Double-robot collaborative welding dynamic compensation system

Abstract

The invention provides a double-robot collaborative welding dynamic compensation system which is characterized by comprising two welding robots slidably mounted on two ground rails, a six-axis laser welding seam locating device is mounted on each welding robot, and a machine vision assembly is arranged in each laser welding seam locating device. A stand column cantilever used for driving the welding robot to move is installed on the ground rails in a sliding mode, a rotary table position changing machine is arranged between the two ground rails, a powder concentrator rotor is placed on the rotary table position changing machine, and a control system is electrically arranged beside the welding robot. The double robots cooperatively work through the laser distance measuring sensor and the visual assembly, rapid concentric alignment of a rotor, blade angle calibration and real-time tracking of a welding seam are achieved, high-precision positioning of the laser distance measuring instrument is combined with intelligent image processing of the visual module, the accuracy of welding path planning is remarkably improved, deviation caused by manual intervention is avoided, and the working efficiency is improved. The two robots cooperate with each other, so that the welding efficiency is improved by more than 40% compared with a single-machine system.

Description

Technical Field

[0001] The present invention relates to the field of intelligent manufacturing technology, and in particular to a dual-robot collaborative welding dynamic compensation system. Background Art

[0002] In the intelligent manufacturing powder classifier rotor welding scenario, the powder classifier rotor is the core component of the powder classifier equipment, and its performance directly affects the material classification efficiency and equipment operation stability. In the intelligent manufacturing powder classifier rotor welding scenario, when a single welding robot directly completes the welding work in the automatic welding production line, it is easier to control the working order and welding path of the welding robot. However, in a production line with a large workpiece length and a large number of welds to be welded, only a single welding robot is used for welding, which has low work efficiency and cannot reflect the high efficiency of automated production. When the existing dual welding robots work together, they are prone to interlacing when welding at similar weld positions, making the welded rotor not fully applicable. In order to solve the above problems, a dual robot collaborative welding dynamic compensation system is proposed. Summary of the invention

[0003] According to the above existing technical problems, the present invention provides a dual-robot collaborative welding dynamic compensation system, characterized in that it comprises two welding robots slidably mounted on two sets of ground rails, a six-axis laser weld positioner is mounted on the welding robot, a machine vision component is included in the laser welding positioner, a column cantilever for driving the welding robot to move is slidably mounted on the ground rail, a turntable positioner is provided between the two ground rails, a powder selector rotor is placed on the turntable positioner, and a control system is electrically arranged beside the welding robot;

[0004] The machine vision component includes a visual image acquisition module, a weld detection module and a welding control module, wherein the visual image acquisition module is used to acquire the weld area image of the powder classifier rotor blade, the weld detection module is used to perform image segmentation and weld recognition on the processed weld area image of the powder classifier rotor blade, and determine the position information of the weld, and the welding control module is used to control the welding gun of the welding robot to weld along the weld according to the obtained weld position information;

[0005] The control system includes a welding track editing system, an instruction editing system, a robot control system and a laser ranging sensor. The robot control system is used to control the opening and closing operations of the welding robot. The welding track editing system is used to process and edit the weld information captured by the machine vision component. The instruction editing system is used to process and edit the issued control instructions. The instruction editing system is adapted to the welding track editing system. The instruction editing system is electrically connected to the robot control system and two welding robot bodies. The instruction editing system has a PLC in it to perform TCP communication with the welding robot. The PLC receives and calculates the parameters preset manually on the touch screen. The PLC stores 50 types of robot trial teaching positions of rotors to avoid repeated trial teaching of rotors of the same model. The laser ranging sensor is used for concentric alignment when the powder selector rotor is fed and the initial angle alignment of the turntable positioner for welding the first blade. A plurality of laser ranging sensors are provided and correspond to the two welding robots respectively.

[0006] The visual components specifically include:

[0007] A1: Read the image data to be processed, which is a visual image obtained in the working scene of the welding robot;

[0008] A2: For each pixel in the image, build a 5×5 window K with it as the center, and set the center coordinates of the current window K to (m, n);

[0009] A3: For window K, calculate the maximum grayscale value and the minimum grayscale value inside it;

[0010] A4: Traverse each pixel point in the window K, whose coordinates are (x, y), and obtain the gray value h(x, y) of the pixel point;

[0011] A5: Based on the noise detection formula To determine whether the pixel is a noise pixel; when F(x,y)=1, the pixel is a noise pixel; when F(x,y)=0, the pixel is a signal pixel;

