A robotic intelligent welding method and system

Abstract

The application relates to the technical field of welding automation and robot visual perception, and discloses a robot intelligent welding method and system, wherein the method comprises the following steps: constructing a three-dimensional point cloud image by adopting a laser triangulation method; filtering and enhancing the point cloud image, and extracting a weld point cloud set by adopting a U-Net semantic segmentation model; identifying a space path and regressing a pose based on the set to obtain a position-pose pair trajectory primitive; fitting the path by adopting a cubic B-spline curve and generating a continuous welding trajectory by interpolation; and combining visual feedback in the welding process, introducing proportional differential control logic to compensate for closed-loop trajectory errors. Compared with the technical problem that the existing technology cannot stably realize weld path identification and accurate tracking under complex structures such as curved surfaces and variable cross sections, the application realizes high-precision automatic welding of complex weld structures by constructing a perception control closed loop based on visual servoing, and improves trajectory fitting precision.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of welding automation and robot vision perception, and particularly relates to a robot intelligent welding method and system. BACKGROUND

[0002] At present, robot automatic welding technology is widely used in automobile manufacturing, shipbuilding, engineering machinery and other fields, and is particularly suitable for high repeatability and regular structure welding seam scenes. However, with the development of industrial structure towards large-scale, complex and multi-curvature surface, the traditional welding robot still faces many technical bottlenecks when facing complex three-dimensional welding structure such as curved surface, stepped surface, pipe seat interface and special-shaped section.

[0003] The prior art mainly adopts preset trajectory or offline programming method for welding path planning, which seriously depends on the positioning accuracy of the workpiece, lacks the perception ability of the actual welding seam geometry, and is difficult to adapt to the welding seam drift caused by the change of welding seam position, the assembly deviation of the workpiece or the thermal deformation. In addition, some schemes use single frame image or fixed angle vision system to extract the welding seam edge, which cannot construct complete spatial information and is also difficult to obtain the tangent direction or attitude change characteristics of the welding seam.

[0004] In terms of path control, the common robot trajectory planning only considers the sequence of position points, and ignores the dynamic change of the end attitude, which leads to the inconsistency between the welding gun angle and the welding seam tangent, resulting in uneven welding seam penetration and decreased forming quality. In terms of control loop, the existing method is mainly position feedforward control, which lacks error correction ability based on real-time visual feedback, and cannot dynamically correct the error during welding.

[0005] Therefore, the existing technology cannot fully meet the high-precision, self-adaptive and closed-loop control welding requirements of complex three-dimensional structure welding seam. There is an urgent need for a robot intelligent welding method that can still realize three-dimensional welding seam identification, attitude estimation, trajectory generation and dynamic error compensation in complex welding seam structure, multi-attitude change and real-time disturbance environment, in order to improve the welding path fitting degree, welding consistency and product quality. SUMMARY

[0006] In view of the above technical deficiencies, the purpose of the present application is to provide a robot intelligent welding method, which aims to solve the technical problem that the existing technology cannot stably realize welding seam path identification and accurate tracking under complex structures such as curved surface and variable cross-section.

[0007] To solve the above technical problems, the present application adopts the following technical scheme: the present application provides a robot intelligent welding method,

[0008] The robot intelligent welding method comprises:

[0009] Step S10: scanning the target welding workpiece by a laser emitter and an industrial camera to obtain a weld area image, constructing a space mapping model according to the weld area image by using a laser triangulation method, and outputting a three-dimensional point cloud image of the target welding workpiece;

[0010] Step S20: sequentially performing Gaussian filter denoising processing and limited contrast self-adaptive histogram equalization processing on the three-dimensional point cloud image of the target welding workpiece to obtain a preprocessed point cloud image, performing pixel-level weld segmentation by using a pre-trained Python semantic segmentation model U-Net, and generating a weld point cloud set;

[0011] Step S30: performing space path structure recognition and pose recognition based on the weld point cloud set to obtain a position-pose pair trajectory primitive;

[0012] Step S40: fitting the position-pose pair trajectory primitive by using a cubic B-spline curve, generating a continuous executable welding trajectory path by using a pose interpolation method, and generating a welding instruction, wherein the robot reads the welding instruction and performs a welding operation;

[0013] Step S50: during the welding operation of the robot, performing closed-loop trajectory tracking control based on the welding trajectory path combined with a real-time detection result of the weld, and introducing proportional differential control logic to perform error compensation.

