A flange pipe welding robot, system and method based on laser tracking and force perception

Abstract

The present application relates to the technical field of automatic welding equipment, and particularly relates to a flange pipeline welding robot, system and method based on laser tracking and force perception, the welding robot comprising a chassis, a 6-degree-of-freedom mechanical arm, a rotary positioner, a laser tracking module and a force perception module, the welding system integrating a weld image processing module, a parameter adjustment module and a data storage module through a control system, adopting PID, fuzzy PID, improved YOLOv8, multivariate linear regression and gradient descent algorithms, and the welding method being based on the system to complete clamping positioning, parameter setting, laser tracking, force feedback and adaptive adjustment; the present application realizes intelligent and efficient operation of flange pipeline welding, the welding efficiency being 3-5 times higher than that of manual operation, the changeover time being shortened to within 5 minutes, the weld trajectory precision being controlled within ±0.3 mm, and the qualified rate being greatly improved, so that the present application can be widely applied in the fields of petrochemical industry, municipal pipe network and power engineering.

Description

Technical Field

[0001] This invention relates to the field of automated welding equipment technology, and in particular to a flange pipe welding robot, system and method based on laser tracking and force sensing. Background Technology

[0002] In the field of pipeline and flange welding, welding quality and efficiency directly determine the safety and progress of the project. Traditional welding operations mostly use manual argon arc welding. Although this method can ensure the aesthetic appearance of the weld, it has obvious drawbacks: First, the welding efficiency is low, and the speed of manual operation is limited, making it difficult to meet the needs of large-scale mass production; second, it requires extremely high skill levels from welders, and the training period for skilled welders is long. Moreover, manual operation is easily affected by factors such as fatigue and emotions, resulting in large fluctuations in weld quality; third, the welding working environment is harsh, with welders exposed to harmful environments such as high temperature, dust, and arc radiation for a long time, resulting in high labor intensity and safety hazards.

[0003] While various automated welding equipment has emerged in the current technology, several technical problems remain to be solved: First, existing automated welding equipment has poor adaptability to pipe "out-of-roundness" deviations and cannot compensate for pipe shape deviations in real time during welding, easily leading to defects such as weld misalignment, uneven weld formation, incomplete penetration, or excessive deposition. Second, most existing MIG / MAG welding robots on the market are designed for regular welds and lack adaptive welding integration solutions for complex pipe flange structures, failing to achieve coordinated control of laser tracking, force sensing, and welding parameters. Third, existing welding equipment often lacks effective process monitoring and feedback mechanisms, failing to perceive changes in welding pressure and weld position in real time, making it difficult to achieve adaptive adjustments during the welding process, resulting in insufficient welding quality stability.

[0004] Therefore, developing a dedicated welding robot system for flanges and pipes that is integrated, intelligent, easy to program, highly efficient in welding, and can effectively solve problems such as poor adaptability to pipe out-of-roundness, complex teaching, and difficulty in displacement coordination has become a technical challenge that urgently needs to be addressed by those skilled in the art. Summary of the Invention

[0005] To address the shortcomings of existing technologies, the present invention aims to provide a flange and pipe welding robot, system, and method based on laser tracking and force perception. This invention achieves integrated and intelligent control of flange and pipe welding, solves technical problems such as poor adaptability to pipe out-of-roundness, cumbersome teaching programming, low production changeover efficiency, difficulty in positional coordination, and unstable welding quality of existing automatic welding equipment, improves welding efficiency and weld quality, reduces reliance on manual labor and labor intensity, and adapts to the welding needs of flanges and pipes of various specifications and working conditions.

[0006] A flange pipe welding robot based on laser tracking and force perception includes a base frame, a robot body, and a rotary positioner;

[0007] The robot body includes a base, a 6-DOF robotic arm, and a welding torch. The base is fixedly mounted on a frame, the 6-DOF robotic arm is mounted on the base, and the welding torch is mounted on the execution end of the 6-DOF robotic arm.

[0008] The rotary positioner includes a motor, a mounting plate, and a circular docking plate. The mounting plate is fixedly connected to the base, the motor is fixedly connected to the outer wall of the mounting plate, and its output shaft extends through the mounting plate to the inner side. The circular docking plate is fixedly installed on the output shaft of the motor.

[0009] The rotary positioner also includes a pipe clamping assembly, which includes an arc-shaped clamping plate, a slider, and an electric push rod.

[0010] The inner wall of the circular docking plate is symmetrically provided with sliding grooves, the slider is slidably adapted to the sliding grooves, the arc-shaped clamp is fixedly connected to the side wall of the slider, and the electric push rod is located in the sliding grooves on both sides respectively, with one end connected to the inner wall of the sliding groove and the other end connected to the slider.

[0011] The robot body also includes a laser tracking module and a force sensing module. The laser tracking module is installed at the front end of the welding torch, and the force sensing module is installed between the 6-DOF robotic arm actuator and the welding torch.

[0012] Furthermore: the rotary positioner also includes a dedicated quick-change fixture.

[0013] The special quick-change fixture includes a mounting block, a clamping roller, a drive motor, a screw, and a moving plate. The mounting block is fixedly mounted on the base frame, and a mounting groove is provided on the inner wall of the mounting block. The screws are symmetrically mounted in the mounting grooves. The drive motor is fixedly connected to the side wall of the mounting block, and its output shaft is fixedly connected to one end of the corresponding screw.

[0014] The movable plate is slidably connected to the corresponding screw, and a receiving plate is fixedly connected to the top of the movable plate. The clamping rollers are rotatably connected to the inner wall of the receiving plate and are arranged in pairs symmetrically above the mounting block to clamp the pipe to be welded from both sides and assist the pipe to rotate.

[0015] Furthermore: the laser tracking module includes a CCD camera, a semiconductor laser, an optical filter, and a cooling component, used to scan the weld position of the flange and pipe to be welded in real time, capture the geometric features of the weld, and transmit position deviation signals;

[0016] The cooling component is used to maintain the temperature of the electronic components of the laser tracking module below 50°C, and the optical filter is used to filter out welding arc interference to ensure clear acquisition of weld images.

