Demonstration-free automatic welding method and system

Abstract

The invention provides a teaching-free automatic welding method and system. The system comprises a control center, a welding execution mechanism, a welding machine, a 3D sensor and a teaching-free control system. Wherein the control center is externally connected with the welding executing mechanism and the 3D sensor to obtain the state and observation data of the welding executing mechanism and the 3D sensor; the welding machine is an independent power supply and supplies power to the welding gun, and the welding execution mechanism controls the posture of the welding gun to complete the welding task. The 3D sensor is used for shooting a welding workpiece and obtaining point cloud information of the welding workpiece. The teaching-free control system is used for calculating the optimal shooting position of the welding seam, planning the welding path of the welding seam, optimizing the welding sequence and controlling the welding execution mechanism. The invention further provides a teaching-free automatic welding method, the method is divided into an offline stage and an online stage, and the two stages are matched with each other to achieve teaching-free automatic welding. According to the method, the manual teaching workload of the robot before customized welding can be reduced, the labor cost and the time cost are reduced, and the purposes of reducing cost and improving efficiency are achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of robot welding, and in particular to a teaching-free automatic welding method and system. Background Art

[0002] Since its birth, welding technology has always played an important role in industrial production and manufacturing. With the proposal and implementation of intelligent manufacturing and Industry 4.0 strategic plans, it is of great significance to realize the automation and intelligence of welding technology.

[0003] With the development of economy and society, welding robots are being used more and more widely and their technology is becoming more and more perfect. However, existing welding robots are still pre-programmed welding robots that require manual teaching. With the shortening of product iteration cycles and the increase in demand for small batch customized products, the existing teaching welding method requires a lot of time in the teaching stage and can no longer quickly adapt to customers' customized welding requirements. For this reason, a teaching-free automatic welding method and system are proposed. Summary of the invention

[0004] The object of the present invention is to provide a teaching-free automatic welding method and system.

[0005] In the first aspect, a teaching-free automatic welding system is provided, which is composed of a control center, a welding actuator, a welding machine, a 3D sensor, and a teaching-free control system.

[0006] The control center is connected to external peripherals such as robots and 3D sensors, and communicates with the peripherals to obtain their status and observation data; the control center is loaded with a teaching-free control system to process system status and collect environmental information.

[0007] The welding actuator consists of a welding robot and a welding gun. The welding actuator can control the position of the welding gun to complete the welding task.

[0008] The welding machine is an independent power supply, and its function is to directly power the welding gun.

[0009] The 3D sensor is installed at the end of the welding actuator through a fixture. The 3D sensor can capture the welding workpiece and obtain the corresponding point cloud information.

[0010] The teaching-free control system runs on the control center and can obtain the optimal shooting position of the weld, plan the welding path of the weld, optimize the welding sequence of the weld, and control the movement of the welding actuator.

[0011] In a second aspect, a teaching-free automatic welding method includes:

[0012] The offline simulation stage calculates the 3D sensor shooting posture and plans the optimal welding path and welding sequence.

[0013] During the online control stage, workpiece information is actually collected, weld features are located and extracted, welding paths are planned in real time, and the welding actuator is controlled to perform welding.

[0014] The robot simulation planning module is used to simulate the scene, optimize the 3D sensor shooting position and plan the robot welding trajectory.

[0015] The weld recognition module is used to locate the welding workpiece and extract the weld features.

[0016] Furthermore, a weld recognition module is loaded in the teaching-free control system. It includes: 3D point cloud preprocessing module, 3D point cloud model matching module, 3D point cloud weld recognition module, etc. The 3D point cloud preprocessing module is used to filter and downsample the point cloud, where the point cloud filtering algorithm includes: through filtering, voxel filtering, radius filtering, statistical filtering, geometric shape filtering and other algorithms; the point cloud downsampling algorithm includes: uniform downsampling, voxel downsampling, curvature downsampling and other algorithms; the 3D point cloud model recognition module directly matches the welding workpiece model with the 3D point cloud processed by the 3D point cloud preprocessing module, and its function is to locate the actual welding workpiece, determine the position of the welding workpiece, and use the model matching results to correct the workpiece and robot posture calculated in the offline simulation stage. At the same time, it is used to accurately extract the weld of the welding workpiece. Its function is to segment, extract features and calculate the actual photographed workpiece point cloud. Commonly used algorithms include: normal calculation, point cloud segmentation, curve fitting, and surface fitting.

