining that the above feature pattern exists; The slope value acquisition unit is used to acquire the trend of the deviation of the maximum change area of ​​the laser projection pattern of the robot before and after the change within a preset time interval: the slope value of the second slope value. Angle correction value adjustment unit is used to set the second slope value as a correction factor, and correct the initial angle correction value through the correction factor to obtain the corrected angle correction value; The correction value application unit is used to apply the corrected angle correction value to the welding trajectory planning of the robot in the handover welding area of ​​other workpieces to be welded in the target batch of workpieces, so as to achieve compensatory optimization of the initial angle correction value.

Images