[0017] The force sensing module uses a six-dimensional force sensor to detect in real time the linear force and torque in three directions of space that the welding torch (23) is subjected to during the welding process, capture the changes in welding pressure and transmit force feedback signals to realize tactile perception of the welding process.

[0018] Furthermore, the laser tracking module also includes a temperature monitor and a calibration data storage unit;

[0019] The temperature monitor is used to monitor the temperature of the semiconductor laser and triggers a protection mechanism when the cooling system fails.

[0020] The calibration data storage unit is used to store sensor head calibration data, enabling rapid interchange of the laser tracking module sensor heads and reducing downtime.

[0021] The force sensing module also includes a calibration component, which is used to periodically calibrate the six-dimensional force sensor by applying known forces and torques in different directions to correct the sensor output data and ensure the accuracy of force sensing.

[0022] Furthermore: A flange pipe welding system based on laser tracking and force perception includes a welding robot, a PLC control system, a weld image processing module, a welding parameter adjustment module, and a data storage module;

[0023] The PLC central control system is electrically connected to the welding robot, laser tracking module, force sensing module, weld seam image processing module, welding parameter adjustment module and data storage module respectively. It adopts multi-axis linkage control technology to realize the coordinated motion control of the 6-DOF robotic arm and rotary positioner.

[0024] The weld image processing module is used to receive weld image data transmitted by the laser tracking module, analyze the weld position and bevel shape through image recognition algorithms, extract the weld center coordinates and deviation information, and transmit the processed signal to the control system.

[0025] The welding parameter adjustment module is used to receive control signals from the control system and adjust the welding current, welding voltage, welding speed and shielding gas flow rate of MIG / MAG welding in real time to achieve adaptive optimization of welding process parameters.

[0026] The data storage module is used to store weld position data, force feedback data, welding parameter data, and quality inspection data after welding, which facilitates subsequent traceability and process optimization.

[0027] Furthermore, the PLC control system also includes a human-machine interface, which is used to display real-time data during the welding process, including weld deviation, force feedback value, welding parameters and equipment operating status. It also supports manual input of welding parameters, setting of welding paths and manual control of equipment start and stop.

[0028] The weld seam image processing module uses an AI image recognition algorithm, which can automatically identify weld seams of different flange specifications and different bevel forms, predict weld seam deviation trends, and send adjustment signals to the control system in advance to achieve predictive adjustment of weld seam tracking, ensuring that the welding trajectory accuracy is controlled within ±0.3mm.

[0029] Furthermore: A smart MIG welding method for flange pipes based on laser tracking and force sensing includes the following steps:

[0030] S1. Pipe and flange clamping and positioning: Fix one end of the pipe to be welded to the flange, place the other end of the pipe between the clamping rollers of the special quick-change fixture, start the electric actuator, drive the slider to make the arc-shaped clamping plate slide along the groove of the circular butt plate, and cooperate with the clamping rollers to achieve bidirectional clamping and positioning of the pipe from the outer periphery, while fixing the flange to the circular butt plate.

[0031] S2. Equipment initialization and parameter setting: Start the welding system through the human-machine interface, calibrate the laser tracking module and force sensing module, input the specifications, bevel type and preset welding parameters of the flange pipe to be welded, including welding current, welding voltage, welding speed and shielding gas flow rate, and the PLC control system plans the welding path.

[0032] S3. Laser tracking and positioning: The laser tracking module is activated, and the semiconductor laser projects laser stripes onto the weld area of ​​the flange and pipe. The CCD camera captures the weld image through the optical filter. The weld image processing module analyzes the image, extracts the weld center position and deviation data, and transmits it to the PLC control system.

[0033] S4. Force sensing and welding start: Start the 6-DOF robotic arm and rotary positioner. The rotary positioner drives the pipe and flange to rotate synchronously. The robotic arm drives the welding torch to move to the weld start position and start MIG welding. The force sensing module detects the force and torque signals on the welding torch in real time and transmits them to the PLC control system.

[0034] S5. Real-time adaptive adjustment: The PLC control system adjusts the posture of the robotic arm and the position of the welding torch according to the weld deviation data transmitted by the laser tracking module to ensure that the welding torch is always aligned with the center of the weld. At the same time, according to the force feedback data transmitted by the force sensing module, the welding current, voltage and welding speed are adjusted through the welding parameter adjustment module to keep the welding pressure within the preset range and avoid defects such as weld deviation, incomplete penetration or excessive deposition.

[0035] S6. Welding Completion and Inspection: After the welding torch completes all weld seams along the preset path, the welding power supply and shielding gas are turned off, the robotic arm and rotary positioner are stopped, the pipe clamping assembly and special quick-change fixture are released, and the welded flange pipe is removed; the data storage module saves all process data of this welding, and the welding operation is completed.

[0036] Furthermore: In S2, the preset welding parameters can be adaptively matched according to the material of the pipe to be welded, the flange specifications and the bevel form. The PLC central control system calls the historical welding data in the data storage module and generates the optimal welding parameter combination through algorithm optimization.

[0037] In S5, when the laser tracking module detects that the weld deviation exceeds the preset threshold, or the force sensing module detects that the welding pressure is abnormal, the PLC control system immediately issues an early warning signal and suspends the welding operation. Welding will resume after manual confirmation or automatic adjustment to the normal range.

[0038] In step S6, after welding is completed, the weld seam can be visually inspected, and the inspection data is entered into the data storage module for subsequent iterative optimization of the welding process.

