A welding station and a welding process based on machine vision

Abstract

The application discloses a welding work station and a welding process based on machine vision and belongs to the technical field of welding. The technical scheme is as follows: the welding work station comprises a material table, a feeding robot and a welding workbench, and further comprises a visual fine positioning device; the welding workbench comprises a reference table top, and a lateral positioning mechanism is arranged on the reference table top; and a mobile visual device is arranged on a moving end of the feeding robot. According to the scheme, the mobile visual device is arranged on the feeding robot, the positioning precision during material taking is improved through the mobile visual device, the visual fine positioning device is used to accurately collect position information during workpiece conveying, the processing program or the feeding position is updated, and higher-precision processing is realized.

Description

Technical Field

[0001] This invention belongs to the field of welding technology, and in particular relates to a welding workstation and a welding process based on machine vision. Background Technology

[0002] Welding technology is currently one of the most commonly used industrial production technologies, with an extremely wide range of applications. While welding equipment performance has greatly improved in industrial production, the welding environment remains extremely harsh, posing certain risks to welders. Furthermore, with industrial development, precision welding is becoming increasingly stringent, placing higher demands on the technical skills of welders. Given the working environment and skill requirements, maintaining a high level of productivity for extended periods is difficult for workers. Therefore, the emergence of automated welding workstations has significantly replaced manual labor, offering extremely high precision and stability, and enabling continuous welding operations over long periods.

[0003] An automated welding workstation mainly consists of a welding robot, welding equipment (such as welding power source and welding torch), a control system, and other auxiliary equipment (such as workpiece fixtures and conveying equipment). It uses a computer system to intelligently sense the welding environment, weld seam tracking, and the dynamic welding process, and performs real-time tracking and control of various complex spatial curve weld seams based on the sensor information. In current automated welding workstations, the system controls the welding torch to move along a planned trajectory using preset programs and welding parameters, and performs real-time intelligent control of the welding dynamic process. Before welding, a computer software system for the welding path and welding parameters should be installed. The software plans and designs the continuous trajectory of the weld seam, the collision-free path of the welding motion, and the welding torch posture, and optimizes the welding parameters according to the welding process.

[0004] With the welding trajectory already optimized and standardized, improving the positioning accuracy of the workpiece in the workstation has become an important consideration for further improving welding accuracy. Summary of the Invention

[0005] This invention addresses current automated welding workstations and, in order to further improve welding accuracy, provides a welding workstation that uses a machine vision system to reposition the workpiece, resulting in higher welding precision.

[0006] To achieve the above objectives, the technical solution adopted by this invention is a welding workstation, including a material platform, a loading robot, and a welding worktable, and further including a vision-based precision positioning device; the welding worktable includes a reference platform with a lateral positioning mechanism on it; the moving end of the loading robot is equipped with a mobile vision device. This solution uses a mobile vision device on the loading robot to improve the positioning accuracy during material handling, and accurately collects position information during workpiece transport using the vision-based precision positioning device, updating the processing program or loading position to achieve higher precision processing.

[0007] Preferably, the lateral positioning mechanism includes a column, one side of which is a positioning surface and perpendicular to the reference platform. A clamping cylinder is provided at the upper end of the column, and a pressure plate is provided at the telescopic end of the clamping cylinder. The pressure plate is parallel to the positioning surface.

[0008] Preferably, two lateral positioning mechanisms are provided, and the positioning surfaces of the two mechanisms are coplanar. This achieves dual-point positioning of the workpiece and improves positioning stability.

[0009] Preferably, the system also includes a welding robot. Along the first direction, the material platform, the vision positioning device, the welding worktable, and the welding robot are arranged sequentially, with the loading robot positioned to one side of the vision positioning device. This optimizes the equipment layout and improves the workpiece turnover efficiency.

[0010] On the other hand, the present invention also provides a welding process based on machine vision, including the following steps:

[0011] S1. Write the processing program;

[0012] S2. Set the reference position of the motherboard, and achieve precise loading of the motherboard through a vision precision positioning device and a moving vision device;

[0013] S3. Set the reference position of the material, and realize the disordered feeding of the material through the vision precision positioning device and the moving vision device;

[0014] S4. Weld each material in sequence to complete the welding.

[0015] Preferably, in step S2:

[0016] Set the coordinate compensation amount (δ1,0,0) of the mother plate and the reference position L0 (X0,Y0,Z0) of the mother plate;

[0017] The loading robot places the mother plate in the initial position on the reference table of the welding workbench, and the lateral positioning mechanism on the reference table fixes the position of the mother plate in the Y direction in the horizontal plane.

[0018] The mobile vision device of the loading robot extracts the motherboard position data L2(X1,Y1,Z1);

[0019] The difference between L0 and L2 is used as an adjustment value to adjust the position of the motherboard and the coordinates of the motherboard in the industrial control system so that they coincide with the coordinates in the robot world.