[0014] Preferably, in step S10, the step of scanning the target welding workpiece by the laser emitter and the industrial camera to obtain the weld area image, constructing the space mapping model according to the weld area image, and outputting the three-dimensional point cloud image of the target welding workpiece specifically comprises:

[0015] forming a laser line on the surface of the target welding workpiece by the laser emitter, and capturing the projection position of the laser line on the surface of the target welding workpiece by the industrial camera at a preset fixed frame rate to obtain the weld area image and the position of the laser line pixel;

[0016] performing depth inverse calculation on the parallax value d of the position of the laser line pixel in the weld area image by using the laser triangulation method to calculate the depth z value of the position of the laser line pixel in the preset camera coordinate system;

[0017] further mapping the position of the laser line pixel to the camera coordinate system by inverse projection in combination with a preset camera intrinsic matrix K, and obtaining the corresponding three-dimensional point cloud image of the target welding workpiece after distortion correction of the camera coordinate system.

[0018] Preferably, in step S30, the step of performing spatial path structure recognition and pose recognition based on the weld seam point cloud set to obtain the position-pose pair trajectory primitive, specifically comprises: predicting the pose direction vector of the weld i based on the weld seam point cloud set using a pose regression model constructed based on a ResNet or MobileNet backbone network And applying a rigid body transformation to map the weld seam point cloud set to a preset robot base coordinate system to obtain the position-pose pair trajectory primitive Wherein, p i is the three-dimensional coordinate position of the weld i in the robot base coordinate system.

[0019] Preferably, in step S40, the step of fitting the position-pose pair trajectory primitive using a cubic B-spline curve, the formula used is: Wherein, C(u) is the path fitting function, n is the number of weld seams, N i,k (u) is a B-spline basis function of order k, P i is the three-dimensional coordinate position of the weld i in the robot base coordinate system.

[0020] Preferably, in step S40, the step of generating a continuous executable welding trajectory path through pose interpolation, specifically comprises: for the pose direction vector Apply linear interpolation or Slerp spherical interpolation to generate a path pose function Sample the path fitting function C(u) and the pose function synchronously, construct a trajectory point sequence as a continuous executable welding trajectory path.

[0021] Preferably, in step S50, during the execution of the welding process, the step of performing closed-loop trajectory tracking control based on the welding trajectory path combined with the real-time detection results of the weld seam and introducing proportional-derivative control logic for error compensation, specifically comprises: during the execution of the welding process, the actual position of the weld j is obtained in real time through a camera and a sensor And calculate the current frame error vector e j ; using proportional-derivative control logic to calculate the trajectory correction amount, and error compensating the welding trajectory path according to the trajectory correction amount to obtain an optimized welding trajectory path.

[0022] Preferably, in step S50, the proportional-derivative control logic uses PD control method.

[0023] The application also provides a robot intelligent welding system comprising:

[0024] The three-dimensional mapping modeling module is used for obtaining a weld area image by scanning a target welding workpiece through a laser emitter and an industrial camera, constructing a space mapping model by using a laser triangulation method according to the weld area image, and outputting a three-dimensional point cloud image of the target welding workpiece.

[0025] The weld point cloud segmentation module is used for sequentially performing Gaussian filter denoising processing and limited contrast self-adaptive histogram equalization processing on the three-dimensional point cloud image of the target welding workpiece to obtain a pretreated point cloud image, performing pixel-level weld segmentation by using a pre-trained Python semantic segmentation model U-Net, and generating a weld point cloud set.

[0026] The posture recognition and trajectory primitive extraction module is used for performing space path structure recognition and posture recognition based on the weld point cloud set to obtain a position-posture pair trajectory primitive.

[0027] The welding trajectory fitting module is used for fitting a path of the position-posture pair trajectory primitive by using a cubic B-spline curve, generating a continuous executable welding trajectory path by using an attitude interpolation method, and generating a welding instruction, so that a robot reads the welding instruction and performs a welding operation.