[0039] Furthermore: the weld position deviation in S3 is calculated using the following function formula:

[0040]

[0041] In the formula: This refers to the spatial positional deviation between the welding torch and the center of the weld. The three-dimensional coordinates of the weld center detected in real time by the laser tracking module; The preset three-dimensional coordinates of the weld center;

[0042] The AI ​​image recognition in the weld seam image processing module of S3 uses the improved YOLOv8 algorithm, and the weld seam feature extraction uses a convolutional neural network (CNN). The core feature extraction formula is as follows:

[0043]

[0044] In the formula: This is the output feature map of the l-th convolutional layer; The ReLU function is used as the activation function. ; Let be the weight matrix of the convolutional kernel in the l-th layer; This is the output feature map of the (l-1)th layer; The bias vector of the l-th layer; the weld deviation trend prediction uses an LSTM neural network, and the prediction formula is:

[0045]

[0046] In the formula: This is the predicted output at time t; The hidden layer state at time t-1; The laser tracking input data is given at time t; This is the weight matrix; This is the bias vector.

[0047] Furthermore: the parameter optimization in S5 adopts a fuzzy PID algorithm based on force feedback and weld deviation, and the core optimization formula is as follows:

[0048]

[0049]

[0050]

[0051] In the formula: The welding current, welding voltage, and welding speed are adjusted in real time. Preset baseline parameters; This refers to the parameter adjustment coefficient; The difference (N) between the actual welding pressure detected by the force sensing module and the preset pressure. The function is a fuzzy inference function. It maps pressure deviation and position deviation into parameter adjustment values ​​through fuzzy rules. The fuzzy rules adopt the if-then form.

[0052] In step S5, a PID algorithm is used to adjust the welding torch position in real time. The expression for the PID control algorithm is:

[0053]

[0054] In the formula: This is the amount of adjustment for the welding torch position; This is the proportionality coefficient; This represents the real-time position deviation. The integral time constant; is the differential time constant.

[0055] The present invention has the following beneficial effects:

[0056] 1. Solves the problem of poor adaptability to pipe non-roundness: The laser tracking module captures weld position deviation in real time, and combined with the PID algorithm, it achieves rapid response and adjustment within 0.1s, dynamically correcting the welding torch position with an adjustment accuracy of ±0.05mm. At the same time, with the six-dimensional force feedback of the force sensing module, it captures welding pressure changes in real time, realizing bidirectional adaptive compensation in the welding process. It can effectively offset the non-roundness deviation of the pipe within 0.5mm, ensuring uniform weld formation and welding trajectory accuracy controlled within ±0.3mm, avoiding common defects such as weld deviation, incomplete penetration, and excessive deposition. Compared with existing automatic welding equipment, the weld pass rate is improved by more than 14%.

[0057] 2. Simplified teaching programming and improved changeover efficiency: Through the improved YOLOv8AI image recognition algorithm, it can automatically identify welds of different flange specifications and different bevel forms without manual teaching; combined with the multiple linear regression algorithm, it can call historical welding data in the data storage module and generate the optimal welding parameter combination within 10 seconds, which greatly shortens the changeover time and improves the changeover efficiency by more than 80%. It can flexibly adapt to the production needs of multiple varieties and small batches. No professional programmers are required. Ordinary operators can start operating after simple training.

[0058] 3. Achieve intelligent collaborative control of the welding process: Integrate technologies such as laser tracking, force sensing, and multi-axis linkage control, and realize synchronous collaboration between the 6-DOF robotic arm and the rotary positioner through the PLC central control system. The collaboration error is ≤0.05mm. During the welding process, there is no need for manual intervention in the welding torch position, welding parameters, and positioner speed, realizing fully automated welding. Compared with manual argon arc welding, the welding efficiency is increased by 3-5 times, while reducing the dependence on the welder's skill level, avoiding quality fluctuations caused by manual operation, and reducing the harm to operators caused by high temperature, dust and other harmful environments.

[0059] 4. Strong equipment adaptability and convenient operation: The dedicated quick-change clamps and pipe clamping components work together to adapt to flange pipes of different diameters from 50mm to 200mm. The human-machine interface intuitively displays real-time welding data and equipment operating status, supports one-button start, manual and automatic mode switching, and has a fault warning function to quickly locate equipment fault points, which is convenient for maintenance. The welding torch and laser tracking module adopt quick-change interfaces, and the disassembly and assembly time is ≤5 minutes, reducing equipment maintenance costs and downtime. Attached Figure Description

[0060] Figure 1 A schematic diagram of the overall structure of an intelligent MIG / MAG welding robot for flanges and pipes based on laser tracking and force perception;

[0061] Figure 2 A schematic diagram of a 6-DOF robotic arm structure for an intelligent MIG welding robot for flanges and pipes based on laser tracking and force perception.

[0062] Figure 3 A schematic diagram of a rotary positioner structure for a flange pipe welding robot based on laser tracking and force perception;

[0063] Figure 4 A schematic diagram of a dedicated quick-change fixture structure for a flange pipe welding robot based on laser tracking and force perception;

[0064] Figure 5 A schematic diagram of the pipe clamping assembly structure for a flange pipe welding robot based on laser tracking and force perception;

[0065] Figure 6 A schematic diagram of the arc-shaped clamping plate installation structure for a flange pipe welding robot based on laser tracking and force perception;

[0066] Figure 7 This is a flowchart of a flange pipe welding method based on laser tracking and force sensing.

[0067] The reference numerals in the figures are as follows: 1. Base frame; 2. Robot body; 21. Base; 22. 6-DOF robotic arm; 23. Welding torch; 3. Rotary positioner; 31. Motor; 32. Mounting plate; 33. Circular docking plate; 34. Pipe clamping assembly; 341. Arc-shaped clamping plate; 342. Slider; 343. Electric actuator; 35. Special quick-change fixture; 351. Mounting block; 352. Clamping roller; 353. Drive motor; 354. Screw; 355. Moving plate; 4. Slide groove; 5. Mounting groove; 6. Flange pipe; 7. Receiving plate. Detailed Implementation

[0068] The present invention will be further described in detail below with reference to specific embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0069] Please see the appendix Figures 1-6 The present invention provides an embodiment 1: a flange pipe 6 welding robot based on laser tracking and force perception, including a base frame 1, a robot body 2 and a rotary positioner 3;

[0070] The robot body 2 includes a base 21, a 6-DOF robotic arm 22, and a welding torch 23. The base 21 is fixedly mounted on the base frame 1, the 6-DOF robotic arm 22 is mounted on the base 21, and the welding torch 23 is mounted on the execution end of the 6-DOF robotic arm 22.

