Submerged-arc welding laser tracking control system and method

Abstract

The invention provides a submerged-arc welding laser tracking control system and method, and relates to the technical field of welding. A submerged-arc welding trolley is provided with a line laser camera and a welding gun; in each control period, data points obtained by laser scanning of the line laser camera on the weld joint are obtained, and feature extraction is conducted on the data points to obtain weld joint features; and according to the welding seam theoretical coordinates and the welding seam characteristics, deviation correction amount is calculated to obtain the deviation correction amount of the current control period, and then the welding position of a welding gun is adjusted according to the deviation correction amount to achieve laser tracking welding. The method has the beneficial effects that in the welding process, data points obtained through laser scanning and the theoretical welding seam position are calculated to obtain the deviation correction amount to adjust the welding position of the welding gun, self-adaptive adjustment according to the deformation amount of the welding seam is achieved, welding seam forming is accurately controlled, the welding quality is stable, dependence on high-skill-level welders is reduced, and the welding quality is improved. The welding quality and efficiency are improved, the construction efficiency of a construction site is further promoted, economic benefits are increased, and the welding technology is high in applicability.

Description

Technical Field

[0001] This invention relates to the field of welding technology, and in particular to a laser tracking control system and method for submerged arc welding. Background Technology

[0002] The welding of LNG cryogenic storage tank bottom plates (5-6mm) typically employs two processes: manual shielded metal arc welding (SMAW) and submerged arc welding (SAW), or a combination of both. SMAW is the most common method for welding cryogenic storage tank bottom plates. Therefore, the welding of LNG cryogenic storage tank bottom plates has long relied on traditional manual welding. In practice, manual welding requires welders to concentrate for extended periods on meticulous operations, resulting in high labor intensity. Furthermore, it is not only inefficient, but also suffers from inconsistent weld quality due to varying skill levels among welders. With the increasing demand for LNG storage tanks in the energy industry, traditional manual welding is gradually revealing its limitations in the face of rapid construction and high-quality requirements, failing to meet the needs of industry development.

[0003] The subsequent introduction of high-efficiency submerged arc welding has a high welding efficiency. However, since the base plate material is a thin plate of 5-6mm, the large current used in submerged arc welding increases the heat input, making deformation difficult to control. Furthermore, the deformation leads to frequent adjustments of the trolley or welding torch during the process. Although it is a mechanized welding process, in actual production, due to factors such as plate length and flux, the process adjustment is frequent and difficult, and the welding quality cannot be guaranteed. Therefore, it requires experienced and highly skilled welders to operate the welding process. Summary of the Invention

[0004] To address the problems existing in the prior art, the present invention provides a submerged arc welding laser tracking control system, including a submerged arc welding carriage, wherein a line laser camera and a welding torch are mounted on the submerged arc welding carriage; The submerged arc welding laser tracking control system also includes: The weld feature extraction module is used to acquire data points obtained by the line laser camera from the laser scanning of the weld in each control cycle, and to extract features from the data points to obtain weld features. The position correction module, connected to the weld feature extraction module, is used to calculate the correction amount for the current control cycle based on the theoretical coordinates of the weld and the weld features. Then, the welding position of the welding torch is adjusted according to the correction amount to achieve laser tracking welding.

[0005] Preferably, the nozzle of the welding torch is directed toward the recognition point of the line laser camera, and the direction of travel of the submerged arc welding carriage is parallel to the direction of welding of the welding torch.

[0006] Preferably, the weld features include the center coordinates of the weld and the weld width; The weld feature extraction module includes: A gradient detection unit is used to calculate the gradient change rate at the data point when acquiring the data point in each control cycle, and then determine the edge point of the weld based on each gradient change rate. A weld fitting unit, connected to the gradient detection unit, is used to determine the left and right edges of the weld based on each edge point when at least two edge points are detected, and to obtain the center coordinates and weld width based on the left and right edges.

[0007] Preferably, the weld feature extraction module further includes a prediction unit connected to the gradient detection unit, used to select the center coordinates and the weld width from the historical control cycles as the center coordinates and the weld width of the current control cycle when less than two edge points are detected.