[0017] Furthermore, a robot simulation planning module is loaded into the teaching-free control system, including a robot kinematics model, a robot motion planning module, and a robot simulation module. The robot kinematics model is used to calculate the robot forward and inverse solutions; the robot motion planning module is used to plan the robot motion path that meets certain constraints and has no collision with the environment; and the robot simulation module provides a planning environment for the robot motion planning module.

[0018] Furthermore, the teaching-free control system can be divided into an offline simulation stage and an online control stage. The offline simulation stage is based on the workpiece drawing and the workstation model, and obtains the photographing position and welding sequence of the weld through 3D simulation calculation. The online control stage runs under actual working conditions, locates and plans the weld in the actual scene based on the results of the offline simulation stage, and actually controls the welding actuator to weld.

[0019] Furthermore, the offline simulation stage includes the following steps:

[0020] S11. Data import and processing: import workpiece drawings, import workstation models, and mark welds on workpieces.

[0021] S12. Coordinate system binding: Bind the workpiece model and the workstation model to a unified coordinate system, and place the workpiece and workstation in a fixed position.

[0022] S13. Determine the robot's photographing position: determine the optimal position for the robot to photograph the weld according to the shape of the workpiece, the position of the weld in space, and the shape of the weld.

[0023] S14. Determine the welding sequence of the welds: determine the optimal sequence of welding the welds by the welding actuator according to the spatial distribution of the welds.

[0024] The online control phase includes the following steps:

[0025] S21. Select welding workpiece: select the workpiece model according to the actual welding workpiece.

[0026] S22. Model matching: The 3D sensor takes a photo of the welding workpiece and sends the point cloud data of the welding workpiece to the 3D point cloud model matching module for matching to obtain a rough positioning result, and then corrects the photographic pose generated in the offline simulation stage.

[0027] S23. Robot photography: The robot moves to the optimal photography position of the robot corrected in step 2 and takes a photo of the weld.

[0028] S24. Positioning the weld: Based on the weld image taken in step 3, the weld is identified and positioned using a 3D point cloud weld recognition module.

[0029] S25. Plan welding path: According to the weld position determined in step 4, determine the path of the robot welding the weld, and control the robot to perform welding.

[0030] In the offline simulation stage, the optimal shooting position of the robot needs to be determined through simulation. The following steps are included:

[0031] S31. Establishing a camera model through an algorithm;

[0032] S32. Using computer graphics to simulate and photograph the starting point and end point of the weld, and render images of each surface of the weld point;

[0033] S33. Optimize the projection objective function by using optimization algorithms such as genetic algorithm and ant colony algorithm to determine the best shooting position of the weld;

[0034] The offline simulation stage needs to determine the optimal welding sequence of the weld, including the following steps:

[0035] S41. Plan welding sequence: use optimization algorithms including but not limited to genetic algorithm, particle swarm algorithm and other optimization algorithms to plan the welding sequence of each robot weld.

[0036] S42. Robotic arm path planning: Use robot path planning algorithms including but not limited to RRT, artificial potential field method, etc. to plan the robot's motion trajectory.

[0037] S43. Collision detection: Use collision detection algorithm to detect collisions and impose penalties.

[0038] S44. Calculate time: Calculate the robot working time.

[0039] S45. Replanning: Replan the motion trajectory of the robot arm that caused the collision.

[0040] S46. Determine termination: Check whether the optimization stop condition is met. If so, stop; otherwise, return to S41 for a new round of optimization.

[0041] According to a third aspect, a computing device includes:

[0042] one or more processors;

[0043] The storage device is used to store one or more programs. When the one or more programs are executed by the one or more processors, the one or more processors implement the method described.

[0044] In a fourth aspect, a computer-readable storage medium stores a program, and when the program is executed by a processor, the method described is implemented.

[0045] The above solution of the present invention includes at least the following beneficial effects:

[0046] Solve the problem that the current welding system is not intelligent enough, requires a lot of manpower and time for manual teaching, and has a heavy workload for operators, and improve the working environment of operators.

[0047] Improve the debugging speed of customized products and increase enterprise output.

[0048] Reduce the demand for welding personnel in enterprises, lower the quality requirements for welding equipment operators, and lower the threshold for enterprises to recruit personnel. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 A structural diagram of a teaching-free automatic welding system provided in an embodiment of the present invention.

[0050] Figure 2 A flow chart of an offline simulation phase of a teaching-free automatic welding method and system provided in an embodiment of the present invention.

