Double-welding-robot automatic programming method based on digital twinning

Abstract

The invention provides an automatic programming method for double welding robots based on digital twinning. The double-welding-robot automatic programming method based on digital twinning specifically comprises the steps of twinning of a robot and a working scene, weld joint track optimization and automatic programming of robot motion codes. Acquiring a three-dimensional model of the robot and a workpiece, and establishing a twin space; adding set membership among joints in the robot model, and establishing a robot kinematics model; the workpiece is shot through a binocular camera, and welding seam position information is collected and processed; planning a double-robot multi-welding-seam task, extracting task parameters, establishing an algorithm operator, and carrying out iteration until a genetic algorithm is converged; a motion code of the robot is generated and debugged in the twin space, and if the debugging is passed, the motion code is sent to the entity robot; the method is convenient, fast, good in flexibility, low in requirement for the proficiency of operators, simple and easy to operate, and the efficiency and precision of welding operation can be improved.

Description

Technical Field

[0001] The present invention provides an automatic programming method for a welding robot based on digital twin, which specifically relates to the technical field of welding robot teaching. Background Art

[0002] Due to its high repeatability accuracy, fast working efficiency, stability and reliability, and the ability to perform welding work in extreme environments, welding robots have become an important means to reduce production costs and improve product competitiveness. Welding robots need to clarify the relative position and posture between the weld seam and itself in space through certain methods. The common methods can be divided into drag teaching, teach pendant teaching, and offline programming teaching.

[0003] Drag teaching is that workers perform teaching by dragging the end of the robot along the welding trajectory. Although this method can make the robot weld according to a fixed trajectory, due to the observation error of the operator and the joint damping of the robot itself, etc., this method cannot achieve high accuracy. In addition, this technology has poor adaptability to the external environment. When the environment is complex or the clamping error of the workpiece is large, greater deviation will occur.

[0004] Teach pendant teaching refers to controlling the robot through a teach pendant, walking along the required working trajectory and recording the trajectory, so as to complete the teaching. This industrial robot teaching programming method puts forward higher requirements for the programming level of practitioners, and there are problems of too many teaching points and long programming cycle for complex trajectory programming.

[0005] Offline programming teaching is a robot offline programming teaching technology based on computational graphics. By reconstructing the models of the robot and the working environment, the motion trajectory of the robot is calculated and planned offline for programming. Conventional offline programming is fast, but has poor adaptability. When the actual welding scenario changes, it is more troublesome to adjust and is prone to be out of touch with the actual situation. Summary of the Invention

[0006] In order to solve the deficiencies of the existing welding robot programming technology, the present invention provides an automatic programming method for a welding robot based on digital twin technology, which provides a more flexible and fast programming method while achieving high-precision acquisition of the spatial position of the weld seam.

[0007] To achieve the above object, the technical solution adopted by the present invention is: an automatic programming method for a dual welding robot based on digital twin. The specific steps are as follows:

[0008] Step S1: Install a binocular point cloud camera at the end of robot R1, calibrate the positional relationship between robot R1, robot R2, and the binocular point cloud camera, and obtain the homogeneous transformation matrix M1 from robot R2 to the base coordinates of robot R1 and the homogeneous transformation matrix M2 from the binocular camera to the base coordinates of robot R1.

[0009] Step S2: Use the D-H robot kinematic modeling method to establish the forward and inverse kinematic models of robot R1 and robot R2.

[0010] Step S3: Draw the three-dimensional models of robot R1, robot R2, the workpiece to be welded and related equipment, establish a digital twin programming system for the welding robot, import the drawn three-dimensional models, and restore the digital twin space of the welding robot.

[0011] Step S4: Obtain the workpiece point cloud data through a binocular point cloud camera, perform processing such as filtering, resampling, and segmentation on the data, extract the weld seam point cloud information, and apply the homogeneous transformation matrix M2 to convert the point cloud coordinates into the coordinates in the base coordinate system of robot R1.

[0012] Step S5: Twin the weld seam point cloud extracted in Step S4, perform operations such as B-spline curve fitting and equally spaced re-discretization on the weld seam to obtain discrete welding trajectory points.

[0013] Step S6: Perform robot welding task planning, and assign weld seam tasks to robot R1 and robot R2. According to the robot motion code format used and the discrete welding trajectory points obtained in Step S5, pre-give the welding process parameters required for the robot motion code, and automatically generate the robot motion code.