[0008] Preferably, the position correction module includes: The initialization unit is used to initialize the correction parameters for the current control cycle and store the historical errors calculated in the previous control cycles. A parameter adjustment unit, connected to the initialization unit, is used to calculate the current error between the theoretical coordinates of the weld and the weld features in the current control cycle, and then adjust each of the correction parameters according to the current error. The error correction calculation unit, connected to the parameter adjustment unit, is used to perform error integration based on the current error to obtain an integral term, perform error differentiation calculation based on the current error and the previous cycle error in the historical errors to obtain a differential term, and then process the adjusted error correction parameters, the current error, the integral term and the differential term to obtain the error correction amount.

[0009] Preferably, the position correction module further includes a correction limiting unit connected to the correction calculation unit, used to store a correction amount threshold, and output the correction amount threshold as the correction amount when the correction amount exceeds the threshold.

[0010] The present invention also provides a submerged arc welding laser tracking control method, which is applied to the above-mentioned submerged arc welding laser tracking control system and includes: Step S1: In each control cycle, the submerged arc welding laser tracking control system acquires data points obtained by the line laser camera scanning the weld seam, and extracts features from the data points to obtain weld seam features. In step S2, the submerged arc welding laser tracking control system calculates the correction amount for the current control cycle based on the theoretical coordinates of the weld and the characteristics of the weld. Then, it adjusts the welding position of the welding torch according to the correction amount to achieve laser tracking welding.

[0011] Preferably, the weld features include the center coordinates of the weld and the weld width; Step S1 includes: Step S11: When the submerged arc welding laser tracking control system acquires the data point in each control cycle, it calculates the gradient change rate at the data point, and then determines the edge point of the weld seam based on each gradient change rate. In step S12, when the submerged arc welding laser tracking control system detects at least two edge points, it determines the left and right edges of the weld based on each edge point, and processes the left and right edges to obtain the center coordinates and the weld width.

[0012] Preferably, step S1 further includes step S13, whereby the submerged arc welding laser tracking control system selects the center coordinates and the weld width from the historical control cycles as the center coordinates and the weld width of the current control cycle when it detects that there are fewer than two edge points.

[0013] Preferably, step S2 includes: Step S21: The submerged arc welding laser tracking control system initializes the correction parameters for the current control cycle and reads the historical errors calculated in the stored historical control cycles. Step S22: The submerged arc welding laser tracking control system calculates the current error between the theoretical coordinates of the weld and the weld features in the current control cycle, and then adjusts each of the correction parameters according to the current error. Step S23: The submerged arc welding laser tracking control system performs error integration based on the current error to obtain an integral term, performs error differentiation based on the current error and the previous cycle error in the historical error to obtain a differential term, and then processes the adjusted correction parameters, the current error, the integral term and the differential term to obtain the correction amount.

[0014] The above technical solution has the following advantages or beneficial effects: During the welding process, the welding torch position is adjusted based on the data points obtained from laser scanning of the weld seam and the theoretical weld seam position, calculating the correction amount. This allows for adaptive adjustment according to the weld seam deformation, offering advantages such as precise control of weld formation and stable welding quality. It reduces reliance on highly skilled welders and improves welding quality and efficiency. This further enhances construction efficiency on-site, increases economic benefits, and demonstrates the applicability of this welding process. Attached Figure Description

[0015] Figure 1 A schematic diagram of the process operation at the welding torch and weld seam in a preferred embodiment of the present invention; Figure 2A schematic diagram of the structure of a submerged arc welding laser tracking control system is shown in a preferred embodiment of the present invention. Figure 3 A schematic diagram of modules in a submerged arc welding laser tracking control system is provided in a preferred embodiment of the present invention. Figure 4 A flowchart illustrating a submerged arc welding laser tracking control method is provided in a preferred embodiment of the present invention. Figure 5 A flowchart illustrating step S1 is shown in a preferred embodiment of the present invention. Figure 6 The flowchart of step S2 is shown in a preferred embodiment of the present invention.

[0016] Figure reference numerals: 1. Upper 9Ni steel plate; 2. Lower 9Ni steel plate; 3. Flux nozzle; 4. Welding wire; 5. Welding direction; 6. Trolley track; 7. Welding trolley; 8. Integrated control panel; a. Extension length; b and c. Base metal thickness; d. Root pass weld; e. Weld leg size; f. Cover pass weld; g. Distance between line laser camera and bevel software identification point; h. Line laser camera; i. Software identification point; β. Welding torch angle; j. Distance between welding torch and software identification point; 100. Weld feature extraction module; 200. Position correction module. Detailed Implementation

[0017] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. The present invention is not limited to this embodiment; other embodiments that conform to the spirit of the present invention may also fall within the scope of the present invention.