[0071] The rotary positioner 3 includes a motor 31, a mounting plate 32, and a circular docking plate 33. The mounting plate 32 is fixedly connected to the base 21, the motor 31 is fixedly connected to the outer wall of the mounting plate 32, and its output shaft extends through the mounting plate 32 to the inner side. The circular docking plate 33 is fixedly installed on the output shaft of the motor 31.

[0072] The rotary positioner 3 also includes a pipe clamping assembly 34, which includes an arc-shaped clamping plate 341, a slider 342, and an electric push rod 343.

[0073] The inner wall of the circular docking plate 33 is symmetrically provided with grooves 4. The slider 342 slides and adapts to the groove 4. The arc-shaped clamping plate 341 is fixedly connected to the side wall of the slider 342. The electric push rods 343 are located in the two side grooves 4 respectively, with one end connected to the inner wall of the groove 4 and the other end connected to the slider 342.

[0074] The robot body 2 also includes a laser tracking module and a force sensing module. The laser tracking module is installed at the front end of the welding torch 23, and the force sensing module is installed between the execution end of the 6-DOF robotic arm 22 and the welding torch 23.

[0075] The rotary positioner 3 also includes a dedicated quick-change fixture 35:

[0076] The special quick-change fixture 35 includes a mounting block 351, a clamping roller 352, a drive motor 353, a screw 354, and a moving plate 355. The mounting block 351 is fixedly mounted on the base frame 1. The inner wall of the mounting block 351 is provided with a mounting groove 5. The screw 354 is symmetrically mounted in the mounting groove 5. The drive motor 353 is fixedly connected to the side wall of the mounting block 351, and its output shaft is fixedly connected to one end of the corresponding screw 354.

[0077] The movable plate 355 is slidably connected to the corresponding screw 354. The top of the movable plate 355 is fixedly connected to the receiving plate 7. The clamping roller 352 is rotatably connected to the inner wall of the receiving plate 7 and is arranged in pairs symmetrically above the mounting block 351 to clamp the pipe to be welded from both sides and assist the pipe to rotate.

[0078] The laser tracking module includes a CCD camera, a semiconductor laser, an optical filter, and a cooling component. It is used to scan the weld position of the flange and pipe to be welded in real time, capture the geometric features of the weld, and transmit the position deviation signal.

[0079] The cooling system is used to keep the temperature of the electronic components of the laser tracking module below 50°C, and the optical filter is used to filter out welding arc interference to ensure clear acquisition of weld images.

[0080] The force sensing module uses a six-dimensional force sensor, which is installed between the execution end of the six-degree-of-freedom robotic arm 22 and the welding torch 23. It is used to detect the linear force and torque in three spatial directions that the welding torch 23 experiences in real time during the welding process, capture changes in welding pressure and transmit force feedback signals to realize tactile perception of the welding process.

[0081] The laser tracking module also includes a temperature monitor and a calibration data storage unit;

[0082] Temperature monitors are used to monitor the temperature of semiconductor lasers and trigger protection mechanisms when the cooling system fails.

[0083] The calibration data storage unit is used to store sensor head calibration data, enabling rapid interchange of the laser tracking module sensor heads and reducing downtime.

[0084] The force sensing module also includes a calibration component, which is used to periodically calibrate the six-dimensional force sensor by applying known forces and torques in different directions to correct the sensor output data and ensure the accuracy of force sensing.

[0085] An embodiment 2 of the present invention provides: a flange pipe 6 welding system based on laser tracking and force perception, including a welding robot, a PLC control system, a weld image processing module, a welding parameter adjustment module and a data storage module;

[0086] The PLC control system is electrically connected to the welding robot, laser tracking module, force sensing module, weld seam image processing module, welding parameter adjustment module and data storage module respectively. It adopts multi-axis linkage control technology to realize the coordinated motion control of the 6-DOF robotic arm 22 and the rotary positioner 3.

[0087] The weld image processing module receives weld image data transmitted by the laser tracking module, analyzes the weld position and bevel shape through image recognition algorithms, extracts the weld center coordinates and deviation information, and transmits the processed signal to the PLC control system.

[0088] The welding parameter adjustment module is used to receive control signals from the PLC main control system and adjust the welding current, welding voltage, welding speed and shielding gas flow rate of MIG / MAG welding in real time to achieve adaptive optimization of welding process parameters.

[0089] The data storage module is used to store weld position data, force feedback data, welding parameter data, and quality inspection data after welding, which facilitates subsequent traceability and process optimization.

[0090] The PLC control system also includes a human-machine interface, which displays real-time data during the welding process, including weld deviation, force feedback value, welding parameters and equipment operating status. It also supports manual input of welding parameters, setting of welding paths and manual control of equipment start and stop.

[0091] The weld image processing module uses an AI image recognition algorithm to automatically identify welds of different flange specifications and different bevel forms, predict weld deviation trends, and send adjustment signals to the PLC control system in advance to achieve predictive adjustment of weld tracking, ensuring that the welding trajectory accuracy is controlled within ±0.3mm.

[0092] Please see the appendix Figure 1 The present invention provides an embodiment 3: a welding method for flange pipe 6 based on laser tracking and force sensing, comprising the following steps:

[0093] S1. Pipe and flange clamping and positioning: One end of the pipe to be welded is fixedly connected to the flange, and the other end of the pipe is placed between the clamping rollers 352 of the special quick-change fixture 35. The electric actuator 343 is activated to drive the slider 342 to make the arc-shaped clamping plate 341 slide along the groove 4 of the circular butt plate 33. The clamping rollers 352 work together to achieve bidirectional clamping and positioning of the pipe from the outer periphery of the pipe, and at the same time, the flange is fixedly connected to the circular butt plate 33.

