Method and apparatus for applying the base coat of a welding robot

Abstract

This invention proposes a method and apparatus for root pass welding of a welding robot. Based on a weld seam recognition model, point cloud data of the weld seam in the robot arm coordinate system is identified. Welding trajectory points are generated based on the weld seam point cloud data. Gap information for each welding trajectory point is obtained based on the point cloud data of each welding trajectory point. Based on the maximum gap value in the gap information, a pre-designed welding process library is searched, and the root pass process corresponding to the maximum gap value is selected. Based on the root pass process and the welding trajectory points, the welding robot is controlled to execute the corresponding root pass welding flow. Therefore, by solving for the gap information of each welding trajectory point, adaptive matching of the root pass process is achieved, greatly improving the generalization ability of the welding robot, making it applicable to more non-standard scenarios, and improving welding efficiency.

Description

Technical Field

[0001] This invention relates to the field of welding technology, and in particular to a method and apparatus for applying the base coat using a welding robot. Background Technology

[0002] Currently, some workpieces to be welded are not processed to standard upstream, and the gaps between welds are not uniform. Some welds with larger gaps require manual root pass welding based on the weld width before the robot can weld, which is quite troublesome. In related technologies, without human intervention, it is impossible to determine the size of the weld gap. When the gap is large, the molten pool cannot fill the weld, requiring manual repair welding, which greatly reduces efficiency. Summary of the Invention The present invention aims to at least partially solve one of the technical problems in the related art.

[0003] Therefore, the first objective of this invention is to provide a method for the base coat of a welding robot.

[0004] The second objective of this invention is to provide a base-forming device for a welding robot.

[0005] The third objective of this invention is to provide an electronic device.

[0006] The fourth objective of this invention is to provide a non-transitory computer-readable storage medium storing computer instructions.

[0007] To achieve the above objectives, a first aspect of the present invention provides a method for applying the base coat using a welding robot, the method comprising: Based on the weld recognition model, the spatial position of the weld in the robot arm coordinate system is identified; Based on the spatial location of the weld, welding trajectory points are generated; Based on the point cloud structure around each welding trajectory point, the gap information of each welding trajectory point is obtained; Based on the maximum value of the gap in the gap information, search the pre-designed welding process library and select the root pass process corresponding to the maximum value of the gap; Based on the aforementioned root pass process and welding trajectory points, the welding robot is controlled to execute the corresponding root pass welding flow.

[0008] To achieve the above objectives, a second aspect of the present invention provides a base-forming device for a welding robot. The device includes: an identification module for identifying the spatial position of the weld in the robot arm coordinate system based on a weld identification model; a generation module for generating welding trajectory points based on the spatial position of the weld; a determination module for obtaining gap information for each welding trajectory point based on the point cloud structure around each welding trajectory point; a selection module for searching a pre-designed welding process library based on the maximum gap value in the gap information and selecting the base-forming process corresponding to the maximum gap value; and an execution module for controlling the welding robot to execute the corresponding base-forming welding flow based on the base-forming process and the welding trajectory points.

[0009] To achieve the above objectives, a third aspect of the present invention provides an electronic device comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the method described in the first aspect.

[0010] To achieve the above objectives, a fourth aspect of the present invention provides a non-transitory computer-readable storage medium storing computer instructions for causing the computer to perform the method described in the first aspect.

[0011] The welding robot's root pass welding method, apparatus, electronic device, and storage medium provided in this invention identify the point cloud data of the weld seam in the robotic arm coordinate system based on a weld seam recognition model; generate welding trajectory points based on the weld seam point cloud data; obtain gap information for each welding trajectory point based on the point cloud data of each welding trajectory point; search a pre-designed welding process library based on the maximum gap value in the gap information and select the root pass welding process corresponding to the maximum gap value; and control the welding robot to execute the corresponding root pass welding flow based on the root pass welding process and the welding trajectory points. Therefore, by solving for the gap information of each welding trajectory point, adaptive matching of the root pass welding process is achieved, greatly improving the generalization ability of the welding robot, making it applicable to more non-standard scenarios, and improving welding efficiency.