[0028] The closed-loop control and error compensation module is used for performing closed-loop trajectory tracking control based on the welding trajectory path in combination with a real-time detection result of a weld during welding performed by the robot, and introducing proportional-differential control logic to perform error compensation.

[0029] The application further provides a computer program product, which comprises a robot intelligent welding program.

[0030] The application has the beneficial effects that the application implements high-precision recognition and posture estimation of structural welds such as complex curved surfaces and variable cross sections by constructing a three-dimensional visual detection method based on laser triangulation and deep learning, can dynamically extract weld position and tangent direction information, makes up for the deficiency of a traditional two-dimensional image detection method in the spatial structure perception dimension, and improves the path recognition accuracy of a robot in a complex welding environment.

[0031] The application introduces real-time visual detection and proportional-differential control closed-loop mechanisms during welding execution, dynamically adjusts a robot welding gun trajectory in combination with a weld detection result, can effectively suppress weld drift problems caused by workpiece assembly errors, thermal deformation or environmental disturbances, and thus improves weld path adhesion and weld forming consistency. BRIEF DESCRIPTION OF DRAWINGS

[0032] In order to make the technical solutions of the embodiments of the present application or the prior art clearer, the accompanying drawings needed in the embodiments or prior art description will be briefly introduced. Obviously, the accompanying drawings in the following description only aim to some embodiments of the present application, and all other embodiments obtained by those of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0033] Figure 1 A flowchart of a first embodiment of a robot intelligent welding method of the present application.

[0034] Figure 2 A structural diagram of a robot intelligent welding system of the present application

[0035] Figure 3 A device diagram of a robot intelligent welding method of the present application. DETAILED DESCRIPTION

[0036] The technical solutions in the embodiments of the present application will be described clearly and completely in combination with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only some embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0037] As shown in the drawings, Figure 1 in an embodiment of the present application, a flowchart of a robot intelligent welding method includes the following steps:

[0038] Step S10: obtaining a weld area image by scanning a target welding workpiece through a laser emitter and an industrial camera, constructing a space mapping model according to the weld area image by using a laser triangulation method, and outputting a three-dimensional point cloud image of the target welding workpiece;

[0039] Specifically, the step of obtaining a weld area image by scanning a target welding workpiece through an industrial camera in step S10, constructing a space mapping model according to the weld area image by using a laser triangulation method, and outputting a three-dimensional point cloud set of the weld, specifically includes: forming a laser line on the surface of the target welding workpiece through a laser emitter, and capturing the projection position of the laser line on the surface of the workpiece through an industrial camera at a preset fixed frame rate to obtain the weld area image and the position of the laser line pixel; performing depth inversion calculation on the parallax value d of the position of the laser line pixel in the weld area image by using the laser triangulation method, calculating the depth z value of the position of the laser line pixel in the preset camera coordinate system; and then combining the preset camera intrinsic matrix K, mapping the position of the laser line pixel to the camera coordinate system through back projection, and after distortion correction of the camera coordinate system, obtaining the corresponding three-dimensional point cloud image of the target welding workpiece.

[0040] Specifically, the laser triangulation system established by the laser emitter and the industrial camera in step S10 can obtain the three-dimensional depth information of the weld area in real time without contacting the workpiece, so that the robot has the spatial perception ability of the weld geometry profile, and is particularly suitable for complex structure scenes such as curved surface, variable cross-section and inclined weld, thereby improving the integrity and robustness of the weld detection.

[0041] Specifically, in step S10, the laser line pixel points are combined with the camera intrinsic matrix for back projection and distortion correction, which can effectively eliminate the positioning deviation caused by image edge distortion, and ensure the accuracy of the generated three-dimensional point cloud in geometric position, thereby providing high-precision visual detection input for subsequent weld recognition, path fitting and robot end posture control, and improving the accuracy of the whole weld path planning and the consistency of the weld forming.

[0042] Step S20: sequentially performing Gaussian filter denoising processing and limited contrast adaptive histogram equalization processing on the target weld workpiece three-dimensional point cloud image to obtain a preprocessed point cloud image, and using a pre-trained Python semantic segmentation model U-Net to perform pixel-level weld segmentation to generate a weld point cloud set;

[0043] Specifically, the above image preprocessing operation significantly improves the gray distribution difference between the weld area and the background, and enhances the discrimination ability of the segmentation model for the weld boundary, and is particularly suitable for low-contrast, unevenly illuminated or strong reflection interference weld site image environment.