[0014] Step S7: Combine the robot kinematic model established in Step S2, calculate the rotation angles of the joint axes corresponding to the welding trajectory points, send them to the twin space for motion simulation, observe whether the welding task requirements can be met, and if not, repeat Steps S4 - S7.

[0015] Step S8: If the requirements are met, send the motion code generated in Step S6 to the physical robot to complete the digital twin automatic programming of the welding robot.

[0016] Among them, in the said Step S3, set the initial position of the coordinate system fixedly connected to the rotation axis according to the external dimensions of robot R1, and set the initial position of the coordinate system fixedly connected to the rotation axis according to the external dimensions of robot R2; further, import the joint models of robot R1 and robot R2 in sequence according to the rotation order; add the parent-child relationship of the robot joint axes and bind the angle variables to facilitate subsequent motion simulation through data in the twin.

[0017] Among them, in the said Step S6, the steps for optimizing the sorting of multiple welding tasks are:

[0018] S6-1. For a single-weld welding task, if the weld is only within the working space of robot R1 or robot R2, the weld is assigned to robot R1 or robot R2; if the weld is at the intersection of the working spaces of robot R1 and robot R2, the weld is divided at the intersection of the working spaces of robot R1 and robot R2 and the respective welds are assigned.

[0019] S6-2. For a multi-weld welding task, perform an optimal sorting of the two robots. The specific steps are as follows:

[0020] Extract the weld task assignment parameters. Constants: weld number sequence S, weld length L, initial point P of the robot end, robot running speed V, working space O occupied by robot R i Algorithm iteration times z. Variables: weld sequence U assigned to the robot i : i :

[0021] U i ={u i1 ,u i2 ,u i3 ,u i4 ,u i5 ,......,u ij}

[0022] In the formula, U i is the weld sequence assigned to the i-th robot, and u ij is the weld with the serial number j assigned to the i-th robot.

[0023] Welding direction sequence D of the welds assigned to the robot i :

[0024] D i ={d i1 ,d i2 ,d i3 ,d i4 ,d i5 ,......,d ij}

[0025] In the formula, D i is the welding direction sequence of the welds assigned to the i-th robot, and d ij is the welding direction of the weld with the serial number j assigned to the i-th robot. d ij ∈{0, 1}, where 0 represents that the welding robot welds from the starting point to the ending point of the weld, and 1 represents that the welding robot welds from the ending point to the starting point of the weld.

[0026] S6-3. Judge the relationship between the weld and the working space of each robot in turn. If it is exclusive to robot R iThe workspace will classify it into the weld sequence U i If it is in the common workspace of two robots, those with odd numbers will be assigned to robot R1, and those with even numbers will be assigned to robot R2. The initial robot-assigned weld sequence U is generated according to this rule. If the weld position exceeds the workspaces of both robots simultaneously, an error will be reported.

[0027] S6-4. Establish double-robot genetic algorithm operators, including a chromosome of the genetic algorithm representing weld attribution and sorting, and a chromosome of the genetic algorithm representing the welding direction of the weld; its structure is as Figure 7 shown;

[0028] S6-5. Randomly generate 8 groups of operators as the initial genetic population.

[0029] S6-6. Take the shortest total distance of the two robots as the fitness function of the genetic algorithm. The total distance W can be expressed by the formula:

[0030]

[0031] In the formula, l i represents the length of the i-th weld to be welded, represents the non-welding travel distance between the i-th weld and the next weld to be welded, represents the non-welding travel distance from the initial position of the robot end to the starting point of the first weld, represents the non-welding travel distance from the end point of the last weld back to the starting point of the robot end.

[0032] S6-7. Calculate the genetic fitness functions of the 8 groups of operators in the genetic population, and extract the optimal operator with the smallest fitness function value.

[0033] S6-8. Perform operations such as mutation and crossover on the seven genetic operators except the optimal operator.

[0034] The schematic diagram of the mutation of the operator sequence is as Figure 8 shown;

[0035] The schematic diagram of the mutation of the operator direction is as Figure 9 shown;

[0036] S6-9. Calculate the fitness function W of the new operator, and judge the magnitude of the fitness function between the new operator and the optimal operator. If the new operator is smaller, replace it; otherwise, discard it.