[0018] In a preferred embodiment of the present invention, based on the above-mentioned problems existing in the prior art, a submerged arc welding laser tracking control system is provided, including a submerged arc welding carriage 7, on which a wire laser camera and a welding torch are provided; Submerged arc welding laser tracking control system, such as Figure 1-3 The diagram also includes: The weld feature extraction module 100 is used to acquire data points obtained by the line laser camera scanning the weld in each control cycle, and to extract features from the data points to obtain weld features. The position correction module 200 is connected to the weld feature extraction module 100. It is used to calculate the correction amount for the current control cycle based on the theoretical coordinates and features of the weld. Then, the welding position of the welding torch is adjusted according to the correction amount to achieve laser tracking welding.

[0019] In a preferred embodiment of the present invention, the nozzle of the welding torch is directed toward the identification point of the line laser camera, and the direction of travel of the submerged arc welding carriage is parallel to the direction of welding of the welding torch.

[0020] Specifically, this embodiment uses the submerged arc welding laser tracking control system applied to submerged arc lap welds with an I-groove as an example to study the weld formation, mechanical properties, and welding quality of LNG cryogenic storage tank bottom plate welds. This leads to the development of a laser tracking submerged arc welding process for LNG cryogenic storage tank bottom plates. The welding process requires no manual intervention, significantly improving automation. It solves problems such as poor weld formation, hot cracking, and incomplete root penetration caused by human error during the lap welding of LNG cryogenic storage tank bottom plates. By using a combination of software (laser tracking control) and hardware (line laser camera + submerged arc welding carriage), it automatically and effectively corrects and adjusts the X-axis (left / right of the welding torch) and Y-axis (up / down of the welding torch) during the welding process, with an error of less than 0.10mm and a correction range of ±10mm. It adaptively adjusts according to the weld deformation, offering advantages such as precise control of weld formation and stable welding quality. This reduces reliance on highly skilled welders and improves welding quality and efficiency. Ultimately, this further improves construction efficiency on-site, increases economic benefits, and provides a highly applicable welding process.

[0021] More specifically, in this embodiment, 5-6mm 9Ni steel plates were used for the experiment. The welding of the bottom plate of the LNG cryogenic storage tank was carried out using a submerged arc laser tracking control system. The results met the requirements of visual inspection, penetrant testing, and macroscopic metallography.

[0022] The embodiment includes the following steps: Bevel type: This embodiment uses a type I bevel; Then, perform beveling: use an angle grinder to clean and grind the lap joint (cleaning area 50mm from the end of the joint) until the end face shows a metallic luster; The overlapping operation involves horizontally overlapping two 9Ni steel plates, forming an overlapping joint at an overlap distance of approximately 50mm. Pre-welding setup: Adjust the distance between the trolley track and the bevel end face by adjusting the position of the welding trolley and the lap joint, so that the two are relatively parallel.

[0023] like Figure 1 , Figure 2 As shown in the attached diagram, 1 represents the upper 9Ni steel plate; 2 represents the lower 9Ni steel plate. A trolley track 6 is installed on one side of the overlapped 9Ni steel plate. The submerged arc welding trolley 7 travels along the trolley track 6 in the travel direction 5, and the trolley track 6 is parallel to the area to be welded on the overlapped 9Ni steel plate. a represents the extension length; b and c represent the base metal thickness; d represents the root pass weld; e represents the weld leg size; f represents the cover pass weld; g represents the distance between the line laser camera and the bevel software identification point; h represents the line laser camera; i represents the software identification point; β represents the welding torch angle; and j represents the distance between the welding torch and the software identification point.

[0024] The welding torch is mounted on the submerged arc welding carriage 7 and moves with the carriage to weld. In the attached figure, 3 is the flux nozzle; 4 is the welding wire; and 5 is the welding direction. The submerged arc welding carriage 7 is also equipped with an integrated control panel 8. The weld feature extraction module and the position correction module in this system are configured in the integrated control panel 8 in the form of program code. They are used to control the welding torch to perform adaptive adjustment of the X-axis (left and right of the welding torch) and Y-axis (up and down of the welding torch) at a frequency of 60HZ to correct the welding process.