[0051] Figure 3 A flow chart of the online control phase of a teaching-free automatic welding method and system provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0052] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0053] like Figure 1 The present invention provides a teaching-free automatic welding system, which mainly includes: a control center 2, a welding actuator 1, a welding machine 4, a 3D sensor 3, and a teaching-free control system. Among them, the teaching-free control system written by the present invention is installed in the control center 2. The functions of the control system are divided into an offline simulation stage and an online control stage. The control system can plan and control the workflow of the entire welding system. The welding actuator 1 is a task execution module of the system, which can control the position of the welding gun to complete the welding task, and the welding machine 4 directly powers the welding gun. The 3D sensor 3 can take pictures of the welding workpiece and the weld, and obtain the point cloud information of the workpiece. The teaching-free control system can obtain the optimal shooting position of the weld, plan the welding path of the weld, optimize the welding sequence of the weld, and control the movement of the welding robot.

[0054] Exemplarily, the 3D sensor 3 is installed near the welding gun through a special fixture, and the sensor can be freely adjusted within the range of 0 to 60 degrees through the fixture.

[0055] Exemplarily, the present invention provides a teaching-free automatic welding method, wherein:

[0056] The teaching-free control system includes a weld recognition module and a robot simulation planning module.

[0057] The weld recognition module mainly includes: 3D point cloud preprocessing module and 3D point cloud model recognition module.

[0058] The 3D point cloud preprocessing module mainly includes the following functions: point cloud filtering, point cloud downsampling, outlier removal, etc. Among them, the point cloud filtering algorithm includes statistical filtering algorithm and geometric shape based filtering algorithm; the point cloud downsampling algorithm includes voxel downsampling and uniform downsampling algorithms; the outlier removal algorithm includes MAD algorithm, standard deviation method, etc.

[0059] The 3D point cloud model recognition module mainly includes the following functions: point cloud matching, point cloud segmentation, normal calculation, point cloud segmentation, curve fitting, surface fitting, etc. Among them, the point cloud matching algorithm includes ICP, PL-ICP, NICP and other algorithms, and the curve fitting and surface fitting algorithms include the least squares fitting algorithm.

[0060] The robot motion planning module mainly includes: robot kinematics model, which is used to calculate the robot kinematics forward and inverse solutions; robot motion planning algorithms, including RRT, RRT*, artificial potential field method and other robot path planning algorithms. The robot simulation module includes model import functions, including but not limited to AABB, SAT, GJK and other collision detection algorithms, providing a planning environment for the robot motion planning module.

[0061] Exemplarily, the teaching-free control system is divided into an offline simulation stage and an online control stage.

[0062] The input of the offline simulation stage is: the workpiece drawing with welds marked and the welding workstation; the output is: the optimal shooting position of the camera and the optimal welding sequence of the welds.

[0063] like Figure 2 As shown, the offline simulation stage specifically includes the following steps:

[0064] Step 1: Import the workpiece drawing and mark the weld on the workpiece model.

[0065] Step 2: The 3D module loads the welding workpiece model and the welding workstation model. The welding workstation model includes the installation relationship between the robot and the camera, and the robot forward and inverse solutions.

[0066] Step 3: Manually establish the workpiece coordinate system and associate the workpiece coordinate system with the workstation coordinate system. The software automatically translates and rotates the workpiece according to the association results.

[0067] Step 4: Generate an objective function whose optimization goal is to maximize the projection area of ​​the three faces and make the area ratio closest.

[0068] Step 5: Build a 3D sensor model and use computer graphics to render the weld start and end positions marked on the workpiece model

[0069] Step 6: Using the robot end position as the independent variable, solve the optimization goal in step 4 through heuristic algorithms such as genetic algorithms to determine the optimal shooting position of the camera.

[0070] Step 7: The optimization goal of planning the optimal welding sequence of the weld seam is to minimize the working time of the welding workstation and avoid collision of the robot arm.

[0071] Step 8: Use optimization algorithms such as genetic algorithm and particle swarm algorithm to plan the welding sequence of each robot weld, and use robot path planning algorithms such as RRT and artificial potential field method to plan the robot's motion trajectory.

[0072] Step 9: According to the robot motion trajectory determined in step 8, the collision detection algorithm is used to detect whether the robot arm has a collision, and the robot arm that has a collision is penalized, and finally the robot arm working time is determined. When the optimization termination condition is met, jump to step 11.

[0073] Step 10: Replan the motion trajectory of the robot arm that caused the collision to avoid the collision and return to step 9.

[0074] Step 11: Optimal welding sequence: The welding sequence of the welds in step 9 is determined as the welding sequence of the system.

[0075] The inputs of the online control stage are: actual welding workpiece point cloud, optimal camera shooting position, and optimal welding sequence of welds.