[0037] S6-10. Judge the number of algorithm iterations. If the number of algorithm iterations has reached the set number z, exit the iteration and output the weld allocation plan. Otherwise, return to S6-8 to perform mutation and crossover operations on the seven genetic operators except the optimal operator again.

[0038] The present invention has the following beneficial effects compared with the prior art:

[0039] 1. The automatic programming method for dual welding robots based on digital twin of the present invention has good economic benefits. By obtaining the weld positions through vision methods, it is simple and easy to operate. Compared with traditional manual teaching, the operators using the method of the present invention can easily operate the welding robot to complete tasks such as welding of workpieces without professional robot technology training, and the cost is more advantageous.

[0040] 2. The automatic programming method for dual welding robots based on digital twin of the present invention has relatively high overall safety. Compared with traditional manual dragging teaching, using the method of the present invention allows workers to complete the programming of the robot under the condition of being far away from the operation scene, away from toxic and harmful substances such as arc light and fumes generated during the welding process.

[0041] 3. The automatic programming method for dual welding robots based on digital twin of the present invention is convenient, fast and has high precision. Compared with traditional manual dragging teaching, the overall time used by the method of the present invention is shorter, and the observation errors in manual teaching are avoided, and the overall precision is higher. Description of the Drawings

[0042] Figure 1 is the system flow chart of the present invention.

[0043] Figure 2 is the schematic diagram of the hardware system device used in the present invention.

[0044] Figure 3 is the schematic diagram of the digital twin established by importing the robot model in the present invention.

[0045] Figure 4 is the schematic diagram of the twin space display interface, robot information and function operation interface established based on the digital twin concept in the present invention.

[0046] Figure 5 is the automatic programming function interface of the present invention.

[0047] Figure 6 is the schematic diagram of the weld seam extracted in the present invention.

[0048] Figure 7 is the schematic diagram of the chromosome of the genetic algorithm representing the welding direction of the weld seam in the present invention.

[0049] Figure 8 is the schematic diagram of the operator sequence mutation in the present invention.

[0050] Figure 9 is the schematic diagram of the operator direction mutation in the present invention.

[0051] Among them, 3.1-robot R1, 3.2-robot R2, 2-binocular point cloud camera, 1-host computer. DETAILED DESCRIPTION

[0052] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0053] The embodiments of the present invention are implemented based on the technical solutions of the present invention, and detailed implementation methods and specific operation processes are given, but the protection scope of the present invention is not limited to the following embodiments.

[0054] Reference Figure 2 The present invention provides a fast programming method based on digital twins. The hardware device used is shown in the figure, which mainly includes two programmable 3.1-robots R1 and 3.2-robots R2, 2-a binocular point cloud camera, and 1-a host computer. Figure 3 , 4 As described in , 5, and 6, the binocular vision module first takes a picture of the workpiece after calibration, then converts it into a point cloud file, transmits it to the host computer after filtering and sampling, and finally identifies the position of the weld through an algorithm and converts it into a readable code for the robot to use in the robot welding operation. The specific implementation principle flow chart is shown in Figure 1 shown.

[0055] S1. Install a binocular point cloud camera at the end of robot R1, calibrate the positional relationship between robot R1, robot R2 and the binocular point cloud camera, and obtain the homogeneous transformation matrix M1 from robot R2 to the base coordinates of robot R1, and the homogeneous transformation matrix M2 from the binocular camera to the base coordinates of robot R1.

[0056] S2. Use the DH robot kinematic modeling method to establish the forward and inverse kinematic models of robot R1 and robot R2.

[0057] S3. Draw the 3D models of robots R1 and R2, the workpieces to be welded, and related equipment, establish a digital twin programming system for the welding robot, import the drawn 3D models, and restore the digital twin space of the welding robot.

[0058] S4. Obtain workpiece point cloud data through a binocular point cloud camera, filter, resample, segment and process the data, extract weld point cloud information, and apply the homogeneous transformation matrix M2 to transform the point cloud coordinates into coordinates in the robot R1 base coordinate system.

[0059] S5. Twinning the weld point cloud extracted in step S4, performing B-spline curve fitting, equal-interval re-discretization and other operations on the weld to obtain discrete welding trajectory points.

[0060] S6. Perform robotic welding task planning, and assign welding seam tasks to robots R1 and R2. Based on the discrete welding trajectory points obtained according to the robotic motion code format used and step S5, pre-give the welding process parameters required for the robotic motion code, and automatically generate the robotic motion code.