[0012] A6: The noise pixels detected in window K are grouped into a set Z(m,n), and the signal pixels are grouped into a set X(m,n);

[0013] A7: Calculate the mean of the pixel grayscale values ​​in the signal set and median

[0014] A8: Count the number of pixels s in the noise set and the total number of pixels in the window K centered at the coordinates (m, n);

[0015] A9: Determine the weight coefficient α(s), which can be adjusted according to the actual welding scene and noise characteristics;

[0016] A10: Apply filtering formula Calculate the gray value h′(m,n) at the center position (m,n) of the filtered window K;

[0017] A11: Slide the window K pixel by pixel on the image, and repeat steps 2 and 3 for each window until all pixels in the image are processed;

[0018] A12: Save or output the filtered image data for use by the subsequent welding robot collaborative control algorithm;

[0019] A13: Evaluate the quality of the processed image using indicators such as image clarity and signal-to-noise ratio;

[0020] A14: If the evaluation result does not meet expectations, adjust the weight coefficient α(s) or other related parameters and execute the above steps again;

[0021] The overall system operation method includes the following steps:

[0022] S1: Place the powder classifier rotor on the turntable positioner by crane;

[0023] S2: After the rotor is loaded, a plurality of laser distance measuring sensors corresponding to the two welding robots are used to perform concentric alignment operations;

[0024] S3: After completing the concentric alignment, the initial angle of the rotor blade is aligned again with the aid of the laser distance sensor;

[0025] S4: Manual trial teaching is performed on two welding robots that are slidably mounted on two sets of ground rail column cantilevers;

[0026] S5: The operator inputs the relevant parameters of the rotor on the touch screen, and these parameters will be received and calculated by the PLC in the instruction editing system;

[0027] S6: connect the communication between PLC and welding robot, and store the position information obtained by manual teaching into PLC;

[0028] S7: Start the automatic welding command on the touch screen, the command editing system will process the command and convert it into specific control instructions and send it to the robot control system;

[0029] S8: The two welding robots first scan the two horizontal and two vertical sides of the blade using the laser weld position finder installed on the six axes;

[0030] S9: When the PLC receives the welding completion signal from the two robots, it controls the vertical axis to raise the robots to the second layer of the rotor;

[0031] S10: When all blades of a layer are welded, the PLC controls the turntable positioner to rotate to the angle of the next blade, and the robot repeats the above welding action; this cycle continues until all blades are welded;

[0032] S11: For subsequent rotors of the same model, after completing concentricity and angle alignment, the corresponding parameters stored in the PLC can be directly called without manual trial teaching again, and welding can be performed directly.

[0033] Beneficial effects of the present invention:

[0034] The dual robots work together through laser rangefinder sensors and visual components to achieve rapid concentric alignment of the rotor, blade angle calibration and real-time tracking of welds. The high-precision positioning of the laser rangefinder (error ≤ 0.1mm) combined with the intelligent image processing of the visual module (supporting 5×5 window noise filtering and feature segmentation) significantly improves the accuracy of welding path planning and avoids deviations caused by manual intervention. The two robots work together (such as regional welding or multi-layer synchronous operation) to increase welding efficiency by more than 40% compared with a single-machine system.

[0035] The system monitors the welding status in real time through PLC and robot TCP communication, and automatically adjusts the welding gun posture and welding parameters (current, voltage) according to dynamic changes such as weld offset and thermal deformation. The welding control module adopts the mean-median weighted filtering algorithm to effectively suppress welding spatter and noise interference, and ensure the consistency of weld formation (penetration fluctuation <5%).

[0036] The PLC pre-stores trial teaching parameters for 50 types of rotors, supports quick calling of products of the same model (shortening the changeover time by 80%), and enables flexible editing of multiple weld trajectories through the welding track editing system. The instruction editing system is compatible with manual parameter input and automatic program generation, and combined with the multi-angle positioning (±0.05° accuracy) of the turntable positioner, it can adapt to the fully automatic welding of complex impeller structures (such as multi-layer blades and special-shaped welds), reducing the scrap rate to below 0.5%. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 It is a schematic diagram of the overall structure of the dual-robot collaborative welding dynamic compensation system of the present invention;

[0038] Figure 2 It is a schematic diagram of the operation method of the dual-robot collaborative welding dynamic compensation system of the present invention; DETAILED DESCRIPTION

[0039] The technical solution of the present invention will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0040] In the description of the present invention, it should be noted that the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc., indicating the orientation or positional relationship, are based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first", "second", and "third" are used for descriptive purposes only, and cannot be understood as indicating or implying relative importance.