[0076] like Figure 3 As shown, the online control stage specifically includes the following steps:

[0077] Step 1: Select the workpiece model according to the actual workpiece to be welded.

[0078] Step 2: The 3D sensor takes a picture of the welding workpiece and sends the point cloud image and data to the 3D point cloud model matching module.

[0079] Step 3: The 3D point cloud model matching module performs point cloud registration on the workpiece model point cloud and the shooting point cloud to obtain the rotation and translation matrix between the two point clouds. According to the rotation and translation matrix, the optimal camera shooting position calculated in the offline simulation stage is corrected. According to the optimal welding sequence, welding tasks are generated and put into the welding task queue in sequence.

[0080] Step 4: The teaching-free control system selects the first task in the welding task queue, extracts the shooting tasks in turn, plans the motion trajectory from the robot's current position to the shooting position in real time, sends it to the robot body through the robot control module, and controls it to move to the shooting position.

[0081] Step 5: The 3D sensor takes a photo of the weld and sends the point cloud image of the weld to the 3D point cloud weld recognition module. The point cloud recognition module uses prior knowledge, such as the normal line near the weld, to segment the point cloud at the weld and fit the surface. It also calculates the intersection points and intersection lines of the surface as weld feature parameters.

[0082] Step 6: The robot planning module plans the welding path of the weld in real time, and sends the result to the robot body through the robot control module to control it to perform welding work.

[0083] Step 7: If the welding task queue is not empty, return to step 4, otherwise, end the welding work.

[0084] It should be noted that the device is a device corresponding to the above method, and all implementation methods in the above method embodiment are applicable to this embodiment and can achieve the same technical effect.

[0085] The embodiment of the present invention further provides a computing device, comprising: a processor, a memory storing a computer program, wherein when the computer program is executed by the processor, the method described above is executed. All implementations in the above method embodiment are applicable to this embodiment and can achieve the same technical effect.

[0086] The embodiment of the present invention also provides a computer-readable storage medium storing instructions, which, when executed on a computer, enable the computer to execute the method described above. All implementations in the above method embodiment are applicable to this embodiment and can achieve the same technical effect.

[0087] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0088] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0089] In the embodiments provided by the present invention, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0090] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0091] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0092] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard disks, ROM, RAM, magnetic disks, or optical disks.

[0093] In addition, it should be noted that in the apparatus and method of the present invention, it is obvious that each component or each step can be decomposed and / or recombined. These decompositions and / or recombinations should be regarded as equivalent schemes of the present invention. Moreover, the steps of performing the above-mentioned series of processing can naturally be performed in chronological order according to the order of description, but it is not necessary to perform them in chronological order, and some steps can be performed in parallel or independently of each other. For those of ordinary skill in the art, it is understood that all or any steps or components of the method and apparatus of the present invention can be implemented in any computing device (including processors, storage media, etc.) or a network of computing devices in hardware, firmware, software or a combination thereof, which can be achieved by those of ordinary skill in the art using their basic programming skills after reading the description of the present invention.

[0094] Therefore, the purpose of the present invention can also be achieved by running a program or a group of programs on any computing device. The computing device can be a well-known general device. Therefore, the purpose of the present invention can also be achieved by simply providing a program product containing a program code that implements the method or device. That is to say, such a program product also constitutes the present invention, and the storage medium storing such a program product also constitutes the present invention. Obviously, the storage medium can be any well-known storage medium or any storage medium developed in the future. It should also be pointed out that in the device and method of the present invention, it is obvious that each component or each step can be decomposed and / or recombined. These decompositions and / or recombinations should be regarded as equivalent schemes of the present invention. In addition, the steps of performing the above-mentioned series of processing can naturally be performed in chronological order according to the order of description, but it is not necessary to perform them in chronological order. Some steps can be performed in parallel or independently of each other.