[0061] S7. Combine the robotic kinematic model established in step S2, calculate the rotational angles of the joint axes corresponding to the welding trajectory points, send them to the twin space for motion simulation, observe whether the welding task requirements can be met, and if not, repeat steps S4 - S7.

[0062] S8. If the requirements are met, send the motion code generated in step S6 to the physical robot to complete the digital twin automatic programming of the welding robot.

[0063] When importing into the 3D model in step 2, it is necessary to disassemble the overall robot model into individual joint models, and fix a coordinate system at the rotational axis of each joint. Set the initial position of its coordinate system according to its external dimensions, import the joint models in sequence according to the rotational order, add parent - child relationships and bind variables, so as to control the motion of the twin - space robot through data in the twin.

[0064] Step S6 plans the multi - welding tasks of dual robots. First, extract the task parameters of the welding seam assignment, set the total distance of the welding trajectory as the genetic fitness function, establish algorithm operators, and then iterate until the genetic algorithm converges.

[0065] The following is the specific implementation process of the present invention:

[0066] An automatic programming method for robots based on digital twin of the present invention has a work process as Figure 1 shown, mainly including twin - space establishment, welding trajectory optimization and task sequencing, and automatic generation of robotic motion code.

[0067] Twin - space establishment. Use Visual studio software to build a twin - space based on the.net framework on the WPF (Windows Presentation Foundation) platform. The specific steps are as follows: Introduce the namespace Presentation3D in the xaml code of the WPF designer, import the 3D design function methods in the namespace for subsequent use. Use the MeshGeometry3D method to define 3D objects, and the Material and Light methods to define the material of the 3D model and the spatial lighting effect respectively. Use the PerspectiveCamera method to define the viewing angle of the overall interface. Add a Viewport3D control, and load the above elements into the control for displaying the 3D model.

[0068] Obtain 3D models of robots, workpieces, etc. If using robots of mainstream brands, they can be obtained from the official websites of these brands. If an accurate 3D model cannot be obtained, it can be established using Solidworks software based on their external dimensions. The obtained 3D model must be in the.stl format. If the model format does not match, it can be converted through Solidworks software. Import the 3D model into the twin space, and at the same time add attributes such as rotation axes and parent-child relationships, and bind data for subsequent motion simulation.

[0069] Weld task sequencing, including the following steps:

[0070] S1 For a single weld welding task, if the weld is only within the working space of robot R1 or robot R2, then assign the weld to robot R1 or robot R2; if the weld is at the intersection of the working spaces of robot R1 and robot R2, then use the intersection of the working spaces of robot R1 and robot R2 as the dividing point to assign their respective welds.

[0071] S2 For a multi-weld welding task, perform optimal sorting for dual robots. The specific steps are as follows:

[0072] Extract weld task allocation parameters. Constants: weld number sequence S, weld length L, initial point P of the robot end, robot running speed V, working space O occupied by robot R i , algorithm iteration times z. Variables: weld sequence U assigned to the robot i : i :

[0073] U i = {u i1 , u i2 , u i3 , u i4 , u i5 ,......, u ij}

[0074] In the formula, U i is the weld sequence assigned to the i-th robot, and u ij is the weld with the serial number j assigned to the i-th robot.

[0075] Welding direction sequence D of the welds assigned to the robot i :

[0076] D i = {d i1 , d i2 , d i3 , d i4 , d i5 ,......, d ij}

[0077] where D i is the welding direction sequence of the welds assigned to the i-th robot, and d ij is the welding direction of the weld with the serial number j assigned to the i-th robot. d ij ∈{0, 1}, where 0 means the welding robot welds from the starting point to the ending point of the weld, and 1 means the opposite.

[0078] S3 judges the relationship between the welds and the working spaces of each robot in turn. If it is uniquely within the working space of robot R i it is classified into the weld sequence U i . If it is in the common working space of two robots, those with odd numbers are assigned to robot R1, and those with even numbers are assigned to robot R2. The initial robot-assigned weld sequence U is generated according to this rule. If the weld position exceeds the working spaces of both robots at the same time, an error is reported.

[0079] S4 establishes a double-robot genetic algorithm operator, the structure of which is as Figure 7 shown;

[0080] S5 randomly generates 8 groups of operators as the initial genetic population.