[0012] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0013] The above-described and / or additional aspects and advantages of the present invention will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein: Figure 1 This is a schematic flowchart of a welding robot's base-forming method provided in an embodiment of the present invention; Figure 2 This is a schematic diagram illustrating the principle of a weld seam recognition model provided in an embodiment of the present invention; Figure 3 An example diagram of gap information provided in an embodiment of the present invention; Figure 4 This is a schematic diagram illustrating the principle of a welding robot's base-forming method provided in an embodiment of the present invention. Figure 5 This is a schematic diagram of the structure of a welding robot's base-forming device provided in an embodiment of the present invention. Detailed Implementation

[0014] Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain the present invention, and should not be construed as limiting the present invention.

[0015] It should be noted that the acquisition, storage, use, and processing of data in the technical solution of this invention all comply with the relevant provisions of relevant laws and regulations.

[0016] The following description, with reference to the accompanying drawings, describes the welding robot's base-forming method, apparatus, electronic equipment, and storage medium according to embodiments of the present invention.

[0017] Figure 1 This is a schematic flowchart of a welding robot's base coat method provided in an embodiment of the present invention.

[0018] like Figure 1 As shown, the method includes the following steps: Step 101: Based on the weld recognition model, identify the spatial position of the weld in the robot arm coordinate system.

[0019] To achieve rapid and accurate acquisition of the spatial location of welds, as an example, grayscale images and point cloud data of historical welds can be used as inputs, and the spatial location of historical welds in the robot arm coordinate system can be used as outputs for model training to obtain a weld recognition model; based on the weld recognition model, the spatial location of the weld in the robot arm coordinate system can be identified.

[0020] In some possible implementations, the weld seam can be observed by a binocular structured light camera at the end of the robotic arm of the welding robot to obtain grayscale images and point cloud data of the observed weld seam.

[0021] It should be noted that, in order to clearly describe the weld identification model, this invention proposes a schematic diagram of the principle of the weld identification model, as shown below. Figure 2As shown, grayscale values ​​from the grayscale image and depth map (point cloud data) are input into the model. Then, the shape, position, and size information of the weld can be accurately extracted by using a mask that identifies individual weld objects. At the same time, classification and identification are performed by classifying weld instances to obtain the spatial position of different types of welds in the robotic arm coordinate system. The classification and identification include the classification of local welds formed in the spot welding process and linear welds formed by continuous welding, as well as the marking of the start or end position of the weld and the identification of the surface area after the weld is formed.

[0022] Step 102: Generate welding trajectory points based on the spatial location of the weld.

[0023] To meet the demands of automated welding, the welding trajectory points can be precisely controlled based on the spatial location of the weld. As an example, the endpoints and extension direction of the weld can be determined based on its spatial location. Based on the extension direction, an initial straight line is fitted at the endpoints of the weld, and multiple sample point clouds are collected along this direction. These sample point clouds are the point clouds in the weld's point cloud data that lie along the initial straight line. A plane is then determined that passes through the multiple sample point clouds and is perpendicular to the extension direction. Point clouds in the weld's point cloud data that lie both on the plane and on the corresponding two side plates are selected, and straight lines are fitted to both sides. The intersection of these two straight lines is taken as the weld point location. Welding trajectory points are generated based on the serialized weld point locations corresponding to each plane.

[0024] In some possible implementations, in order to improve efficiency by avoiding blind searching, the results of the positioning of the middle part of the weld (such as key points and direction vectors) are determined based on the spatial location of the weld. The endpoints and extension direction of the weld are then determined using the positioning results.

[0025] To ensure that no real endpoints are missed, a segment is taken on the initial straight line at the endpoint (the search range is set along the direction of the initial straight line, for example, extending 20mm forward / backward from the estimated position of the endpoint), and then a series of sample point clouds are sampled along the direction of the initial straight line in 1mm increments.

[0026] To filter out point clouds orthogonal to the extension direction, a series of sample point clouds are sampled, and a plane passing through the series of sample point clouds and perpendicular to the extension direction is calculated.