[0041] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected" and "connected" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal connection of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances. In addition, the technical features involved in the different embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0042] Example 1

[0043] The present invention provides a dual-robot collaborative welding dynamic compensation system, which is characterized by comprising two welding robots slidably mounted on two sets of ground rails, a laser weld positioner is mounted on the six axes of the welding robot, a machine vision component is arranged in the laser welding positioner, a column cantilever for driving the welding robot to move is slidably mounted on the ground rail, a turntable positioner is arranged between the two ground rails, a powder selector rotor is placed on the turntable positioner, and a control system is electrically arranged beside the welding robot;

[0044] The machine vision component includes a visual image acquisition module, a weld detection module and a welding control module, wherein the visual image acquisition module is used to acquire the weld area image of the powder classifier rotor blade, the weld detection module is used to perform image segmentation and weld recognition on the processed weld area image of the powder classifier rotor blade, and determine the position information of the weld, and the welding control module is used to control the welding gun of the welding robot to weld along the weld according to the obtained weld position information;

[0045] The control system includes a welding track editing system, an instruction editing system, a robot control system and a laser ranging sensor. The robot control system is used to control the opening and closing operations of the welding robot. The welding track editing system is used to process and edit the weld information captured by the machine vision component. The instruction editing system is used to process and edit the issued control instructions. The instruction editing system is adapted to the welding track editing system. The instruction editing system is electrically connected to the robot control system and two welding robot bodies. The instruction editing system has a PLC in it to perform TCP communication with the welding robot. The PLC receives and calculates the parameters preset manually on the touch screen. The PLC stores 50 types of robot trial teaching positions of rotors to avoid repeated trial teaching of rotors of the same model. The laser ranging sensor is used for concentric alignment when the powder selector rotor is fed and the initial angle alignment of the turntable positioner for welding the first blade. A plurality of laser ranging sensors are provided and correspond to the two welding robots respectively.

[0046] The visual components specifically include:

[0047] A1: Read the image data to be processed, which is a visual image obtained in the working scene of the welding robot;

[0048] A2: For each pixel in the image, build a 5×5 window K with it as the center, and set the center coordinates of the current window K to (m, n);

[0049] A3: For window K, calculate the maximum grayscale value and the minimum grayscale value inside it;

[0050] A4: Traverse each pixel point in the window K, whose coordinates are (x, y), and obtain the gray value h(x, y) of the pixel point;

[0051] A5: Based on the noise detection formula To determine whether the pixel is a noise pixel; when F(x,y)=1, the pixel is a noise pixel; when F(x,y)=0, the pixel is a signal pixel;

[0052] A6: The noise pixels detected in window K are grouped into a set Z(m,n), and the signal pixels are grouped into a set X(m,n);

[0053] A7: Calculate the mean of the pixel grayscale values ​​in the signal set and median

[0054] A8: Count the number of pixels s in the noise set and the total number of pixels in the window K centered at the coordinates (m, n);

[0055] A9: Determine the weight coefficient α(s), which can be adjusted according to the actual welding scene and noise characteristics;

[0056] A10: Apply filtering formula Calculate the gray value h′(m,n) at the center position (m,n) of the filtered window K;

[0057] A11: Slide the window K pixel by pixel on the image, and repeat steps 2 and 3 for each window until all pixels in the image are processed;

[0058] A12: Save or output the filtered image data for use by the subsequent welding robot collaborative control algorithm;

[0059] A13: Evaluate the quality of the processed image using indicators such as image clarity and signal-to-noise ratio;

[0060] A14: If the evaluation result does not meet expectations, adjust the weight coefficient α(s) or other related parameters and execute the above steps again;

[0061] The overall system operation method includes the following steps:

[0062] S1: Place the powder classifier rotor on the turntable positioner by crane;

[0063] S2: After the rotor is loaded, a plurality of laser distance measuring sensors corresponding to the two welding robots are used to perform concentric alignment operations;

[0064] S3: After completing the concentric alignment, the initial angle of the rotor blade is aligned again with the aid of the laser distance sensor;

[0065] S4: Manual trial teaching is performed on two welding robots that are slidably mounted on two sets of ground rail column cantilevers;

[0066] S5: The operator inputs the relevant parameters of the rotor on the touch screen, and these parameters will be received and calculated by the PLC in the instruction editing system;

[0067] S6: connect the communication between PLC and welding robot, and store the position information obtained by manual teaching into PLC;