[0025] After installation, adjust the welding gun position and welding length through the integrated control panel. The welding gun angle β can be fixed at 45°. The distance between the welding wire 4 and the root of the weld is about 2mm. The laser tracking software recognition point (i.e. the recognition point of the line laser camera) is located at the center of the weld. Complete the parameter settings according to the process. The integrated control panel can be started with one click, activating the laser tracking function. After the system is verified to be correct, welding will start automatically (if a fault occurs, the arc initiation function will be stopped). Welding will be completed according to the preset weld length. Since there is a 10mm gap between the welding torch and the identification point, this distance helps to control the weld deformation in real time (pre-scan). In addition, 10mm away from the arc extinguishing point, where there may be an arc initiation plate, the laser tracking function will automatically turn off. The subsequent 10mm of welding will be completed according to the weld tracking data before the arc extinguishing. The automatic shutdown of the laser tracking function can be adjusted according to the actual application.

[0026] Then, perform the root pass welding: Fill the flux hopper with flux, use 1.6mm submerged arc welding wire, and weld in constant current mode with an arc voltage of 25~28V, welding current of 220~240A, extension length of 20~30mm, and welding speed of 60~80 mm / min. Start with one key (the system can only be started after verification). The laser tracking function and welding will be automatically activated. No manual intervention is required during the welding process. After the set welding length is completed, the arc will be automatically extinguished. After the root pass welding, proceed with the cover pass welding: Keep the relative position of the welding torch and the joint unchanged, adjust the welding wire to the middle position of the root pass weld formed by the integrated control panel, and adjust the process parameters: voltage 25~30V, welding current 240~260A, extension length 15~30mm, welding speed 65~80 mm / min. After adjustment, start welding with one key.

[0027] Since the base plates are all thin plates (5-6mm), according to GB / T 26978-2021 "Design and Construction of On-Site Assembled Vertical Cylindrical Flat-Bottom Steel Cryogenic Liquefied Gas Storage Tanks", the maximum post-weld dent deformation should not exceed 1.5% of the deformation length and should not exceed 30mm. Therefore, a submerged arc DC power supply and a two-layer, two-pass welding process are used, employing a fine submerged arc welding wire (φ1.6mm) to effectively reduce heat input, thereby controlling welding deformation and ensuring weld quality and efficiency. This forms an adaptive submerged arc welding process for the cryogenic storage tank base plate welding process. It effectively reduces reliance on highly skilled welders, improves on-site construction capabilities, and enhances adaptability.

[0028] The welding process requires no manual intervention, boasts a high degree of automation, and significantly improves construction efficiency. By horizontally overlapping two 5-6mm thick 9Ni steel plates, using a φ1.6mm submerged arc welding wire and a 2-layer, 2-pass welding process, with an arc voltage of 25-30V, welding current of 200-260A, weld extension of 20-30mm, welding speed of 60-80 mm / min, and a wire angle of 45°, a full 5-6mm weld bead is formed without post-weld grinding, effectively reducing heat input and thus minimizing welding defects and deformation. Furthermore, it reduces reliance on highly skilled welders, ensuring weld quality and efficiency. Extensive testing has verified its feasibility and stability, resolving issues such as poor weld formation, hot cracking, and incomplete root fusion. It also improves the welding efficiency of 5-6mm thick base plate lap fillet welds in on-site production, significantly enhancing weld quality. This provides crucial technical support for the company's LNG storage tank project to promote automated welding, improving the company's on-site welding capabilities while reducing project costs and increasing efficiency.

[0029] Specifically, the laser tracking welding torch correction process is achieved through the following steps: Weld feature extraction → Position deviation calculation → Adaptive PID control → Actuator output. This is primarily based on software implementation configured in the integrated control panel. Several examples are provided below, along with software code, to illustrate the implementation process.