[0027] To eliminate noise interference and improve endpoint positioning accuracy, point clouds that are both on the plane and on the corresponding two sides of the weld can be selected from the point cloud data of the weld. The two straight lines are fitted, and the intersection of the two straight lines is taken as the weld point position. The two straight lines can be fitted using the Random Sample Consensus (RANSAC) algorithm.

[0028] In addition, the angle between the two straight lines can be used to determine whether the weld has been pushed to the end point along the weld direction, and the angle between the two straight lines can be used to judge whether the weld is complete, thus avoiding misjudgment of the end point due to local noise or missing data.

[0029] Step 103: Based on the point cloud structure around each welding trajectory point, obtain the gap information of each welding trajectory point.

[0030] To dynamically generate the optimal welding strategy, the gap information for each welding trajectory point needs to be accurately calculated. As an example, based on the point cloud structure around each welding trajectory point, the point cloud data can be divided into the weld centerline and the two side plate areas. The point cloud structure consists of point cloud data surrounding the weld and its surrounding area. Based on the point cloud data of the weld centerline and the two side plate areas, it is determined whether a gap exists in the weld. In response to the existence of a gap, at each welding trajectory point, the straight lines of the two side plates are fitted, and effective point cloud data is searched along the direction of the straight lines of the two side plates. The effective point cloud data includes point clouds whose distance from the straight lines of the two side plates is less than a set threshold. The target point cloud data corresponding to the intersection of the gap on the other side and the weld centerline and the welding trajectory point is determined. Based on the effective point cloud data and the target point cloud data, the gap information for each welding trajectory point is calculated.

[0031] In some possible implementations, clustering algorithms (such as DBSCAN) or region growing methods can be used to divide the point cloud data surrounding the weld and its surrounding area into the weld centerline and the two side plate regions; the conditions for the existence of gaps may include: the difference in distance between the point cloud data of the two side plates and the weld centerline exceeds a threshold (e.g., >1mm); the distribution width of the point cloud data of the weld centerline in the direction perpendicular to the weld is abnormal (e.g., >3mm).

[0032] To achieve effective point cloud data search, the point cloud data within a 20mm range along straight lines L1 and L2 can be searched starting from the intersection of the weld centerline and the welding trajectory point, and outlier point clouds (such as those with a straight distance greater than or equal to 1mm from the two side plates) can be excluded to obtain effective point cloud data.

[0033] To quickly find the target point cloud data corresponding to the intersection of the weld centerline and the welding trajectory point on the other side of the gap, it can be obtained through the symmetry constraint method. For example, if the gap is symmetrical, the effective point cloud data P1 searched along the L1 direction can be directly translated by d along the weld normal vector to obtain P2 (target point cloud data). If it is asymmetrical, the distance between the two plate surfaces needs to be fitted by the least squares method to calculate the optimal position of P2. In summary, taking the two plate surfaces as side plates and stiffeners as examples, the gap information (d) of the welding trajectory points on both sides can be obtained as follows: Figure 3 As shown, Figure 3 This is an example diagram of gap information provided in an embodiment of the present invention.

[0034] To quantify the gap information, the gap information can be calculated using the three-dimensional coordinates of corresponding points on both sides of P1 and P2.

[0035] Furthermore, when gap information is directly calculated for a welding trajectory point due to missing point cloud data, a reasonable estimate of gap information can be obtained by interpolating the gap information data of adjacent points.

[0036] Step 104: Based on the maximum value of the gap in the gap information, search the pre-designed welding process library and select the root pass process corresponding to the maximum value of the gap.

[0037] To ensure high-quality fusion and structural integrity in the most unfavorable (largest gap) area of ​​the weld joint, thereby improving the overall weld quality, as an example, the mapping relationship between different gap values ​​and welding processes in the welding process library can be determined; based on the mapping relationship, the root pass process corresponding to the maximum gap value can be found from the pre-designed welding process library.

[0038] As an example, a base coat process A can be set, including base coat processes with gaps of 0 to 3 mm, base coat processes with gaps of 3 to 6 mm, and so on. Then, based on the maximum value of the gap given by the algorithm, the corresponding process is matched from the process library. For example, if the maximum value of the gap is calculated to be 4.2 mm, then the 3 to 6 mm base coat process in process A will be matched.