[0020] Preferably, let:

[0021] δ0 is a fixed value for the welding process.

[0022] δ2=|(Y1-Y0) / Y0|

[0023] δ3=|(Z1-Z0) / Z0|

[0024] When δ2 and δ3 are greater than δ0, it indicates that there is an error in the clamping position of the mother plate on welding platform 6. Manual intervention is required to re-clamp the mother plate and repeat step S1.

[0025] When δ2 and δ3 are less than δ0, the difference δ1 = Z1 - Z0 is transmitted to the 7-level master control system, which then uses δ1 as the adjustment value.

[0026] Preferably, in step S3:

[0027] The mobile vision device on the feeding robot collects the position of each material:

[0028] According to the welding sequence, the feeding robot grabs the materials one by one, and the vision positioning device accurately identifies the grabbed materials and calculates the difference between the material and the origin of the vision positioning device in turn.

[0029] Adjust the variables in the machining program based on the difference;

[0030] The loading robot is based on the updated processing program to guide materials, load materials, and accurately position materials relative to the motherboard.

[0031] Preferably, the locations of each material are as follows:

[0032] L101(x 101 ,y 101 ,z 101 ,θ 101 )

[0033] L100(x 102 ,y 102 ,z 102 ,θ 102 ) ......

[0034] L1n(x 1n ,y 1n ,z 1n ,θ 1n ) ......

[0035] After the feeding robot grabs the material, it moves to the designated position L above the 4-vision precision positioning system. 21 (x 21 ,y 21 ,z 21 ), and make initial adjustments to material C during the movement. n The angle error;

[0036] The visual positioning system further identifies the material coordinates and extracts the material coordinate value L. 2n (x 2n ,y 2n ,z 2n ,θ 2n The material's current coordinates and the coordinates of the origin L of the vision-based precision positioning system. 20 (x 20 ,y 20 ,z 20 , θ 20 The difference is:

[0037] δ 21 = x 2n - x 20

[0038] δ 22 = y 2n - y 20

[0039] δ 23 = z 2n - z 20

[0040] δ 24 =θ 2n -θ 20 .

[0041] Preferably, when the visual precision positioning system performs the next recognition, the obtained difference data overwrites the difference data from the previous one.

[0042] As can be seen from the above technical solutions, the advantages of this invention lie in its intelligent welding workstation that achieves disordered material feeding and offline programming through material coordinate error compensation. Offline programming improves programming efficiency and expands its applicability. Coordinate error compensation enhances welding accuracy and expands the range of applications for welding. Attached Figure Description

[0043] To more clearly illustrate the technical solution of the present invention, the accompanying drawings used in the description will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0044] Figure 1 This is a schematic diagram of the structure of Embodiment 1 of the present invention.

[0045] Figure 2 This is a schematic diagram of the material platform in Embodiment 1 of the present invention.

[0046] Explanation of main figure symbols

[0047] 1. Material platform, 2. Loading robot, 3. Mobile vision device, 4. Vision positioning device, 5. End effector, 6. Welding workbench, 601. Reference table, 602. Lateral positioning mechanism, 7. Welding robot, 8. Industrial control system. Detailed Implementation

[0048] To make the objectives, features, and advantages of this invention more apparent and understandable, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings of the specific embodiments. Obviously, the embodiments described below are only some embodiments of this invention, and not all embodiments. Based on the embodiments of this patent, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this patent.

[0049] Example 1

[0050] like Figure 1 As shown, a welding workstation includes a material platform 1, a loading robot 2, a welding worktable 6, a welding robot 7, and a vision-based precision positioning device 4. Along a first direction, the material platform 1, the vision-based precision positioning device 4, the welding worktable 6, and the welding robot 7 are arranged sequentially. Along a second direction, the loading robot 2 is positioned to one side of the vision-based precision positioning device 4. The first and second directions are perpendicular to each other in the horizontal plane. Figure 2 As shown, the welding workbench 6 includes a reference table 601, on which a lateral positioning mechanism 602 is provided. The lateral positioning mechanism includes a column, one side of which is a positioning surface and is perpendicular to the reference table. A clamping cylinder is provided at the upper end of the column, and a pressure plate is provided at the telescopic end of the clamping cylinder. The pressure plate is parallel to the positioning surface. There are two lateral positioning mechanisms, and the positioning surfaces of the two positioning mechanisms are coplanar.