[0030] Specifically, in this embodiment, program initialization is performed first: pid_x = AdaptivePIDController() pid_y = AdaptivePIDController() Used to initialize the adaptive PID controller pid_x, pid_y (X-axis and Y-axis independent control); target_pos = (0, 0) Used to initialize the theoretical center position of the weld (preset target (X, Y) - (0, 0)); while welding: This indicates that the process will be repeated cyclically as welding continues. raw_data = get_laser_sensor_data() raw_data is used to read weld contour point cloud data from a line laser camera; processed_data = preprocess_laser_data(raw_data) This is used to filter and reduce noise in weld contour point cloud data to obtain preprocessed weld contour point cloud data (processed_data), and to extract the effective weld region. center_x,center_y,width=extract_weld_seam_features(processed_data) This function calculates the weld center coordinates (center_x, X-axis coordinate; center_y, Y-axis coordinate) and weld width based on the preprocessed weld contour point cloud data. The extract_weld_seam_features() function represents a pre-packaged weld feature extraction algorithm. The input of this function is the preprocessed laser data points, and the output is the weld center position and weld width. dt = get_time_interval() Used to obtain the control cycle time dt (approximately 16ms, corresponding to a 60Hz frequency). correction_x = pid_x.compute(target_pos[0], center_x, dt) is used to calculate the X-axis correction amount correction_x (left and right directions), where target_pos[0] represents the Y coordinate of the initial theoretical center position of the weld; correction_y = pid_y.compute(target_pos[1], center_y, dt) is used to calculate the Y-axis correction amount correction_y (up and down direction), where target_pos[1] represents the Y coordinate of the initial theoretical center position of the weld; The compute() function represents a pre-packaged correction calculation algorithm. The inputs to this function are the theoretical weld coordinates target_pos, the weld feature center_x or center_y, and the control period dt. The output is the X-axis correction amount correction_x or the Y-axis correction amount correction_y.

[0031] In a preferred embodiment of the present invention, the weld features include the center coordinates of the weld and the weld width; The weld feature extraction module 100 includes: The gradient detection unit 110 is used to calculate the gradient change rate at the data point when acquiring the data point in each control cycle, and then determine the edge point of the weld based on each gradient change rate. The weld fitting unit 120 is connected to the gradient detection unit 110. When at least two edge points are detected, the left and right edges of the weld are determined based on each edge point, and the center coordinates and weld width are obtained by processing the left and right edges.

[0032] In this embodiment, weld feature extraction is implemented based on the extract_weld_seam_features() function, the specific content of which is as follows: gradients = np.gradient(data_points) The gradient is used to calculate the first derivative (gradient rate of change) of the data points (data_points). The gradient of each data point (i.e., the height data of the weld cross-section profile) is calculated using NumPy's `gradient` function. The gradient reflects the rate of change of the data points; the absolute value of the gradient is larger at the edges.

[0033] edge_points = find_peaks(abs(gradients), prominence=0.3) Used to find peaks in the gradient absolute value sequence (i.e., locations where the gradient changes drastically, corresponding to the edge points of the weld). Prominence=0.3 sets the minimum prominence of the peak to 0.3 (filtering out small fluctuations and ensuring a clear edge).

[0034] Subsequently, weld center fitting was performed based on the least squares method; if len(edge_points) >= 2: If the number of detected edge points is greater than or equal to 2 (i.e., at least two edge points, representing the left and right edges), then the following steps are performed: left_edge = min(edge_points) Take the minimum value of all edge point indices as the left edge position left_edge.

[0035] right_edge = max(edge_points) Take the maximum value of all edge point indices as the right edge position, right_edge.

[0036] center_x = (left_edge + right_edge) / 2 Calculate the x-coordinate of the weld center, center_x (center in the x-direction), which is the midpoint between the left and right edges.

[0037] center_y = np.mean(data_points[left_edge:right_edge]) Calculate the ordinate center_y (center in the y-direction) of the weld center region, which is the average of all data points (height values) from the left edge to the right edge. This represents the average height of the weld center.

[0038] width = right_edge - left_edge The weld width is the index of the right edge minus the index of the left edge (the unit is the number of data points; the actual width needs to be multiplied by the point spacing to get the physical width).

[0039] In a preferred embodiment of the present invention, the weld feature extraction module further includes a prediction unit connected to the gradient detection unit, which is used to select the center coordinates and weld width from the historical control cycles as the center coordinates and weld width of the current control cycle when less than two edge points are detected.