[0039] Step 105: Based on the root pass process and welding trajectory points, control the welding robot to execute the corresponding root pass welding flow.

[0040] In some possible implementations, taking a robotic arm as an example, the robotic arm performs the initial welding and subsequent welding processes based on the initial welding process information, thereby achieving process adaptation, improving the generalization ability of the welding robotic arm, and making it applicable to more non-standard scenarios, improving efficiency while reducing human intervention.

[0041] The welding robot's root pass method according to this invention identifies the point cloud data of the weld seam in the robotic arm coordinate system based on a weld seam recognition model; generates welding trajectory points based on the weld seam point cloud data; obtains gap information for each welding trajectory point based on the point cloud data of each welding trajectory point; searches a pre-designed welding process library based on the maximum gap value in the gap information and selects the root pass process corresponding to the maximum gap value; and controls the welding robot to execute the corresponding root pass welding flow based on the root pass process and the welding trajectory points. Therefore, by solving for the gap information of each welding trajectory point, adaptive matching of the root pass process is achieved, greatly improving the generalization ability of the welding robot, making it applicable to more non-standard scenarios, and improving welding efficiency.

[0042] In summary, to clearly explain how automated and rapid base plating is achieved by welding robots, Figure 4This is a schematic diagram illustrating the principle of a welding robot's base-forming method according to an embodiment of the present invention. Specifically, a grayscale image and depth image (point cloud data) of the weld are acquired by a camera at the end of the robotic arm, and then input into a pre-trained weld recognition model to obtain the spatial position of the weld in the robotic arm coordinate system. Based on the spatial position of the weld, welding trajectory points are generated. Based on the point cloud structure around each welding trajectory point, the gap information of each welding trajectory point is obtained. The maximum value of the gap in the gap information is calculated, and a pre-designed welding process library is searched to select the base-forming process corresponding to the maximum value of the gap. Based on the base-forming process and the welding trajectory points, the welding robot is controlled to execute the corresponding base-forming welding process, and subsequent welding is performed after completion.

[0043] To achieve the above embodiments, the present invention also proposes a base-forming device for a welding robot.

[0044] Figure 5 This is a schematic diagram of the structure of a welding robot's base-forming device provided in an embodiment of the present invention.

[0045] like Figure 5 As shown, the welding robot's base-forming device 50 includes: an identification module 51, a generation module 52, a determination module 53, a selection module 54, and an execution module 55.

[0046] The system includes: an identification module 51 for identifying the spatial position of the weld in the robotic arm coordinate system based on the weld identification model; a generation module 52 for generating welding trajectory points based on the spatial position of the weld; a determination module 53 for obtaining the gap information of each welding trajectory point based on the point cloud structure around each welding trajectory point; a selection module 54 for searching a pre-designed welding process library based on the maximum gap value in the gap information and selecting the root pass process corresponding to the maximum gap value; and an execution module 55 for controlling the welding robot to execute the corresponding root pass welding flow based on the root pass process and the welding trajectory points.

[0047] Furthermore, in one possible implementation of this invention, the identification module 51 is specifically used to: take the grayscale image and point cloud data of the historical weld as input, and the spatial position of the historical weld in the robot arm coordinate system as output, perform model training to obtain a weld identification model; and identify the spatial position of the weld in the robot arm coordinate system according to the weld identification model.

[0048] Further, in one possible implementation of this invention, the generation module 52 is specifically used for: determining the endpoints and extension direction of the weld based on the spatial position of the weld; fitting an initial straight line at the endpoints of the weld based on the extension direction, and collecting multiple sample point clouds along the direction of the initial straight line; wherein, the sample point cloud is the point cloud in the weld point cloud data along the direction of the initial straight line; determining a plane that passes through the multiple sample point clouds and is perpendicular to the extension direction; selecting point clouds in the weld point cloud data that are both on the plane and on the corresponding two side plates of the weld, fitting straight lines on both sides, and taking the intersection of the two straight lines as the weld point position; generating welding trajectory points according to the serialized weld point positions corresponding to each plane.