[0044] Specifically, the semantic segmentation model U-Net can learn the typical shape, texture and spatial distribution features of the weld from a large number of labeled weld images through end-to-end training, so as to accurately segment the weld area on the complex workpiece surface; combined with the point cloud back projection and mask mapping relationship, the weld pixel area is finally restored to the weld point cloud set, providing accurate and stable input basis for subsequent path recognition and posture analysis.

[0045] Step S30: performing spatial path structure recognition and posture recognition based on the weld point cloud set to obtain a position-posture pair of trajectory primitives;

[0046] Specifically, the step of performing spatial path structure recognition and posture recognition based on the weld point cloud set in step S30 to obtain a position-posture pair of trajectory primitives, specifically includes: predicting the posture direction vector of the weld i based on the weld point cloud set using a posture regression model constructed based on a ResNet or MobileNet backbone network and mapping the weld point cloud set to a preset robot base coordinate system to obtain a position-posture pair of trajectory primitives wherein, p iThe three-dimensional coordinate position of the welding seam i in the robot base coordinate system.

[0047] The trajectory primitive of the position-attitude pair The precondition is that the local main direction of the welding seam path can be accurately identified by training on a large number of welding seam data sets, and stable direction estimation results can be provided even in areas with complex welding seam shapes or ambiguous boundaries; at the same time, the coordinate mapping method based on rigid body transformation is adopted to ensure that the welding seam establishes a unified spatial reference in the robot control system, avoiding the problems of path distortion or attitude mismatch.

[0048] By obtaining the position-attitude pair of the welding seam points, not only can a geometric continuity basis be provided for subsequent path fitting, but also the welding torch end attitude can be adjusted in real time according to the welding seam direction, realizing uniform welding penetration and angle matching process control, thereby significantly improving the welding quality and welding seam consistency, and being particularly suitable for complex welding seam structures with severe spatial attitude changes or multi-angle splicing.

[0049] Step S40: The trajectory primitive of the position-attitude pair is fitted with a cubic B-spline curve, and a continuous and executable welding trajectory path is generated through an attitude interpolation method to generate welding instructions, and the robot reads the welding instructions and performs welding operations;

[0050] Specifically, in step S40, the trajectory primitive of the position-attitude pair is fitted with a cubic B-spline curve, and the formula used is: Where C(u) is the path fitting function, n is the number of welding seams, N i,k (u) is the B-spline basis function of order k, P i is the three-dimensional coordinate position of the welding seam i in the robot base coordinate system. And the step of generating a continuous and executable welding trajectory path through an attitude interpolation method specifically includes: the attitude direction vector Apply linear interpolation or Slerp spherical interpolation to generate a path attitude function The path fitting function C(u) and the attitude function Synchronize sampling to construct a trajectory point sequence as a continuous and executable welding trajectory path.

[0051] The introduction of the attitude interpolation function can make the welding trajectory not only continuous and smooth in position, but also geometrically consistent in attitude direction, ensuring that the welding torch attitude always fits the welding seam trend. Especially when the welding seam has sharp turns, spatial transitions or height changes, the interpolated attitude path can effectively avoid the risk of welding torch angle jumps or attitude instability.

[0052] By interpolating the path position and orientation together and sampling synchronously to form a trajectory point sequence, an execution path with controllable smoothness, continuity and spatial consistency is constructed, providing complete pose input for subsequent robot welding control, and improving the accuracy of weld trajectory tracking, the accuracy of attitude response, as well as the stability and forming quality of the welding process.

[0053] Step S50: During the welding process, the robot performs closed-loop trajectory tracking control based on the welding trajectory path and the real-time detection results of the weld, and introduces proportional-derivative control logic for error compensation.

[0054] Specifically, the proportional-derivative control logic in step S50 adopts the PD control method.