[0040] In the `extract_weld_seam_features()` function, if fewer than two edge points are detected (e.g., only one edge is detected or no edge is detected due to interference), the `else` branch is executed: else: center_x,center_y,width = predict_from_history() This indicates a call to the `predict_from_history()` function to predict the current weld center position (center_x), center_y, and weld width (width) based on historical data. This is typically used when reliable detection in the current frame is not possible, employing data from the previous frame or several frames to ensure continuity.

[0041] In a preferred embodiment of the present invention, the position correction module 200 includes: The initialization unit 210 is used to initialize the correction parameters of the current control cycle and store the historical errors calculated in the historical control cycles. The parameter adjustment unit 220 is connected to the initialization unit 210 and is used to calculate the current error between the theoretical coordinates of the weld and the characteristics of the weld in the current control cycle, and then adjust each correction parameter according to the current error. The correction calculation unit 230 is connected to the parameter adjustment unit 220. It is used to perform error integration based on the current error to obtain the integral term, perform error differentiation calculation based on the current error and the previous cycle error in the historical error to obtain the differential term, and then process the adjusted correction parameters, the current error, the integral term and the differential term to obtain the correction amount.

[0042] In this embodiment, the correction amount is calculated based on the compute() function. This correction amount calculation is an adaptive PID process, so the PID parameters are adjusted before executing the compute() function, specifically including: 1. PID correction parameter initialization function: def __init__(self): self.Kp = 0.8 indicates the initial value of the scaling factor, which is used to quickly respond to weld offset (such as sudden deformation). self.Ki = 0.05 indicates the initial value of the integral coefficient. A small integral coefficient prevents oversaturation during thin plate welding. self.Kd = 0.2 indicates setting the initial value of the differential coefficient to suppress welding torch oscillation (avoid mechanical vibration); `self.last_error = 0` means storing the previous error value. `self.integral = 0` indicates the cumulative value of the integral term. `self.error_history = []` represents a sequence of historical errors (used for parameter adaptation), recording the most recent 10 errors (approximately 0.16 seconds of data at 60Hz sampling). 2. Adaptive method for correction parameters, adaptive parameter adjustment function def update_gains(self, error): if len(self.error_history) >= 3: This indicates that at least three historical points are needed to analyze trends; error_diff = abs(error - self.error_history[-1]) This represents the absolute difference between the current error (error) and the most recent error (self.error_history[-1]) (i.e., the most recent change in error). if error_diff > 0.5: If the error variation is large (exceeding the threshold of 0.5), then the proportional control is strengthened and the integral control is weakened. self.Kp = 1.2 This indicates that the proportional coefficient has been adjusted from 0.8 to 1.2, enhancing proportional control (fast response). self.Ki = 0.02 This indicates that the integral coefficient has been adjusted from 0.05 to 0.02 to reduce the integral action (prevent overshoot). self.Kd = 0.3 This indicates that the differential coefficient has been adjusted from 0.2 to 0.3, which enhances differential suppression (reduces oscillations). else: When the error change is small, enhance integral control (reduce proportional and derivative control); self.Kp = 0.6 This indicates that the scaling factor has been adjusted from 0.8 to 0.6 to reduce the scaling ratio (to avoid overshoot). self.Ki = 0.1 This indicates that the integral coefficient has been adjusted from 0.05 to 0.1 to enhance integration (eliminate static bias). self.Kd = 0.1 This indicates that the differential coefficient is adjusted from 0.2 to 0.1, thus reducing the differential (smoothing adjustment). self.error_history.append(error) This indicates that the current error (error) is recorded in the history queue (error_history). if len(self.error_history) > 10: self.error_history.pop(0) This means that if the error_history record exceeds 10, the oldest record will be removed (keeping a maximum of 10 records). 3. Core steps in PID calculation: compute() function def compute(self, setpoint, current_pos, dt): error = setpoint - current_pos This means calculating the current error: error = target position setpoint - actual position current_pos; (In `correction_x = pid_x.compute(target_pos[0], center_x, dt)`, `target_pos[0]` represents the Y-coordinate of the initial theoretical center position of the weld as the parameter of the `compute()` function.) The target position is setpoint; center_x is the x-coordinate of the weld center in the weld feature, which is used as the actual position current_pos in the compute() function. self.update_gains(error) This indicates that the `update_gains()` function, which automatically adjusts the execution parameters, is called. Next, integral anti-saturation processing is performed (to prevent excessive accumulation). if abs(self.integral + error*dt) < 10: This means that integration is only performed when the absolute value of the cumulative value of the integral term (self.integral plus the current error term error*dt) is less than 10, to prevent the integral term from becoming too large and causing system instability.