[0051] The loading robot 2 can adopt a multi-joint robotic arm structure, with an end effector 5 at its moving end. This is a conventional structure for loading robots. In addition, this solution also includes a mobile vision device 3 at the moving end. The mobile vision device 3 includes a vision camera and corresponding supplementary lighting. The visual positioning device also uses a vision camera. Furthermore, the material coordinate recognition accuracy of the two cameras in the mobile vision device 3 and the visual positioning device 4 is inversely proportional to the camera's performance and field of view. That is, for cameras with the same performance, a larger field of view means lower accuracy. Therefore, to improve the material positioning accuracy, the visual positioning device 4 adopts a fixed positioning and a small field of view to enhance material recognition accuracy. The mobile vision device 3 needs to handle both motherboard positioning and material recognition functions; therefore, local feature recognition is used for motherboard positioning. The motherboard is fixed in the Y and Z directions. The coordinate values ​​of feature points identified by the mobile vision device can be used to determine the positioning accuracy in the X direction, thereby improving the positioning accuracy in the X direction.

[0052] Example 2

[0053] Based on the workstation provided in Embodiment 1, this embodiment further provides a welding process based on machine vision:

[0054] Step 1: Write an offline welding program based on the intelligent welding workstation, motherboard, and materials. In the offline program, set two sets of compensation variables. By adjusting the values ​​of these compensation variables, achieve unordered material feeding and precise positioning of the motherboard.

[0055] The two sets of compensation variables are

[0056] Mother plate coordinate compensation (δ1,0,0)

[0057] Material coordinate compensation amount (δ) 21 ,δ 22 ,δ 23 ,δ 24 )

[0058] Step 2: Mark the welding positions on the motherboard according to the welding sequence.

[0059] Location numbers A1, A2, A3, ..., A n ...

[0060] Mark according to the welding sequence of the materials.

[0061] Material numbers C1, C2, C3, ..., C n ...

[0062] Step 3: The 602 lateral clamping mechanism clamps the mother plate under the control of the 603 welding station control system.

[0063] Step 4: The loading robot 2, carrying the mobile vision device 3, moves to the positioning position L1 above the welding worktable 6. The position features of the motherboard are extracted through the mobile vision device 3, and the position data L2(X1,Y1,Z1) of the motherboard features are extracted.

[0064] Comparing the feature position data L2(X1,Y1,Z1) with the offline data L0(X0,Y0,Z0), since the height of the welding workbench 6 and the lateral positioning position are fixed values, the errors of Y1 and Y0, and Z1 and Z0 should be within the specified values.

[0065] δ2=|(Y1-Y0) / Y0|

[0066] δ3=|(Z1-Z0) / Z0|

[0067] When δ2 and δ3 are greater than δ0 (the fixed value for welding process), it indicates that there is an error in the clamping position of the mother plate on the welding workbench 6. Manual intervention is required to re-clamp and repeat steps 3 and 4.

[0068] Step 5: When δ2 and δ3 are less than δ0 (the fixed value of the welding process), the difference δ1 = Z1 - Z0 is transmitted to the main control system 8. The main control system uses δ1 as an adjustment value to adjust the coordinate values ​​of the motherboard in the robot industrial control system and the robot industrial control system so that they coincide with the coordinates in the robot world.

[0069] Step 6: The loading robot 2, carrying the mobile vision device 3, moves to the positioning position L above the material platform 1. 11 Take photos sequentially along material platform 1 and identify the materials, then mark each material C. n Location

[0070] L101(x 101 ,y 101 ,z 101 ,θ 101 )

[0071] L100(x 102 ,y 102 ,z 102 ,θ 102 ) ......

[0072] L1n(x 1n ,y 1n ,z 1n ,θ 1n ) ......

[0073] Step 7: According to the welding sequence, the loading robot 2, carrying the end effector 5, sequentially grabs material C. nAnd move to the designated position L above the visual positioning device 4. 21 (x 21 ,y 21 ,z 21 ), and make initial adjustments to material C during the movement. n angular error (θ) 1n ).

[0074] Step 8, the visual precision positioning device 4 pairs with material C n The coordinates were precisely identified again to extract material C. n Coordinate value L 2n (x 2n ,y 2n ,z 2n ,θ 2n ), and calculate material C in sequence. n The current coordinates are consistent with the coordinates of the origin L of the visual precision positioning device 4. 20 (x 20 ,y 20 ,z 20 , θ 20 The difference

[0075] δ 21 = x 2n - x 20 ,

[0076] δ 22 = y 2n - y 20

[0077] δ 23 = z 2n - z 20

[0078] δ 24 =θ 2n -θ 20

[0079] The value is temporarily stored in the industrial control system of the vision positioning device, and the data is transmitted to the central control system, which then writes it into the variables of the offline programming program.

[0080] At the same time, the above values ​​will be overwritten during the next location identification.

[0081] Step 9: The loading robot 2 carries the material C grabbed by the end effector 5. n Material guidance, material loading, and precise material positioning are performed under the updated offline program.

[0082] Step 10, 7: The welding robot performs welding or riveting work under the updated offline program.

[0083] Step 11: The welding robot completes the welding and returns to its original position.