[0055] By introducing real-time visual detection feedback and PD control compensation strategies, the system can dynamically respond to small offsets in the weld position during the welding process. Especially under non-ideal working conditions such as workpiece thermal deformation, assembly tolerance, or posture disturbance, it can still keep the tip of the welding torch running stably along the weld centerline, effectively preventing welding defects such as weld deviation and undercut.

[0056] The aforementioned closed-loop trajectory control mechanism constructs a real-time control chain of "weld visual inspection - error calculation - trajectory compensation," realizing perception-control coordination in robotic welding. Compared to the traditional open-loop execution path approach, this scheme significantly improves the weld path fit, the tracking accuracy of the robot's end effector, and the robustness of the welding process, ensuring the consistency and stability of welding quality under complex weld structures.

[0057] In addition, such as Figure 2 As shown, in one embodiment of the present invention, a robotic intelligent welding system is proposed, the robotic intelligent welding system comprising:

[0058] 3D mapping modeling module: It is used to scan the target welding workpiece to obtain the weld area image by using a laser emitter and industrial camera mounted on the robot, construct a spatial mapping model based on the weld area image using laser triangulation, and output a 3D point cloud image of the target welding workpiece.

[0059] Weld seam point cloud segmentation module: This module performs Gaussian filtering for noise reduction and contrast-limited adaptive histogram equalization on the 3D point cloud image of the target welded workpiece to obtain a preprocessed point cloud image. Then, it uses a pre-trained Python semantic segmentation model U-Net to perform pixel-level weld seam segmentation and generate a weld seam point cloud set.

[0060] Attitude recognition and trajectory primitive extraction module: used to perform spatial path structure recognition and attitude recognition based on weld seam point cloud set, and obtain trajectory primitives of position-attitude pairs;

[0061] The welding trajectory fitting module is configured to fit a path of a trajectory primitive based on a position-pose pair by using a cubic B-spline curve, and generate a continuous executable welding trajectory path by using a pose interpolation method, and the robot reads the welding trajectory path and performs a welding operation;

[0062] The closed-loop control and error compensation module is configured to perform closed-loop trajectory tracking control based on the welding trajectory path and a real-time detection result of a weld seam during the welding process of the robot, and introduce a proportional differential control logic to perform error compensation.

[0063] The robot intelligent welding system provided in the application can solve the technical problem of robot intelligent welding by using the robot intelligent welding method in the above embodiment. Compared with the prior art, the robot intelligent welding system provided in the application has the same beneficial effects as the robot intelligent welding method provided in the above embodiment, and other technical features in the robot intelligent welding system are the same as the features disclosed in the above embodiment method, which will not be repeated here.

[0064] As Figure 3As shown, in an embodiment of the present application, a structural diagram of a robot intelligent welding device suitable for implementing the embodiment of the present application is shown, which comprises: at least one processor; and a memory in communication connection with the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the robot intelligent welding method in the above-mentioned embodiment one. The robot intelligent welding device in the embodiment of the present application can include but is not limited to mobile terminals such as mobile phones, notebook computers, digital broadcast receivers, PDAs (Personal Digital Assistant), PADs (Portable Application Description), PMPs (Portable Media Player), vehicle terminals (such as vehicle navigation terminals), and the like, and fixed terminals such as digital TVs, desktop computers, and the like. The robot intelligent welding device is only an example, and should not bring any limitation to the function and use range of the embodiment of the present application. The robot intelligent welding device can include a processing device 1001 (such as a central processor, a graphics processor, etc.), which can perform various appropriate actions and processes according to programs stored in a read-only memory (ROM) 1002 or loaded from a storage device 1003 to a random access memory (RAM) 1004. In the RAM 1004, various programs and data required for the operation of the robot intelligent welding device are also stored. The processing device 1001, the ROM 1002, and the RAM 1004 are connected to each other through a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Generally, the following systems can be connected to the I / O interface 1006: an input device 1007 including, for example, a touch screen, a touch pad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, and the like; an output device 1008 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, and the like; a storage device 1003 including, for example, a magnetic tape, a hard disk, and the like; and a communication device 1009. The communication device 1009 can allow the robot intelligent welding device to communicate with other devices wirelessly or by wire to exchange data. Although a robot intelligent welding device with various systems is shown in the figure, it should be understood that it is not required to implement or have all the shown systems. More or fewer systems can be alternatively implemented or provided.