[0094] S2. Equipment initialization and parameter setting: Start the welding system through the human-machine interface, calibrate the laser tracking module and force sensing module, input the specifications, bevel type and preset welding parameters of the flange pipe 6 to be welded, including welding current, welding voltage, welding speed and shielding gas flow rate, and the PLC control system plans the welding path.

[0095] In S2, the preset welding parameters can be adaptively matched according to the material of the pipe to be welded, the flange specifications and the bevel form. The PLC central control system calls the historical welding data in the data storage module and generates the optimal welding parameter combination through algorithm optimization.

[0096] The adaptive matching of preset welding parameters in S2 adopts a multiple linear regression algorithm. Based on historical welding data in the data storage module, it generates the optimal parameter combination. The core formula is as follows:

[0097]

[0098] In the formula: These are the optimal preset parameters; The material coefficient of the pipe to be welded; The outer diameter of the pipe; The bevel angle; These are the weighting coefficients for each influencing factor; The baseline offset is used; the weighting coefficients are obtained through training with historical data, and the training objective is to minimize the deviation between the actual welding quality and the standard quality.

[0099] In S2, the calibration uses a least squares fitting algorithm, and the calibration formula is as follows:

[0100]

[0101] In the formula: The calibrated force or torque value; This is the calibration coefficient matrix; This is the sensor's original output value; To calibrate the offset vector; the calibration coefficient matrix A is solved using the least squares method, and the objective function is:

[0102]

[0103] In the formula: This represents the standard force or torque value at the i-th calibration point; Number of calibration points;

[0104] S3. Laser tracking and positioning: The laser tracking module is activated, and the semiconductor laser projects laser stripes onto the weld area of ​​the flange and pipe. The CCD camera captures the weld image through the optical filter. The weld image processing module analyzes the image, extracts the weld center position and deviation data, and transmits it to the PLC control system.

[0105] The following function formula is used to calculate the weld position deviation in S3:

[0106]

[0107] In the formula: This refers to the spatial positional deviation between the welding torch and the center of the weld. The three-dimensional coordinates of the weld center detected in real time by the laser tracking module; The preset three-dimensional coordinates of the weld center;

[0108] In the weld seam image processing module of S3, AI image recognition uses the improved YOLOv8 algorithm, and weld seam feature extraction employs a convolutional neural network (CNN). The core feature extraction formula is as follows:

[0109]

[0110] In the formula: This is the output feature map of the l-th convolutional layer; The ReLU function is used as the activation function. ; Let be the weight matrix of the convolutional kernel in the l-th layer; This is the output feature map of the (l-1)th layer; The bias vector of the l-th layer; the weld deviation trend prediction uses an LSTM neural network, and the prediction formula is:

[0111]

[0112] In the formula: This is the predicted output at time t; The hidden layer state at time t-1; The laser tracking input data is given at time t; This is the weight matrix; It is the bias vector;

[0113] S4. Force sensing and welding start: Start the 6-DOF robotic arm 22 and rotary positioner 3. The rotary positioner 3 drives the pipe and flange to rotate synchronously. The robotic arm drives the welding torch 23 to move to the weld start position and start MIG welding. The force sensing module detects the force and torque signals on the welding torch 23 in real time and transmits them to the PLC control system.

[0114] S5. Real-time adaptive adjustment: The PLC control system adjusts the posture of the robotic arm and the position of the welding torch 23 according to the weld deviation data transmitted by the laser tracking module, ensuring that the welding torch 23 is always aligned with the center of the weld. At the same time, according to the force feedback data transmitted by the force sensing module, the welding current, voltage and welding speed are adjusted through the welding parameter adjustment module to keep the welding pressure within the preset range and avoid defects such as weld deviation, incomplete penetration or excessive deposition.

[0115] In S5, parameter optimization employs a fuzzy PID algorithm based on force feedback and weld deviation. The core optimization formula is as follows:

[0116]

[0117]

[0118]

[0119] In the formula: The welding current, welding voltage, and welding speed are adjusted in real time. Preset baseline parameters; This refers to the parameter adjustment coefficient; The difference (N) between the actual welding pressure detected by the force sensing module and the preset pressure. The function is a fuzzy inference function. It maps pressure deviation and position deviation into parameter adjustment values ​​through fuzzy rules. The fuzzy rules adopt the if-then form.

[0120] In S5, a PID algorithm is used to adjust the welding torch position in real time. The expression for the PID control algorithm is:

[0121]

[0122] In the formula: This is the amount of adjustment for the welding torch position; This is the proportionality coefficient; This represents the real-time position deviation. The integral time constant; The differential time constant;

[0123] In S5, when the laser tracking module detects that the weld deviation exceeds the preset threshold, or the force sensing module detects that the welding pressure is abnormal, the PLC control system immediately issues an early warning signal and suspends the welding operation. Welding will resume after manual confirmation or automatic adjustment to the normal range.

[0124] Anomaly detection uses a threshold-based algorithm, with the core formula as follows:

[0125]

[0126] In the formula: A preset threshold is set for weld position deviation; A preset threshold is set for welding pressure deviation; when an abnormality is detected, the control system triggers an early warning signal.

[0127] The formula for warning signal strength is: ,in These are the weighting coefficients. The intensity of the warning;

[0128] S6. Welding Completion and Inspection: After the welding torch 23 completes all weld seams along the preset path, the welding power supply and shielding gas are turned off, the operation of the robotic arm and rotary positioner 3 is stopped, the pipe clamping assembly 34 and the special quick-change fixture 35 are released, and the welded flange pipe 6 is taken out; the data storage module saves all process data of this welding, and the welding operation is completed;

[0129] In S6, after welding is completed, the weld seam can be visually inspected, and the inspection data is entered into the data storage module for subsequent welding process iteration and optimization.

[0130] An embodiment 4 provided by the present invention:

[0131] I. Specific Implementation Examples of Flange and Pipe Welding Robots

[0132] The welding robot in this embodiment consists of three parts: base frame 1, robot body 2, and rotary positioner 3. It is adapted to flange pipe 6 welding operations in scenarios such as petrochemical and municipal pipeline networks.