[0068] S7: Start the automatic welding command on the touch screen, the command editing system will process the command and convert it into specific control instructions and send it to the robot control system;

[0069] S8: The two welding robots first scan the two horizontal and two vertical sides of the blade using the laser weld position finder installed on the six axes;

[0070] S9: When the PLC receives the welding completion signal from the two robots, it controls the vertical axis to raise the robots to the second layer of the rotor;

[0071] S10: When all blades of a layer are welded, the PLC controls the turntable positioner to rotate to the angle of the next blade, and the robot repeats the above welding action; this cycle continues until all blades are welded;

[0072] S11: For subsequent rotors of the same model, after completing concentricity and angle alignment, the corresponding parameters stored in the PLC can be directly called without manual trial teaching again, and welding can be performed directly.

[0073] Example 2

[0074] The rotor of the powder classifier is placed on the turntable positioner by a crane. The turntable positioner is located between two ground rails to provide stable support and rotation function for subsequent welding work;

[0075] After the rotor is loaded, multiple laser distance measuring sensors corresponding to the two welding robots are used for concentric alignment operations; these laser distance measuring sensors are part of the control system, and their data can be used to accurately determine the center position of the rotor on the turntable positioner to ensure the accuracy of subsequent welding;

[0076] After completing the concentric alignment, the initial angle of the rotor blade is aligned again with the help of the laser ranging sensor; the accurate initial angle is crucial for the subsequent robot to scan and weld the weld according to the predetermined trajectory;

[0077] Manual trial teaching is conducted on two welding robots that are slidably mounted on two sets of ground rail column cantilevers respectively; for the first blade corresponding to each robot, the laser scanning positions of the two horizontal sides, the laser scanning positions of the two vertical sides, the approach point, the welding start point and the welding end point are determined; these position information will be assisted by the visual image acquisition module and the weld detection module in the machine vision component to ensure the accuracy of the trial teaching position; after the trial teaching is completed, the welding program of the first blade is run empty to check the feasibility and accuracy of the program;

[0078] The operator inputs the relevant parameters of the rotor on the touch screen, and these parameters will be received and calculated by the PLC in the command editing system; as the key device for communication between the command editing system and the welding robot, the PLC has powerful data processing and storage capabilities;

[0079] Connect the PLC to the welding robot and store the position information obtained by manual teaching in the PLC. Since the PLC pre-stores 50 types of rotor robot teaching positions, repeated teaching can be avoided for the same type of rotor, thus improving work efficiency.

[0080] Start the automatic welding command on the touch screen, the command editing system will process the command and convert it into a specific control command and send it to the robot control system; the robot control system starts the operation of the welding robot according to the received command;

[0081] The two welding robots first use the laser weld finder installed on the six axes to scan the two horizontal and two vertical sides of the blade; the machine vision component in the laser weld finder works, the visual image acquisition module collects the image of the weld area, and the weld detection module processes and analyzes the image to determine the position information of the weld; then, the welding control module controls the welding robot's welding gun to weld along the weld based on this information and the offset of the coordinate system;

[0082] When the PLC receives the welding completion signal from the two robots, it controls the vertical axis to raise the robot to the second layer of the rotor. At this time, the robot repeats the welding action of the first layer, that is, it scans and welds the weld seam again to ensure the welding quality of each layer of blades.

[0083] When all blades of a layer are welded, the PLC controls the turntable positioner to rotate to the angle of the next blade, and the robot repeats the above welding action; this cycle continues until all blades are welded;

[0084] For subsequent rotors of the same model, after completing the concentricity and angle alignment, the corresponding parameters stored in the PLC can be directly called without manual trial teaching again, and welding can be performed directly, which greatly improves welding efficiency and production continuity.

[0085] The above shows and describes the basic principles and main features of the present invention and the advantages of the present invention. The various components mentioned in the present invention are common technologies in the existing field. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions are only to illustrate the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention to be protected. The scope of protection claimed by the present invention is defined by the attached claims and their equivalents.

Claims

1. Dual-robot collaborative welding dynamic compensation system, characterized in that: It comprises two welding robots slidably mounted on two sets of ground rails, a six-axis laser weld locator is mounted on the welding robot, a machine vision component is installed in the laser welding locator, a column cantilever for driving the welding robot to move is slidably mounted on the ground rail, a turntable positioner is arranged between the two ground rails, a powder selector rotor is placed on the turntable positioner, and a control system is electrically arranged beside the welding robot.