[0084] Steps 12 and 4: The actuator releases the material and returns to the original position of the feeding robot 2.

[0085] Step 13, proceed to the next cycle, grab material C. n+1 Repeat steps 6-12 until the material loading process is complete.

[0086] Step 14: The lateral positioning mechanism of the welding workbench 6 is released, completing the welding work.

[0087] As can be seen from the above embodiments, the beneficial effects of the present invention are as follows: a workstation for intelligent welding that achieves disordered material feeding and offline programming through material coordinate error compensation. Offline programming improves programming efficiency and applicability. Coordinate error compensation improves welding accuracy and the applicability of welding.

[0088] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A machine vision-based welding process, employing a welding workstation for welding, the welding workstation comprising: Material station (1), loading robot (2) and welding workstation (6); characterized in that the welding workstation further includes a visual precision positioning device (4); the welding workstation (6) includes a reference table (601), on which a lateral positioning mechanism (602) is provided; the moving end of the loading robot (2) is provided with a moving vision device; The welding process includes the following steps: S1. Write the processing program; S2. Set the reference position of the motherboard and achieve precise feeding of the motherboard by moving the vision device; S3. Set the reference position of the material and realize the disordered feeding of the material through the visual precision positioning device (4) and the moving vision device; S4. Weld each material in sequence to complete the welding process; In step S2: Set the coordinate compensation amount (δ1,0,0) of the mother plate and the reference position L0 (X0,Y0,Z0) of the mother plate; The loading robot (2) places the mother plate in the initial position on the reference table of the welding workbench, and the lateral positioning mechanism on the reference table fixes the position of the mother plate in the Y direction in the horizontal plane. The mobile vision device of the loading robot (2) extracts the motherboard position data L2(X1,Y1,Z1); The difference between L0 and L2 is used as an adjustment value to adjust the position of the motherboard and the coordinates of the motherboard in the industrial control system so that they coincide with the coordinates in the robot world. set up: δ0 is a fixed value for the welding process. δ2=|(Y1-Y0) / Y0| δ3=|(Z1-Z0) / Z0| When δ2 and δ3 are greater than δ0, it indicates that the welding platform is clamping the mother plate incorrectly, and manual intervention is required to re-clamp the mother plate and repeat step S1. When δ2 and δ3 are less than δ0, the difference δ1 = Z1 - Z0 is transmitted to the central control system, which uses δ1 as the adjustment value. In step S3: The mobile vision device on the feeding robot collects the position of each material: According to the welding sequence, the feeding robot grabs the materials one by one, and the vision positioning device accurately identifies the grabbed materials and calculates the difference between the material and the origin of the vision positioning device in turn. Adjust the variables in the machining program based on the difference; The loading robot is based on the updated processing program to guide materials, load materials, and accurately position materials relative to the motherboard.

2. The machine vision-based welding process according to claim 1, characterized in that: The locations of each material are as follows: L101(x 101 ,y 101 ,z 101 ,θ 101 ) L100(x 102 ,y 102 ,z 102 ,i 102 ) ...... L1n(x 1n ,y 1n ,z 1n ,θ 1n ) ...... After the feeding robot grabs the material, it moves to the designated position L above the vision positioning device. 21 (x 21 ,y 21 ,z 21 ), and make initial adjustments to material C during the movement. n angular error θ 1n ; The visual positioning device accurately identifies the material coordinates again and extracts the material coordinate value L. 2n (x 2n ,y 2n ,z 2n ,θ 2n The material's current coordinates and the coordinates of the origin L of the vision precision positioning device. 20 (x 20 ,y 20 ,z 20 , θ 20 The difference is: δ 21 = x 2n - x 20 δ 22 = and 2n - and 20 δ 23 = with 2n - With 20 d 24 =θ 2n -θ 20 。 3. The machine vision-based welding process according to claim 1, characterized in that, When the visual positioning device performs the next identification, the difference data between the obtained material and the origin of the visual positioning device overwrites the difference data of the previous one.

4. The machine vision-based welding process according to claim 1, characterized in that, The lateral positioning mechanism includes a column, one side of which is a positioning surface and perpendicular to the reference platform. A clamping cylinder is provided at the upper end of the column, and a pressure plate is provided at the telescopic end of the clamping cylinder. The pressure plate is parallel to the positioning surface.

5. The machine vision-based welding process according to claim 4, characterized in that, There are two lateral positioning mechanisms, and the positioning surfaces of the two lateral positioning mechanisms are coplanar.

6. The machine vision-based welding process according to claim 1, characterized in that, It also includes a welding robot (7). Along the first direction, the material table (1), the vision positioning device (4), the welding workbench (6) and the welding robot (7) are arranged in sequence. Along the second direction, the loading robot (2) is arranged on one side of the vision positioning device (4). The first direction and the second direction are perpendicular to each other in the horizontal plane.

Images