[0133] The robot body 2 includes a base 21, a 6-DOF robotic arm 22 and a welding torch 23. The base 21 is stably mounted on the base frame 1, the 6-DOF robotic arm 22 is fixed to the base 21, and the welding torch 23 is mounted on the end effector of the robotic arm, which can flexibly adjust the welding posture.

[0134] The rotary positioner 3 consists of a drive motor 353, a mounting plate 32, and a circular docking plate 33. The mounting plate 32 is fixedly connected to the outside of the robot base 21. The drive motor 353 is installed on the outer wall of the mounting plate 32, and its output shaft extends through the mounting plate 32 to the inner side and is fixedly connected to the circular docking plate 33, which can drive the docking plate to rotate at a uniform speed.

[0135] The rotary positioner 3 is equipped with a pipe clamping assembly 34, which includes an arc-shaped clamping plate 341, a slider 342, and an electric push rod 343.

[0136] The inner wall of the circular docking plate 33 is symmetrically provided with grooves 4, and the slider 342 can slide smoothly in the grooves 4. The arc-shaped clamping plate 341 is fixed to the side wall of the slider 342 to fit and clamp the outer wall of the pipe. The electric push rods 343 are located in the two side grooves 4 respectively, with one end connected to the inner wall of the groove 4 and the other end connected to the slider 342. The slider 342 is driven to slide by the extension and retraction of the electric push rods 343 to complete the clamping and loosening operation of the pipe.

[0137] The rotary positioner 3 is also equipped with a special quick-change fixture 35. The fixture consists of a mounting block 351, clamping rollers 352, a drive motor 353, screws 354, and a moving plate 355. The mounting block 351 is fixed on the base frame 1 and has an mounting groove 5 inside. Two screws 354 are symmetrically installed in the groove. The drive motor 353 is fixed to the side wall of the mounting block 351 and connected to the screws 354. The moving plate 355 is sleeved on the screws 354 and can move back and forth with the rotation of the screws 354. The top of the moving plate 355 is provided with a receiving plate 7. The pair of clamping rollers 352 are rotatably connected to the inside of the receiving plate 7 to clamp the workpiece from both sides of the pipe and at the same time assist the pipe to rotate smoothly.

[0138] The robot body 2 integrates a laser tracking module and a force sensing module. The laser tracking module is installed at the front end of the welding torch 23 and consists of a CCD camera, a semiconductor laser, an optical filter, and a cooling component. The cooling component controls the temperature of the electronic components inside the module below 50°C, and the optical filter filters out welding arc interference to ensure clear acquisition of weld seam images. The module is also equipped with a temperature monitor and a calibration data storage unit to monitor the laser temperature in real time, trigger protection when cooling fails, and store sensor head calibration data for quick interchangeability.

[0139] The force sensing module uses a six-dimensional force sensor, which is installed between the robotic arm's execution end and the welding torch 23 to detect the three-dimensional linear force and three-dimensional torque acting on the welding torch 23 in real time and capture changes in welding pressure. The module is equipped with a calibration component, which can periodically calibrate the sensor, correct the output data, and ensure the accuracy of force sensing.

[0140] II. Specific Implementation of Flange Pipe Welding System

[0141] The welding system in this embodiment is based on the welding robot in embodiment 1, and is coordinated with a PLC control system, a weld seam image processing module, a welding parameter adjustment module, and a data storage module.

[0142] The PLC control system is electrically connected to all components, including the robot, laser tracking module, and force sensing module. It adopts multi-axis linkage control technology to precisely control the synchronous movement of the 6-DOF robotic arm 22 and the rotary positioner 3 with minimal coordination error.

[0143] The system is equipped with a human-machine interface that displays weld deviation, force feedback value, welding parameters, and equipment operating status in real time. It supports manual input of parameters, setting of welding paths, and manual control of equipment start and stop. The weld image processing module receives weld image data transmitted by the laser tracking module and automatically identifies welds of different flange specifications and different bevel forms through AI image recognition algorithms. It analyzes the weld position and bevel shape, extracts the weld center coordinates and deviation information, and can also predict the weld deviation trend and send adjustment signals to the PLC control system in advance to control the welding trajectory accuracy within ±0.3mm.

[0144] The welding parameter adjustment module receives control commands from the PLC central control system and dynamically adjusts the current, voltage, welding speed, and shielding gas flow rate of MIG / MAG welding in real time, performing adaptive optimization of process parameters based on the welding status. The data storage module records weld position data, force feedback data, and welding parameter data throughout the welding process, and stores quality inspection data after welding is completed, facilitating subsequent production traceability and iterative optimization of the welding process.

[0145] III. Specific Implementation of Intelligent MIG Welding Method for Flanges and Pipelines

[0146] The welding method in this embodiment is based on the system of Embodiment 2 and the robot of Embodiment 1, and is completed in a fully automated manner. The specific steps are as follows:

[0147] Pipe and flange clamping and positioning: First, fix one end of the pipe to be welded to the flange. Place the other end of the pipe between the clamping rollers 352 of the special quick-change fixture 35. Start the electric actuator 343 to drive the arc-shaped clamping plate 341 to slide along the groove 4 of the circular docking plate 33. The clamping rollers 352 work together to complete bidirectional clamping and positioning from the outer periphery of the pipe. At the same time, the flange is firmly docked with the circular docking plate 33.

[0148] Equipment initialization and parameter setting: Start the entire welding system through the human-machine interface and complete the calibration of the laser tracking module and force sensing module; input the basic information such as the specifications and bevel type of the pipe to be welded, and the system will automatically call up historical welding data, match and generate the optimal welding parameters, and the PLC control system will simultaneously plan the welding path.

[0149] Laser tracking and positioning: When the laser tracking module is activated, the semiconductor laser projects laser stripes onto the weld area of ​​the flange and pipe. The CCD camera captures a clear image of the weld through an optical filter. The weld image processing module analyzes the image, extracts the weld center position and deviation data, and transmits it to the PLC control system in real time.