[0043] self.integral += error * dt This represents the calculation process of the integral term, where self.integra is the integral term, error is the current error, and dt is the control period. derivative = (error - self.last_error) / dt This represents the rate of change of the differential term, i.e., the error, where the differential term is the current error, the error of the previous control cycle is self.last_error, and the control cycle is dt. output=self.Kp*error+self.Ki*self.integral+ self.Kd*derivative This represents the calculation of the correction factor output, where self.Kp is the proportional coefficient, self.Ki is the integral coefficient, self.Kd is the differential coefficient, error is the current error, self.integra is the integral term, and derivative is the differential term.

[0044] self.last_error = error This indicates that the "previous error" is updated to the current value.

[0045] Through adaptive adjustment, this PID controller can adopt a more aggressive control strategy when the weld deformation is severe (such as sudden misalignment), and a more delicate control strategy when tracking smoothly, thereby improving system stability while ensuring accuracy.

[0046] In a preferred embodiment of the present invention, the position correction module further includes a correction limiting unit connected to the correction calculation unit, which is used to store the correction amount threshold and output the correction amount threshold as the correction amount when the correction amount exceeds the threshold.

[0047] Specifically, the compute() function also includes: output = np.clip(output, -10, 10) This indicates the output amplitude limit (within ±10mm).

[0048] The present invention also provides a submerged arc welding laser tracking control method, which is applied to the above-mentioned submerged arc welding laser tracking control system and includes: Step S1: The submerged arc welding laser tracking control system acquires data points obtained by the line laser camera scanning the weld seam in each control cycle, and extracts features from the data points to obtain weld seam features. In step S2, the submerged arc welding laser tracking control system calculates the correction amount for the current control cycle based on the theoretical coordinates and characteristics of the weld seam. Then, it adjusts the welding position of the welding torch according to the correction amount to achieve laser tracking welding.

[0049] In a preferred embodiment of the present invention, the weld features include the center coordinates of the weld and the weld width; Step S1 includes: Step S11: When the submerged arc welding laser tracking control system acquires data points in each control cycle, it calculates the gradient change rate at the data points and then determines the edge points of the weld seam based on each gradient change rate. In step S12, when the submerged arc welding laser tracking control system detects at least two edge points, it determines the left and right edges of the weld based on each edge point, and obtains the center coordinates and weld width based on the left and right edges.

[0050] In a preferred embodiment of the present invention, step S1 further includes step S13, in which the submerged arc welding laser tracking control system selects the center coordinates and weld width from the historical control cycles as the center coordinates and weld width of the current control cycle when it detects less than two edge points.

[0051] In a preferred embodiment of the present invention, step S2 includes: Step S21: The submerged arc welding laser tracking control system initializes the correction parameters for the current control cycle and reads the historical errors calculated in the stored historical control cycles. Step S22: The submerged arc welding laser tracking control system calculates the current error between the theoretical coordinates of the weld and the characteristics of the weld in the current control cycle, and then adjusts each correction parameter according to the current error. Step S23: The submerged arc welding laser tracking control system performs error integration based on the current error to obtain the integral term, performs error differentiation calculation based on the current error and the previous cycle error in the historical error to obtain the differential term, and then processes the adjusted correction parameters, the current error, the integral term and the differential term to obtain the correction amount.

[0052] The above are merely preferred embodiments of the present invention and are not intended to limit the implementation methods and protection scope of the present invention. Those skilled in the art should recognize that any equivalent substitutions and obvious changes made using the content of this specification and illustrations should be included within the protection scope of the present invention.

Claims

1. A laser tracking control system for submerged arc welding, characterized in that, Includes a submerged arc welding carriage, which is equipped with a wire laser camera and a welding torch; The submerged arc welding laser tracking control system also includes: The weld feature extraction module is used to acquire data points obtained by the line laser camera from the laser scanning of the weld in each control cycle, and to extract features from the data points to obtain weld features. The position correction module, connected to the weld feature extraction module, is used to calculate the correction amount for the current control cycle based on the theoretical coordinates of the weld and the weld features. Then, the welding position of the welding torch is adjusted according to the correction amount to achieve laser tracking welding.