[0065] Further, the present application also provides a computer program product comprising a computer program which, when executed by a processor, implements the steps of a robot intelligent welding method as described above. The computer program product provided by the present application can solve the technical problem of a robot intelligent welding. Compared with the prior art, the computer program product provided by the present application has the same beneficial effects as the robot intelligent welding method provided by the above-mentioned embodiments, which will not be described here in detail.

[0066] In particular, according to the embodiments disclosed by the present application, the processes described above with reference to the flowcharts can be implemented as a computer software program. For example, the embodiments disclosed by the present application include a computer program product comprising a computer program carried on a computer readable medium, the computer program comprising program code for executing the method shown in the flowchart. In such embodiments, the computer program can be downloaded and installed from a network through a communication device, or installed from a storage device 1003, or installed from a ROM 1002. When the computer program is executed by a processing device 1001, the above-mentioned functions defined in the method of the embodiments disclosed by the present application are executed.

[0067] Parts of the present application can be implemented in hardware, software, firmware, or a combination thereof. In the description of the above-mentioned embodiments, specific features, structures, materials or characteristics can be combined in any one or more embodiments or examples in a suitable manner.

[0068] Obviously, those skilled in the art can make various modifications and variations to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application belong to the scope of the claims of the present application and their equivalent technologies, the present application also intends to include these modifications and variations.

Claims

1. A robotic intelligent welding method, characterized in that, The methods include: Step S10: Scan the target welding workpiece with a laser emitter and an industrial camera to obtain an image of the weld area. Based on the weld area image, construct a spatial mapping model using laser triangulation and output a three-dimensional point cloud image of the target welding workpiece. Step S20: Gaussian filtering and contrast-limited adaptive histogram equalization are performed sequentially on the 3D point cloud image of the target welded workpiece to obtain a preprocessed point cloud image. The pre-trained Python semantic segmentation model U-Net is used to perform pixel-level weld seam segmentation to generate a weld seam point cloud set. Step S30: Based on the weld seam point cloud set, perform spatial path structure recognition and attitude recognition to obtain trajectory primitives of position-attitude pairs; Step S40: The trajectory primitives based on the position-attitude pair are fitted with a cubic B-spline curve, and a continuous and executable welding trajectory path is generated by attitude interpolation and welding instructions are generated. The robot reads the welding instructions and executes the welding operation. Step S50: During the welding process, the robot performs closed-loop trajectory tracking control based on the welding trajectory path and the real-time detection results of the weld, and introduces proportional-derivative control logic for error compensation. In step S10, the steps of scanning the target welding workpiece with a laser emitter and an industrial camera to obtain an image of the weld area, constructing a spatial mapping model based on the weld area image using laser triangulation, and outputting a three-dimensional point cloud set of the target welding workpiece specifically include: A laser line is formed on the surface of the target welding workpiece by a laser emitter. An industrial camera captures the projection position of the laser line on the workpiece surface at a preset fixed frame rate, thus obtaining an image of the weld area and the position of the laser line pixels. The laser triangulation method is used to perform depth inverse calculation on the disparity value d of the laser line pixel position in the weld area image, and the depth z value of the laser line pixel position in the preset camera coordinate system is calculated. Then, by combining the preset camera intrinsic parameter matrix K, the position of the laser line pixel is mapped to the camera coordinate system through back projection, and after the distortion correction of the camera coordinate system is performed, the corresponding three-dimensional point cloud image of the target welding workpiece is obtained.

2. The robotic intelligent welding method as described in claim 1, characterized in that, Step S30, which involves spatial path structure recognition and attitude recognition based on the weld point cloud set to obtain trajectory primitives of position-attitude pairs, specifically includes: predicting the attitude direction vector of weld i using an attitude regression model built on a ResNet or MobileNet backbone network based on the weld point cloud set. Rigid body transformation is applied to map the weld point cloud set to a preset robot base coordinate system to obtain trajectory primitives of position-attitude pairs. ,in, Let i be the three-dimensional coordinate position of weld i in the robot's base coordinate system.

