An adaptive welding control system and method for rotating pipe submerged arc welding based on dual monitoring of straightness and weld temperature
By monitoring the straightness of the pipe and the temperature of the weld in real time during submerged arc welding of rotating pipes, and adjusting the welding parameters in combination with a mathematical model, the problems of thermal deformation and defects in the pipe welding process were solved, and real-time control and quality improvement of the welding process were achieved.
Patent Information
- Application Number
- CN202611504275.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-09-24
- Publication Date
- 2026-11-17
Smart Images

Figure CN123103127A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of pipeline welding technology, and specifically to an adaptive welding control system and method for rotary pipeline submerged arc welding based on dual monitoring of straightness and weld temperature. Background Technology
[0002] Fixed-machine, rotating-pipe circumferential submerged arc welding is the mainstream process in the field of automated welding of thick-walled pressure pipelines. In this process, the submerged arc welding machine remains stationary, while the pipeline is driven to rotate at a uniform speed by a roller frame or chuck to complete multi-layer, multi-pass circumferential welding. It has advantages such as uniform weld formation, large penetration depth, and high degree of automation, and is widely used in the circumferential welding production of power plant boilers, pressure vessels, long-distance oil and gas pipelines, and nuclear power pressure pipelines. However, during the rotating welding process, the electric arc always acts on the same side of the pipeline. Multi-layer, multi-pass welding causes heat to continuously accumulate locally in the pipeline. As the number of welding layers increases, the temperature of the heated area on one side of the pipeline continues to rise. Thermal expansion, constrained by the rigidity of the pipeline itself, is converted into compressive plastic strain, which, upon cooling, forms irreversible plastic deformations such as axial bending and warping. Under current production conditions, pipeline straightness can only be statically re-measured manually after welding. The development trend of micro-deformation cannot be captured in real time during the welding process. After the deformation solidifies, remedial measures such as cutting and rework or flame straightening can only be taken, resulting in persistently high pipeline scrap rates and manual rework costs. During submerged arc welding, the peak temperature of the molten pool, the temperature of the heat-affected zone, and the interpass temperature directly determine the amount of thermal deformation of the pipeline, the metallographic structure of the weld, and the tendency for cold and hot cracking. Traditional processes use fixed welding current, arc voltage, and pipeline rotation speed throughout the process, making it impossible to dynamically match the welding heat input according to the real-time weld temperature. Excessively high temperatures will exacerbate pipeline bending deformation and lead to coarse grains, while excessively low temperatures are prone to defects such as incomplete fusion and slag inclusions. Summary of the Invention
[0003] To address the shortcomings and problems of existing fixed-machine, rotary submerged arc welding processes that cannot monitor pipe bending deformation and weld temperature in real time, and whose fixed process parameters lead to excessive pipe straightness deviations and frequent thermal defects, this invention provides an adaptive welding control system and method for rotary submerged arc welding based on dual monitoring of straightness and weld temperature. By simultaneously collecting dynamic straightness and plastic deformation data of the pipe, temperature field data of the entire weld area, and visual data of molten pool formation, and relying on a deformation-temperature coupling mathematical model to predict the trend of pipe thermal deformation, the system adjusts the submerged arc welding current, voltage, and pipe rotation speed in a graded and coordinated manner to achieve integrated temperature and deformation control. The system is equipped with active purge protection hardware with optical lenses to adapt to high dust and strong arc light conditions, and integrates fault interlock shutdown, MES data traceability, and a multi-material process database. This solves the problems of excessive pipe bending deviations, temperature runaway, and frequent welding defects, while also meeting the requirements of digital production management for special equipment.