[0081] S6 takes the shortest total distance of the double robots as the fitness function of the genetic algorithm. The total distance W can be expressed by the formula:

[0082]

[0083] where l i represents the length of the i-th weld to be welded, represents the non-welding travel distance between the i-th weld and the next weld to be welded, represents the non-welding travel distance from the initial position of the robot end to the starting point of the first weld, represents the non-welding travel distance from the ending point of the last weld back to the starting point of the robot end.

[0084] S7 calculates the genetic fitness functions of the 8 groups of operators in the genetic population, and extracts the optimal operator with the smallest fitness function value.

[0085] S8 performs operations such as mutation and crossover on the seven genetic operators except the optimal operator.

[0086] The schematic diagram of the operator sequence mutation is as Figure 8 shown;

[0087] The schematic diagram of the operator direction mutation is as Figure 9 shown;

[0088] S9 calculates the fitness function W of the new operator, and judges the magnitudes of the fitness functions of the new operator and the optimal operator. If the new operator is smaller, it is replaced; otherwise, it is discarded.

[0089] S10 determines the number of algorithm iterations. If the set number z has been reached, the iteration is exited and the weld joint allocation plan is output. Otherwise, return to S8.

[0090] The above are the preferred embodiments of the present invention. All changes made according to the technical solution of the present invention, when the functions and effects produced do not exceed the scope of the technical solution of the present invention, fall within the protection scope of the present invention.

Claims

1. A dual welding robot automatic programming method based on digital twin, characterized in that: The steps include: S1. Install a binocular point cloud camera at the end of robot R1, calibrate the positional relationship between robot R1, robot R2 and the binocular point cloud camera, and obtain the homogeneous transformation matrix M1 from robot R2 to the base coordinates of robot R1, and the homogeneous transformation matrix M2 from the binocular camera to the base coordinates of robot R1; S2, using the DH robot kinematics modeling method to establish the forward and inverse kinematics models of robot R1 and robot R2; S3. Draw the three-dimensional model of the welding robot physical workstation and workpiece, establish the welding robot digital twin programming system, import the drawn three-dimensional model, and restore the welding robot digital twin space; S3, draw the three-dimensional models of robot R1, robot R2 and the workpiece to be welded, establish a welding robot digital twin programming system, import the drawn three-dimensional model, and restore the welding robot digital twin space; S4, obtain the workpiece point cloud data through the binocular point cloud camera, filter, resample, segment and process the data, extract the weld point cloud information, and apply the homogeneous transformation matrix M2 to transform the point cloud coordinates into the coordinates of the robot R1 base coordinate system; S5, twinning the weld point cloud extracted in step S4, wherein the operation of twinning the weld point cloud includes performing B-spline curve fitting and equally spaced discretization on the weld to obtain discrete welding trajectory points; S6, perform robot welding task planning, and assign welding tasks to robot R1 and robot R2; The method includes presetting welding process parameters required for the robot motion code according to the robot motion code format used and the discrete welding trajectory points obtained in step S5, and automatically generating the robot motion code; S7, combining the robot kinematics model established in step S2, calculating the rotation angle of the joint axis corresponding to the welding trajectory point, sending it to the twin space for motion simulation, and observing whether the simulation result of the motion code can meet the welding task requirements. If it cannot meet the welding task requirements, repeat steps S4-S7; S8. When the simulation result of the motion code meets the requirements of the welding task, the motion code generated in step S6 is sent to the physical robot to complete the automatic programming of the welding robot digital twin.

2. According to claim 1, a dual welding robot automatic programming method based on digital twins is characterized in that: In step S3, before importing the three-dimensional models of robots R1 and R2, it is necessary to disassemble the three-dimensional model of robot R1 and the three-dimensional model of robot R2 into single joint models, and fix a coordinate system at the rotation axis of each joint of robot R1 and a coordinate system at the rotation axis of each joint of robot R2.

3. The automatic programming method of dual welding robots based on digital twins according to claim 2 is characterized in that: In step S3, the initial position of the coordinate system fixed at the rotation axis of the robot R1 is set according to the external dimensions of the robot R1, and the initial position of the coordinate system fixed at the rotation axis of the robot R2 is set according to the external dimensions of the robot R2; further, the joint model of the robot R1 and the joint model of the robot R2 are imported in sequence according to the rotation order; further, the parent-child relationship of the joint axis of the robot R1 is added and the angle variable is bound, and the parent-child relationship of the joint axis of the robot R2 is added and the angle variable is bound.