2. The dual-robot collaborative welding dynamic compensation system according to claim 1 is characterized in that: The machine vision component includes a visual image acquisition module, a weld detection module and a welding control module. The visual image acquisition module is used to acquire the weld area image of the powder classifier rotor blade. The weld detection module is used to perform image segmentation and weld recognition on the processed weld area image of the powder classifier rotor blade to determine the position information of the weld. The welding control module is used to control the welding gun of the welding robot to weld along the weld according to the obtained weld position information.

3. The dual-robot collaborative welding dynamic compensation system according to claim 2 is characterized in that: The control system includes a welding track editing system, an instruction editing system, a robot control system and a laser ranging sensor. The robot control system is used to control the opening and closing operations of the welding robot. The welding track editing system is used to process and edit the weld information captured by the machine vision component. The instruction editing system is used to process and edit the issued control instructions. The instruction editing system is compatible with the welding track editing system. The instruction editing system is electrically connected to the robot control system and two welding robot bodies. The instruction editing system has a PLC that communicates with the welding robot via TCP. The PLC receives and calculates the parameters preset manually on the touch screen. The PLC stores 50 types of robot trial teaching positions of rotors to avoid repeated trial teaching of rotors of the same model. The laser ranging sensor is used for concentric alignment when the powder selector rotor is fed and for initial angle alignment of the turntable positioner for welding the first blade. A plurality of laser ranging sensors are provided and correspond to the two welding robots respectively.

4. The dual-robot collaborative welding dynamic compensation system according to claim 3 is characterized in that: The visual components specifically include: A1: Read the image data to be processed, which is a visual image obtained in the working scene of the welding robot; A2: For each pixel in the image, build a 5×5 window K with it as the center, and set the center coordinates of the current window K to (m, n); A3: For window K, calculate the maximum grayscale value and the minimum grayscale value inside it; A4: Traverse each pixel point in the window K, whose coordinates are (x, y), and obtain the gray value h(x, y) of the pixel point; A5: Based on the noise detection formula To determine whether the pixel is a noise pixel; when F(x,y)=1, the pixel is a noise pixel; when F(x,y)=0, the pixel is a signal pixel; A6: The noise pixels detected in window K are grouped into a set Z(m,n), and the signal pixels are grouped into a set X(m,n); A7: Calculate the mean of the pixel grayscale values ​​in the signal set and median A8: Count the number of pixels s in the noise set and the total number of pixels in the window K centered at the coordinates (m, n); A9: Determine the weight coefficient α(s), which can be adjusted according to the actual welding scene and noise characteristics; A10: Apply filtering formula Calculate the gray value h′(m,n) at the center position (m,n) of the filtered window K; A11: Slide the window K pixel by pixel on the image, and repeat steps 2 and 3 for each window until all pixels in the image are processed; A12: Save or output the filtered image data for use by the subsequent welding robot collaborative control algorithm; A13: Evaluate the quality of the processed image using indicators such as image clarity and signal-to-noise ratio; A14: If the evaluation result does not meet expectations, adjust the weight coefficient α(s) or other related parameters and re-execute the above steps.

5. The dual-robot collaborative welding dynamic compensation system according to claim 4 is characterized in that: The operation method of the overall system includes the following steps: S1: Place the powder classifier rotor on the turntable positioner by crane; S2: After the rotor is loaded, a plurality of laser distance measuring sensors corresponding to the two welding robots are used to perform concentric alignment operations; S3: After completing the concentric alignment, the initial angle of the rotor blade is aligned again with the aid of the laser distance sensor; S4: Manual trial teaching is performed on two welding robots that are slidably mounted on two sets of ground rail column cantilevers; S5: The operator inputs the relevant parameters of the rotor on the touch screen, and these parameters will be received and calculated by the PLC in the instruction editing system; S6: connect the communication between PLC and welding robot, and store the position information obtained by manual teaching into PLC; S7: Start the automatic welding command on the touch screen, the command editing system will process the command and convert it into specific control instructions and send it to the robot control system; S8: The two welding robots first scan the two horizontal and two vertical sides of the blade using the laser weld position finder installed on the six axes; S9: When the PLC receives the welding completion signal from the two robots, it controls the vertical axis to raise the robots to the second layer of the rotor; S10: When all blades of a layer are welded, the PLC controls the turntable positioner to rotate to the angle of the next blade, and the robot repeats the above welding action; this cycle continues until all blades are welded; S11: For subsequent rotors of the same model, after completing concentricity and angle alignment, the corresponding parameters stored in the PLC can be directly called without manual trial teaching again, and welding can be performed directly.

Images