[0095] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

[0096] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A teaching-free automatic welding system. It is characterized by The system consists of a control center, welding actuators, welding machines, 3D sensors, and a teaching-free control system. Among them, the control center is connected to external peripherals such as robots and 3D sensors, and communicates with the peripherals to obtain their status and observation data; the control center is loaded with a teaching-free control system to process system status and collect environmental information. Among them, the welding actuator is composed of a welding robot and a welding gun. The welding actuator can control the position and posture of the welding gun to complete the welding task. Among them, the welding machine is an independent power supply, and its function is to directly power the welding gun. The 3D sensor is installed at the end of the welding actuator through a fixture. The 3D sensor can capture the welding workpiece and obtain the corresponding point cloud information. A teaching-free automatic welding method, characterized in that It includes offline simulation stage and online control stage. Among them, the offline simulation stage is used to calculate the 3D sensor shooting posture, plan the optimal welding path and welding sequence. Among them, the online control stage is used to actually collect workpiece information, locate and extract weld features, plan the welding path in real time, and control the welding actuator to perform welding. The teaching-free automatic welding method according to claim 2 is characterized in that Includes robot simulation planning module and weld recognition module. Among them, the robot simulation planning module is used to simulate the scene, optimize the 3D sensor shooting position and plan the robot welding trajectory. Among them, the weld recognition module is used to locate the welding workpiece and extract the weld features. The teaching-free automatic welding method according to claim 3 is characterized in that The weld recognition module includes: a 3D point cloud preprocessing module and a 3D point cloud model recognition module. Among them, the 3D point cloud preprocessing module is used to perform filtering, downsampling and other processing on the point cloud. Among them, the 3D point cloud model recognition module is used to match, locate, segment, identify, and extract weld features of actual welding workpieces. The teaching-free automatic welding method according to claim 4 is characterized in that The robot simulation planning module includes a robot kinematics model, a robot motion planning module, and a robot simulation module. Among them, the robot kinematics model is used to calculate the robot forward and inverse solutions; Among them, the robot motion planning module is used to plan the robot motion path that meets certain constraints and has no collision with the environment; Among them, the robot simulation module provides a planning environment for the robot motion planning module. The teaching-free automatic welding method according to claim 5 is characterized in that The overall workflow can be divided into the offline simulation stage and the online control stage. The offline simulation stage is based on the workpiece and workstation model, and the robot simulation planning module and the weld recognition module calculate the optimal photography position and welding sequence of the weld. The online control stage runs under actual working conditions. Based on the results of the offline simulation stage, data collection, positioning and planning of welds in actual scenarios are performed, and the welding actuator is controlled to perform welding. The teaching-free automatic welding method according to claim 6 is characterized in that The offline simulation phase includes the following steps: S11. Data import and processing: import workpiece drawings, import workstation models, and mark welds on workpieces. S12. Coordinate system binding: Bind the workpiece model and the workstation model to a unified coordinate system, and place the workpiece and workstation in a fixed position. S13. Determine the robot's photographing position: determine the optimal position for the robot to photograph the weld according to the shape of the workpiece, the position of the weld in space, and the shape of the weld. S14. Determine the welding sequence of the welds: determine the optimal sequence of welding the welds by the welding actuator according to the spatial distribution of the welds. The teaching-free automatic welding method according to claim 7 is characterized in that The online control phase includes the following steps: S21. Select welding workpiece: select the workpiece model according to the actual welding workpiece. S22. Model matching: The 3D sensor takes a photo of the welding workpiece and sends the point cloud data of the welding workpiece to the 3D point cloud model matching module for matching to obtain a rough positioning result, and then corrects the photographic pose generated in the offline simulation stage. S23. Robot photography: The robot moves to the optimal photography position of the robot corrected in step 2 and takes a photo of the weld. S24. Positioning the weld: Based on the weld image taken in step 3, the weld is identified and positioned using a 3D point cloud weld recognition module. S25. Plan welding path: According to the weld position determined in step 4, determine the path of the robot welding the weld, and control the robot to perform welding. The teaching-free automatic welding method and system according to claim 8 is characterized in that In the offline simulation stage, the optimal shooting position of the robot is determined through simulation. The following steps are included: S31. Establishing a camera model through an algorithm; S32. Using computer graphics to simulate and photograph the starting point and end point of the weld, and render images of each surface of the weld point; S33. Optimize the projection objective function by using optimization algorithms such as genetic algorithm and ant colony algorithm to determine the best shooting position of the weld; The teaching-free automatic welding method according to claim 9 is characterized in that Determining the optimal welding sequence of the weld in the offline simulation stage includes the following steps: S41. Plan welding sequence: use optimization algorithms including but not limited to genetic algorithm, particle swarm algorithm and other optimization algorithms to plan the welding sequence of each robot weld. S42. Robotic arm path planning: The robot's motion trajectory is planned using robot path planning algorithms including but not limited to RRT and artificial potential field method. S43. Collision detection: Use collision detection algorithm to detect collisions and impose penalties. S44. Calculate time: Calculate the robot working time. S45. Replanning: Replan the motion trajectory of the robot arm that caused the collision. S46. Determine termination: Check whether the optimization stop condition is met, if so, stop, otherwise return to step 41 for a new round of optimization.

Images