[0049] Further, in one possible implementation of this invention, the determining module 53 is specifically used for: dividing the point cloud data into weld centerline and two side plate areas based on the point cloud structure around each welding trajectory point, wherein the point cloud structure is point cloud data surrounding the weld and its surrounding area; determining whether a gap exists in the weld based on the point cloud data of the weld centerline and the two side plate areas; in response to the existence of a gap in the weld, fitting the straight lines of the two side plates at each welding trajectory point, and searching for effective point cloud data along the direction of the straight lines of the two side plates, wherein the effective point cloud data includes point clouds whose distance from the straight lines of the two side plates is less than a set threshold; determining the target point cloud data corresponding to the intersection of the other side of the gap and the weld centerline and the welding trajectory point; and calculating the gap information for each welding trajectory point based on the effective point cloud data and the target point cloud data.

[0050] Furthermore, in one possible implementation of this invention, the selection module 54 is specifically used to: determine the mapping relationship between different gap values ​​and welding processes in the welding process library; and based on the mapping relationship, find the root pass process corresponding to the maximum gap value in the pre-designed welding process library.

[0051] It should be noted that the foregoing explanation of the method embodiments also applies to the apparatus of this embodiment, and will not be repeated here.

[0052] The welding robot's root pass device in this embodiment of the invention identifies the point cloud data of the weld seam in the robotic arm coordinate system based on a weld seam recognition model; generates welding trajectory points based on the weld seam point cloud data; obtains gap information for each welding trajectory point based on the point cloud data of each welding trajectory point; searches a pre-designed welding process library based on the maximum gap value in the gap information and selects the root pass process corresponding to the maximum gap value; and controls the welding robot to execute the corresponding root pass welding flow based on the root pass process and the welding trajectory points. Thus, by solving for the gap information of each welding trajectory point, adaptive matching of the root pass process is achieved, greatly improving the generalization ability of the welding robot, making it applicable to more non-standard scenarios, and improving welding efficiency.

[0053] To achieve the above embodiments, the present invention also proposes an electronic device, comprising: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the aforementioned method.

[0054] To implement the above embodiments, the present invention also proposes a non-transitory computer-readable storage medium storing computer instructions for causing the computer to perform the aforementioned method.

[0055] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0056] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this invention, "a plurality of" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0057] Any process or method description in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing custom logic functions or processes, and the scope of preferred embodiments of the invention includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as should be understood by those skilled in the art to which embodiments of the invention pertain.

[0058] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Alternatively, the computer-readable medium may be paper or other suitable media on which the program can be printed, since the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in a computer memory.

[0059] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware as in another embodiment, it can be implemented using any of the following techniques known in the art, or a combination thereof: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0060] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.

[0061] Furthermore, the functional units in the various embodiments of the present invention can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

[0062] The storage medium mentioned above can be a read-only memory, a disk, or an optical disk, etc. Although embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present invention.

Claims

1. A method for applying a base coat using a welding robot, characterized in that, The method includes: Based on the weld recognition model, the spatial position of the weld in the robot arm coordinate system is identified; Based on the spatial location of the weld, the endpoints and extension direction of the weld are determined; Based on the aforementioned extension direction, an initial straight line is fitted at the endpoint of the weld, and multiple sample point clouds are collected along the direction of the initial straight line; wherein, the sample point cloud is the point cloud along the direction of the initial straight line in the point cloud data of the weld. Determine a plane that passes through the multiple sample point clouds and is perpendicular to the extension direction; Select the point cloud data of the weld that is both on the plane and on the corresponding two sides of the plate, fit the straight lines on both sides, and take the intersection of the two straight lines as the weld point position. Based on the serialized weld point positions corresponding to each plane, welding trajectory points are generated; Based on the point cloud structure around each welding trajectory point, the point cloud data is divided into the weld centerline and the two side plate areas, wherein the point cloud structure is the point cloud data surrounding the weld and its surrounding area. Based on the point cloud data of the weld centerline and the two side plate areas, determine whether there is a gap in the weld. In response to the presence of gaps in the weld, at each welding trajectory point, the straight lines of the two side plates are fitted, and effective point cloud data is searched along the direction of the straight lines of the two side plates. The effective point cloud data includes point clouds whose distance from the straight lines of the two side plates is less than a set threshold. Determine the target point cloud data corresponding to the intersection of the other side of the gap and the weld centerline and the welding trajectory point; Based on the effective point cloud data and the target point cloud data, the gap information of each welding trajectory point is calculated; Based on the maximum value of the gap in the gap information, search the pre-designed welding process library and select the root pass process corresponding to the maximum value of the gap; Based on the aforementioned root pass process and welding trajectory points, the welding robot is controlled to execute the corresponding root pass welding flow.