[0150] Force sensing and welding start-up: The 6-DOF robotic arm 22 and rotary positioner 3 are started simultaneously. The rotary positioner 3 drives the pipe and flange to rotate at a uniform speed and synchronously. The robotic arm drives the welding torch 23 to move precisely to the starting position of the weld and start the MIG / MAG welding operation. The force sensing module detects the force and torque signals of the welding torch 23 in real time and continuously transmits them to the PLC control system.

[0151] Real-time adaptive adjustment: Based on the weld deviation data from laser tracking, the PLC central control system adjusts the robotic arm posture and welding torch 23 position in real time to ensure that the welding torch 23 is always aligned with the weld center. Simultaneously, combined with pressure feedback data from force sensing, the system adjusts the welding current, voltage, and welding speed through the welding parameter adjustment module to stabilize the welding pressure within a preset range, avoiding defects such as weld misalignment, incomplete penetration, and over-deposition. If the weld deviation or welding pressure exceeds the preset threshold, the system immediately issues a warning and pauses welding, automatically restarting the operation once the status returns to normal.

[0152] Welding Completion and Inspection: After the welding torch 23 completes the welding of the entire section of the weld along the preset path, the system automatically shuts off the welding power and shielding gas, and stops the operation of the robotic arm and rotary positioner 3; the pipe clamping assembly 34 and the special quick-change fixture 35 are released, and the welded flange pipe 6 is taken out; all process data of this welding is automatically stored in the data storage module, and the weld can be visually inspected. The inspection data is synchronously entered into the system for subsequent welding process optimization.

[0153] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A flange pipe welding robot based on laser tracking and force perception, characterized in that, Includes a base frame (1), a robot body (2), and a rotary positioner (3); The robot body (2) includes a base (21), a 6-DOF robotic arm (22), and a welding torch (23). The base (21) is fixedly mounted on the base frame (1), the 6-DOF robotic arm (22) is mounted on the base (21), and the welding torch (23) is mounted on the execution end of the 6-DOF robotic arm (22). The rotary positioner (3) includes a motor (31), a mounting plate (32) and a circular docking plate (33). The mounting plate (32) is fixedly connected to the base (21). The motor (31) is fixedly connected to the outer wall of the mounting plate (32), and its output shaft extends through the mounting plate (32) to the inner side. The circular docking plate (33) is fixedly installed on the output shaft of the motor (31). The rotary positioner (3) also includes a pipe clamping assembly (34), which includes an arc-shaped clamping plate (341), a slider (342), and an electric push rod (343). The inner wall of the circular docking plate (33) is symmetrically provided with grooves (4), the slider (342) is slidably adapted to the groove (4), the arc-shaped clamp (341) is fixedly connected to the side wall of the slider (342), and the electric push rod (343) is located in the two side grooves (4) respectively, with one end connected to the inner wall of the groove (4) and the other end connected to the slider (342); The robot body (2) also includes a laser tracking module and a force sensing module. The laser tracking module is installed at the front end of the welding torch (23), and the force sensing module is installed between the execution end of the 6-DOF robotic arm (22) and the welding torch (23).

2. The flange pipe welding robot based on laser tracking and force perception according to claim 1, characterized in that, The rotary positioner (3) also includes a dedicated quick-change fixture (35): The special quick-change fixture (35) includes a mounting block (351), a clamping roller (352), a drive motor (353), a screw (354), and a moving plate (355). The mounting block (351) is fixedly mounted on the base frame (1). The inner wall of the mounting block (351) is provided with a mounting groove (5). The screw (354) is symmetrically mounted in the mounting groove (5). The drive motor (353) is fixedly connected to the side wall of the mounting block (351), and its output shaft is fixedly connected to one end of the corresponding screw (354). The movable plate (355) is slidably connected to the corresponding screw (354), and the top of the movable plate (355) is fixedly connected to the receiving plate (7). The clamping roller (352) is rotatably connected to the inner wall of the receiving plate (7) and is arranged in pairs symmetrically above the mounting block (351) to clamp the pipe to be welded from both sides and assist the pipe to rotate.

3. The flange pipe welding robot based on laser tracking and force perception according to claim 1, characterized in that, The laser tracking module includes a CCD camera, a semiconductor laser, an optical filter, and a cooling component, used to scan the weld position of the flange and pipe to be welded in real time, capture the geometric features of the weld, and transmit position deviation signals. The cooling component is used to maintain the temperature of the electronic components of the laser tracking module below 50°C, and the optical filter is used to filter out welding arc interference to ensure clear acquisition of weld images. The force sensing module uses a six-dimensional force sensor to detect in real time the linear force and torque in three directions of space that the welding torch (23) is subjected to during the welding process, capture the changes in welding pressure and transmit force feedback signals to realize tactile perception of the welding process.

4. The flange pipe welding robot based on laser tracking and force perception according to claim 1, characterized in that, The laser tracking module also includes a temperature monitor and a calibration data storage unit; The temperature monitor is used to monitor the temperature of the semiconductor laser and triggers a protection mechanism when the cooling system fails. The calibration data storage unit is used to store sensor head calibration data, enabling rapid interchange of the laser tracking module sensor heads and reducing downtime. The force sensing module also includes a calibration component, which is used to periodically calibrate the six-dimensional force sensor by applying known forces and torques in different directions to correct the sensor output data and ensure the accuracy of force sensing.

5. A flange pipe welding system based on laser tracking and force sensing, characterized in that, Includes the welding robot, PLC control system, weld image processing module, welding parameter adjustment module and data storage module as described in any one of claims 1-4; The PLC control system is electrically connected to the welding robot, laser tracking module, force sensing module, weld seam image processing module, welding parameter adjustment module and data storage module respectively. It adopts multi-axis linkage control technology to realize the coordinated action control of the 6-DOF robotic arm (22) and the rotary positioner (3). The weld image processing module is used to receive weld image data transmitted by the laser tracking module, analyze the weld position and bevel shape through image recognition algorithms, extract the weld center coordinates and deviation information, and transmit the processed signal to the PLC control system. The welding parameter adjustment module is used to receive control signals from the PLC main control system and adjust the welding current, welding voltage, welding speed and shielding gas flow rate of MIG / MAG welding in real time to achieve adaptive optimization of welding process parameters. The data storage module is used to store weld position data, force feedback data, welding parameter data, and quality inspection data after welding, which facilitates subsequent traceability and process optimization.