2. The submerged arc welding laser tracking control system according to claim 1, characterized in that, The nozzle of the welding torch is directed toward the recognition point of the line laser camera, and the direction of travel of the submerged arc welding carriage is parallel to the direction of welding of the welding torch.

3. The submerged arc welding laser tracking control system according to claim 1, characterized in that, The weld features include the center coordinates of the weld and the weld width; The weld feature extraction module includes: A gradient detection unit is used to calculate the gradient change rate at the data point when acquiring the data point in each control cycle, and then determine the edge point of the weld based on each gradient change rate. A weld fitting unit, connected to the gradient detection unit, is used to determine the left and right edges of the weld based on each edge point when at least two edge points are detected, and to obtain the center coordinates and weld width based on the left and right edges.

4. The submerged arc welding laser tracking control system according to claim 3, characterized in that, The weld feature extraction module further includes a prediction unit connected to the gradient detection unit, used to select the center coordinates and the weld width from the historical control cycles as the center coordinates and the weld width of the current control cycle when less than two edge points are detected.

5. The submerged arc welding laser tracking control system according to claim 1, characterized in that, The position correction module includes: The initialization unit is used to initialize the correction parameters for the current control cycle and store the historical errors calculated in the previous control cycles. A parameter adjustment unit, connected to the initialization unit, is used to calculate the current error between the theoretical coordinates of the weld and the weld features in the current control cycle, and then adjust each of the correction parameters according to the current error. The error correction calculation unit, connected to the parameter adjustment unit, is used to perform error integration based on the current error to obtain an integral term, perform error differentiation calculation based on the current error and the previous cycle error in the historical errors to obtain a differential term, and then process the adjusted error correction parameters, the current error, the integral term and the differential term to obtain the error correction amount.

6. The submerged arc welding laser tracking control system according to claim 5, characterized in that, The position correction module further includes a correction limiting unit connected to the correction calculation unit, which stores a correction amount threshold and outputs the correction amount threshold as the correction amount when the correction amount exceeds the threshold.

7. A laser tracking control method for submerged arc welding, characterized in that, The submerged arc welding laser tracking control system according to any one of claims 1-6 comprises: Step S1: In each control cycle, the submerged arc welding laser tracking control system acquires data points obtained by the line laser camera scanning the weld seam, and extracts features from the data points to obtain weld seam features. In step S2, the submerged arc welding laser tracking control system calculates the correction amount for the current control cycle based on the theoretical coordinates of the weld and the characteristics of the weld. Then, it adjusts the welding position of the welding torch according to the correction amount to achieve laser tracking welding.

8. The submerged arc welding laser tracking control method according to claim 7, characterized in that, The weld features include the center coordinates of the weld and the weld width; Step S1 includes: Step S11: When the submerged arc welding laser tracking control system acquires the data point in each control cycle, it calculates the gradient change rate at the data point, and then determines the edge point of the weld seam based on each gradient change rate. In step S12, when the submerged arc welding laser tracking control system detects at least two edge points, it determines the left and right edges of the weld based on each edge point, and processes the left and right edges to obtain the center coordinates and the weld width.

9. The submerged arc welding laser tracking control method according to claim 8, characterized in that, Step S1 further includes step S13, whereby the submerged arc welding laser tracking control system selects the center coordinates and the weld width from the historical control cycles as the center coordinates and the weld width of the current control cycle when it detects that there are fewer than two edge points.

10. The submerged arc welding laser tracking control method according to claim 7, characterized in that, Step S2 includes: Step S21: The submerged arc welding laser tracking control system initializes the correction parameters for the current control cycle and reads the historical errors calculated in the stored historical control cycles. Step S22: The submerged arc welding laser tracking control system calculates the current error between the theoretical coordinates of the weld and the weld features in the current control cycle, and then adjusts each of the correction parameters according to the current error. Step S23: The submerged arc welding laser tracking control system performs error integration based on the current error to obtain an integral term, performs error differentiation based on the current error and the previous cycle error in the historical error to obtain a differential term, and then processes the adjusted correction parameters, the current error, the integral term and the differential term to obtain the correction amount.

Images