2. The method according to claim 1, characterized in that, The step of identifying the spatial position of the weld in the robotic arm coordinate system based on the weld identification model includes: The grayscale image and point cloud data of historical welds are used as inputs, and the spatial position of the historical welds in the robot arm coordinate system is used as the output to train the model and obtain the weld recognition model. Based on the weld recognition model, the spatial position of the weld in the robot arm coordinate system is identified.

3. The method according to claim 1, characterized in that, The step of searching a pre-designed welding root pass process library based on the maximum value of the gap information and selecting the root pass process corresponding to the maximum value of the gap includes: Determine the mapping relationship between different gap values ​​in the welding process library and the welding processes; Based on the mapping relationship, the root pass process corresponding to the maximum gap value is found from the pre-designed welding process library.

4. A base-forming device for a welding robot, characterized in that, The device includes: The recognition module is used to identify the spatial position of the weld in the robot arm coordinate system based on the weld recognition model; The generation module is used to determine the endpoints and extension direction of the weld seam based on its spatial location; based on the extension direction, fit an initial straight line at the endpoints of the weld seam and collect multiple sample point clouds along the direction of the initial straight line; wherein, the sample point clouds are the point clouds in the weld seam point cloud data along the direction of the initial straight line; determine a plane that passes through the multiple sample point clouds and is perpendicular to the extension direction; select the point clouds in the weld seam point cloud data that are both on the plane and on the corresponding two side plates of the weld seam, fit the straight lines on both sides, and take the intersection of the two straight lines as the weld seam point position; generate welding trajectory points according to the serialized weld seam point positions corresponding to each plane. The determination module is used to divide the point cloud data into weld centerline and two side plate areas based on the point cloud structure around each welding trajectory point, wherein the point cloud structure is the point cloud data surrounding the weld and its surrounding area; based on the point cloud data of the weld centerline and the two side plate areas, determine whether there is a gap in the weld; in response to the existence of a gap in the weld, at each welding trajectory point, fit the straight lines of the two side plate areas, and search for effective point cloud data along the direction of the straight lines of the two side plate areas, wherein the effective point cloud data includes point clouds whose distance from the straight lines of the two side plate areas is less than a set threshold; determine the target point cloud data corresponding to the intersection of the other side of the gap and the weld centerline and the welding trajectory point; and calculate the gap information for each welding trajectory point based on the effective point cloud data and the target point cloud data. The selection module is used to search a pre-designed welding process library based on the maximum value of the gap in the gap information, and select the root pass process corresponding to the maximum value of the gap. The execution module is used to control the welding robot to execute the corresponding root welding flow based on the rooting process and welding trajectory points.

5. The apparatus according to claim 4, characterized in that, The identification module is specifically used for: The grayscale image and point cloud data of historical welds are used as inputs, and the spatial position of the historical welds in the robot arm coordinate system is used as the output to train the model and obtain the weld recognition model. Based on the weld recognition model, the spatial position of the weld in the robot arm coordinate system is identified.

6. The apparatus according to claim 4, characterized in that, The selection module is specifically used for: Determine the mapping relationship between different gap values ​​in the welding process library and the welding processes; Based on the mapping relationship, the root pass process corresponding to the maximum gap value is found from the pre-designed welding process library.

Images