6. A flange pipe welding system based on laser tracking and force sensing according to claim 5, characterized in that, The PLC control system also includes a human-machine interface, which is used to display real-time data during the welding process, including weld deviation, force feedback value, welding parameters and equipment operating status. It also supports manual input of welding parameters, setting of welding paths and manual control of equipment start and stop. The weld seam image processing module uses an AI image recognition algorithm, which can automatically identify weld seams of different flange specifications and different bevel forms, predict weld seam deviation trends, and send adjustment signals to the PLC central control system in advance to realize predictive adjustment of weld seam tracking and ensure that the welding trajectory accuracy is controlled within ±0.3mm.

7. A flange pipe welding method based on laser tracking and force sensing, characterized in that, The welding system and robot according to any one of claims 1-6 includes the following steps: S1. Pipe and flange clamping and positioning: Connect one end of the pipe to be welded to the flange, place the other end of the pipe between the clamping rollers (352) of the special quick-change fixture (35), start the electric push rod (343), drive the slider (342) to make the arc-shaped clamping plate (341) slide along the groove (4) of the circular docking plate (33), and cooperate with the clamping rollers (352) to achieve bidirectional clamping and positioning of the pipe from the outer periphery of the pipe, while fixing the flange to the circular docking plate (33); S2. Equipment initialization and parameter setting: Start the welding system through the human-machine interface, calibrate the laser tracking module and force sensing module, input the specifications, bevel type and preset welding parameters of the flange pipe (6) to be welded, including welding current, welding voltage, welding speed and shielding gas flow rate, and the PLC control system plans the welding path. S3. Laser tracking and positioning: The laser tracking module is activated, and the semiconductor laser projects laser stripes onto the weld area of ​​the flange and pipe. The CCD camera captures the weld image through the optical filter. The weld image processing module analyzes the image, extracts the weld center position and deviation data, and transmits it to the PLC control system. S4. Force sensing and welding start: Start the 6-DOF robotic arm (22) and rotary positioner (3). The rotary positioner (3) drives the pipe and flange to rotate synchronously. The robotic arm drives the welding torch (23) to move to the weld start position and start MIG welding. The force sensing module detects the force and torque signals received by the welding torch (23) in real time and transmits them to the PLC control system. S5. Real-time adaptive adjustment: The PLC control system adjusts the posture of the robotic arm and the position of the welding torch (23) according to the weld deviation data transmitted by the laser tracking module, ensuring that the welding torch (23) is always aligned with the center of the weld. At the same time, according to the force feedback data transmitted by the force sensing module, the welding current, voltage and welding speed are adjusted by the welding parameter adjustment module to keep the welding pressure within the preset range and avoid defects such as weld deviation, incomplete penetration or excessive deposition. S6. Welding completion and inspection: After the welding torch (23) completes all weld seams along the preset path, turn off the welding power and shielding gas, stop the operation of the robotic arm and rotary positioner (3), loosen the pipe clamping assembly (34) and the special quick-change fixture (35), and take out the welded flange pipe (6); the data storage module saves all process data of this welding and completes the welding operation.

8. The flange pipe welding method based on laser tracking and force sensing according to claim 7, characterized in that, In S2, the preset welding parameters can be adaptively matched according to the material of the pipe to be welded, the flange specifications and the bevel form. The PLC central control system calls the historical welding data in the data storage module and generates the optimal welding parameter combination through algorithm optimization. In S5, when the laser tracking module detects that the weld deviation exceeds the preset threshold, or the force sensing module detects that the welding pressure is abnormal, the PLC control system immediately issues an early warning signal and suspends the welding operation. Welding will resume after manual confirmation or automatic adjustment to the normal range. In step S6, after welding is completed, the weld seam can be visually inspected, and the inspection data is entered into the data storage module for subsequent iterative optimization of the welding process.

9. A flange pipe welding method based on laser tracking and force perception according to claim 7, characterized in that, The weld position deviation in S3 is calculated using the following function formula: In the formula: This refers to the spatial positional deviation between the welding torch and the center of the weld. The three-dimensional coordinates of the weld center detected in real time by the laser tracking module; The preset three-dimensional coordinates of the weld center; The AI ​​image recognition in the weld seam image processing module of S3 uses the improved YOLOv8 algorithm, and the weld seam feature extraction uses a convolutional neural network (CNN). The core feature extraction formula is as follows: In the formula: This is the output feature map of the l-th convolutional layer; The ReLU function is used as the activation function. ; Let be the weight matrix of the convolutional kernel in the l-th layer; This is the output feature map of the (l-1)th layer; The bias vector of the l-th layer; the weld deviation trend prediction uses an LSTM neural network, and the prediction formula is: In the formula: This is the predicted output at time t; The hidden layer state at time t-1; The laser tracking input data is given at time t; This is the weight matrix; This is the bias vector.

10. A flange pipe welding method based on laser tracking and force perception according to claim 7, characterized in that, The parameter optimization in S5 adopts a fuzzy PID algorithm based on force feedback and weld deviation. The core optimization formula is as follows: In the formula: The welding current, welding voltage, and welding speed are adjusted in real time. Preset baseline parameters; This refers to the parameter adjustment coefficient; The difference (N) between the actual welding pressure detected by the force sensing module and the preset pressure. The function is a fuzzy inference function. It maps pressure deviation and position deviation into parameter adjustment values ​​through fuzzy rules. The fuzzy rules adopt the if-then form. In step S5, a PID algorithm is used to adjust the welding torch position in real time. The expression for the PID control algorithm is: In the formula: This is the amount of adjustment for the welding torch position; This is the proportionality coefficient; This represents the real-time position deviation. The integral time constant; is the differential time constant.

Images