4. The automatic programming method of dual welding robots based on digital twins according to claim 1 is characterized in that: The extraction and allocation task of weld seam in step S6 includes allocating weld seam numbering sequences, establishing a dual-robot genetic algorithm operator, establishing a dual-robot genetic algorithm fitness function, and performing algorithm iterations.

5. The automatic programming method of dual welding robots based on digital twins according to claim 4 is characterized in that: Assigning a weld numbering sequence also includes the following: Step S6-1: For a single weld task, if the weld is only in the workspace of robot R1 or robot R2, the weld is assigned to robot R1 or robot R2; if the weld is at the intersection of the workspaces of robot R1 and robot R2, the intersection of the workspaces of robot R1 and robot R2 is used as the dividing point to assign the respective welds; Step S6-2: For multi-weld welding tasks, perform optimal dual-robot sorting. Establish the welding sequence U assigned to the robot i : IN i ={in i1 ,in i2 ,in i3 ,in i4 ,in i5 ,......,in ij } Where U i is the weld sequence assigned to the i-th robot, u ij The weld with serial number j assigned to the i-th robot; Establish the welding direction sequence D of the welds assigned by the robot i : D i ={d i1 ,d i2 ,d i3 ,d i4 ,d i5 ,......,d ij } Where D i is the welding direction sequence of the weld assigned to the i-th robot, d ij The welding direction of the weld with serial number j assigned to the i-th robot; d ij ∈{0,1}, where 0 represents that the welding robot welds from the starting point to the end point of the weld, and 1 represents that the welding robot welds from the end point to the starting point of the weld.

6. The automatic programming method of dual welding robots based on digital twins according to claim 5 is characterized in that: Assigning a weld numbering sequence also includes the following: Step S6-3: Determine the relationship between the weld and the workspace of each robot in turn. i The workspace is divided into the weld sequence U i ; When in the common workspace of two robots, those with odd numbers are assigned to robot R1, and those with even numbers are assigned to robot R2; according to this rule, the initial robot allocation weld sequence U is generated; if the weld position exceeds the workspace of both robots at the same time, an error is reported.

7. The automatic programming method of dual welding robots based on digital twins according to claim 4 is characterized in that: The establishment of the dual-robot genetic algorithm operator and the establishment of the dual-robot genetic algorithm fitness function include the following: Step S6-4: Establishing a dual-robot genetic algorithm operator, including a chromosome representing the attribution and sorting of welds and a chromosome representing the welding direction of welds; randomly generating 8 groups of operators as the initial genetic population; Step S6-6: Taking the shortest total distance of the two robots as the fitness function of the genetic algorithm; the total distance W can be expressed by the formula: Among them, l i represents the length of the i-th weld to be welded, It represents the idle distance between the ith weld and the next weld to be welded. It represents the distance traveled from the initial position of the robot end to the starting point of the first weld. Indicates the idle distance from the end point of the last weld to the starting point of the robot end.

8. The automatic programming method of dual welding robots based on digital twins according to claim 7 is characterized in that: The establishment of the dual-robot genetic algorithm operator and the establishment of the dual-robot genetic algorithm fitness function include the following: Step S6-7: Calculate the genetic fitness function of 8 groups of operators in the genetic population, and extract the optimal operator with the smallest fitness function value; Step S6-8: Perform mutation and crossover operations on the seven genetic operators except the optimal operator.

9. The automatic programming method of dual welding robots based on digital twins according to claim 8 is characterized in that: The algorithm iteration of the dual robot genetic algorithm fitness function includes the following: Step S6-9: Calculate the fitness function W of the new operator, and determine the size of the fitness function of the new operator and the optimal operator; when the fitness function of the new operator is smaller than the fitness function of the optimal operator, replace it; when the fitness function of the new operator is larger than the fitness function of the optimal operator, abandon the replacement.

10. The automatic programming method of dual welding robots based on digital twins according to claim 9 is characterized in that: The algorithm iteration of the dual robot genetic algorithm fitness function includes the following: Step S6-10: Determine the number of algorithm iterations; When the number of algorithm iterations reaches the set number z, the iteration is exited and the weld allocation plan is output; On the contrary, the seven genetic operators except the optimal operator are mutated and crossover operated again.

Images