3. The robotic intelligent welding method as described in claim 1, characterized in that, In step S40, the path fitting process is performed on the trajectory primitives of the position-attitude pair using a cubic B-spline curve. The formula used is: ,in, This is the path fitting function. The number of welds. For order is B-spline basis functions, Let i be the three-dimensional coordinate position of weld i in the robot's base coordinate system.

4. The robotic intelligent welding method as described in claim 3, characterized in that, Step S40, which generates a continuous and executable welding trajectory path using attitude interpolation, specifically includes: adjusting the attitude direction vector in the trajectory primitives of the position-attitude pair. Path attitude functions are generated using linear interpolation or Slerp spherical interpolation. , path fitting function With attitude function Synchronous sampling is used to construct a sequence of trajectory points, which serves as a continuously executable welding trajectory path.

5. The robotic intelligent welding method as described in claim 1, characterized in that, In step S50, during the welding process, closed-loop trajectory tracking control is performed based on the welding trajectory path and the real-time weld detection results, and proportional-derivative control logic is introduced for error compensation. Specifically, this includes: during the welding process, the actual position of weld j is acquired in real time through a camera and sensors. And calculate the current frame error vector. The trajectory correction amount is calculated using proportional-derivative control logic, and the welding trajectory path is optimized by compensating for the welding trajectory path error based on the trajectory correction amount.

6. The robotic intelligent welding method as described in claim 1, characterized in that, In step S50, the proportional-derivative control logic adopts the PD control method.

7. A robotic intelligent welding system, applied to the robotic intelligent welding method according to any one of claims 1 to 6, characterized in that, The robotic intelligent welding system includes: 3D mapping modeling module: Used to scan the target welding workpiece with a laser emitter and an industrial camera to obtain the weld area image, construct a spatial mapping model based on the weld area image using laser triangulation, and output a 3D point cloud image of the target welding workpiece; Weld seam point cloud segmentation module: This module performs Gaussian filtering for noise reduction and contrast-limited adaptive histogram equalization on the 3D point cloud image of the target welded workpiece to obtain a preprocessed point cloud image. Then, it uses a pre-trained Python semantic segmentation model U-Net to perform pixel-level weld seam segmentation and generate a weld seam point cloud set. Attitude recognition and trajectory primitive extraction module: used to perform spatial path structure recognition and attitude recognition based on weld seam point cloud set, and obtain trajectory primitives of position-attitude pairs; Welding trajectory fitting module: It is used to fit the trajectory primitives based on position-attitude pairs with cubic B-spline curves, and generate a continuous and executable welding trajectory path and welding instructions through attitude interpolation. The robot reads the welding instructions and executes the welding operation. Closed-loop control and error compensation module: used for closed-loop trajectory tracking control of the robot during welding, based on the welding trajectory path and real-time weld detection results, and introducing proportional-derivative control logic for error compensation; The three-dimensional mapping modeling module includes the following steps: scanning the target welding workpiece with a laser emitter and an industrial camera to obtain an image of the weld area; constructing a spatial mapping model based on the weld area image using laser triangulation; and outputting a three-dimensional point cloud set of the target welding workpiece. A laser line is formed on the surface of the target welding workpiece by a laser emitter. An industrial camera captures the projection position of the laser line on the workpiece surface at a preset fixed frame rate, thus obtaining an image of the weld area and the position of the laser line pixels. The laser triangulation method is used to perform depth inverse calculation on the disparity value d of the laser line pixel position in the weld area image, and the depth z value of the laser line pixel position in the preset camera coordinate system is calculated. Then, by combining the preset camera intrinsic parameter matrix K, the position of the laser line pixel is mapped to the camera coordinate system through back projection, and after the distortion correction of the camera coordinate system is performed, the corresponding three-dimensional point cloud image of the target welding workpiece is obtained.

8. A robotic intelligent welding device, characterized in that, The robotic intelligent welding equipment includes: a memory, a processor, and a robotic intelligent welding program stored in the memory and executable on the processor. When the robotic intelligent welding program is executed by the processor, it implements a robotic intelligent welding method according to any one of claims 1 to 6.

9. A computer program product, characterized in that, The computer program product includes a robot intelligent welding program, which, when executed by a processor, implements a robot intelligent welding method according to any one of claims 1 to 6.

Images