[0004] The solution adopted by this invention to solve its technical problem is: an adaptive welding control system and method for submerged arc welding of rotating pipes based on dual monitoring of straightness and weld temperature, including a pipe reference calibration module, used to construct the three-dimensional absolute reference axis of the pipe before welding and output the initial reference coordinates of the pipe, and lock the reference parameters after calibration; a pipe straightness dynamic detection module, which arranges at least three sets of laser displacement sensor arrays along the pipe axis to collect the radial offset of the pipe wall in real time during the pipe rotation welding process, and outputs the real-time straightness error, bending vector and deformation growth rate; a weld infrared real-time temperature measurement module, used to synchronously collect the temperature data of the weld area; and a visual monitoring module for the molten pool and deposited metal, using... The system includes: a real-time acquisition module for molten pool morphology data; a multi-parameter data fusion and analysis module with a built-in deformation-temperature coupled mathematical model to fuse straightness error, temperature data, molten pool morphology data, and welding electrical parameter data to predict the subsequent bending deformation trend of the pipeline and classify deformation and temperature anomaly levels; a threshold judgment module with pre-stored temperature and straightness deformation thresholds for different pipe materials, diameters, and wall thicknesses; a submerged arc welding current and voltage adjustment module and a pipeline rotation speed frequency conversion control module to adjust the welding current, arc voltage, and pipeline rotation speed in conjunction with the deformation and temperature anomaly levels; and an early warning, storage, and traceability module for outputting early warning signals, executing fault interlock shutdowns, and storing full-process data.
[0005] Furthermore, the pipeline reference calibration module includes a cross laser reference emitting device, a laser receiver, an attitude locking bracket, a coordinate fitting calculation unit, and a reference parameter storage and locking unit; before welding, the pipeline rotates one revolution and automatically collects the distance measurement data of the outer wall at each predetermined rotation angle, fits the standard reference axis of the pipeline by the least squares method, and locks the reference parameters after calibration.
[0006] Furthermore, the pipeline straightness dynamic detection module incorporates a composite filtering algorithm combining time-domain mean filtering and frequency-domain vibration filtering to filter out pipeline rotational periodic jumps, elliptical eccentricity, instantaneous thermal expansion elastic deformation, and mechanical vibration interference, and outputs plastic permanent bending deformation data.
[0007] Furthermore, the lens end of the weld infrared real-time temperature measurement module is equipped with an active lens cleaning protection module, which includes an annular air curtain nozzle, an air source pressure regulating valve, a pulse timing controller, a dust scraper ring, and a narrow band filter, used for continuous dust prevention, light shielding, and anti-splashing during the welding process.
[0008] Furthermore, the input variables of the deformation-temperature coupled mathematical model include real-time straightness deviation, deformation growth rate, peak temperature of the molten pool, interpass temperature, temperature rise rate, pipe wall thickness, pipe linear expansion coefficient, and total pipe length; the model pre-adjusts parameters by pre-setting the welding cycle based on the predicted deformation amount.
[0009] Furthermore, the threshold judgment module is divided into four levels of linkage control strategy, including slight abnormality, moderate abnormality, severe abnormality and insufficient fusion formation. Each level is set with welding current adjustment range, arc voltage adjustment range and pipe rotation speed adjustment range. When the temperature and deformation recover to the safe range, they are gradually raised to the standard process parameters in multiple adjustment cycles, and the single parameter adjustment range does not exceed 10%.
[0010] Furthermore, the early warning storage and traceability module has built-in fault interlocking logic. When sensor signals are lost, deformation continues to exceed the predetermined time, or the molten pool collapses beyond the threshold, the welding power supply is automatically cut off and the pipeline is controlled to decelerate and stop.
[0011] Furthermore, the system has a built-in multi-material process database including carbon steel, chromium-molybdenum heat-resistant steel and stainless steel, and is suitable for circumferential submerged arc welding of straight pipes and bends with pipe diameters of φ50 to φ1500mm and wall thicknesses of 8 to 120mm.
[0012] An adaptive welding control method for submerged arc welding of rotating pipes based on dual monitoring of straightness and weld temperature, implemented using the aforementioned control system, includes the following steps: S1. Pre-welding reference calibration and process initialization: Start the pipeline reference calibration module, rotate the pipeline one revolution to fit the three-dimensional absolute reference axis and lock the reference parameters; according to the pipe material, pipe diameter and wall thickness, retrieve the built-in process database, preset the temperature threshold and straightness deformation threshold, and initialize the standard welding current, arc voltage and pipeline rated rotation speed. S2. Welding Start: Drive the pipeline to rotate at a constant speed to carry out circumferential submerged arc welding; S3. Full-process synchronous multi-parameter acquisition: The laser sensor array acquires the axial straightness deformation data of the pipeline in real time, the infrared temperature measurement unit acquires the temperature data of the weld area in real time, and the visual monitoring unit acquires the molten pool morphology data and acquires the real-time welding current and voltage data in real time. S4. Multi-parameter coupled modeling and predictive analysis: Input the data collected in step S3 into the deformation-temperature coupled mathematical model, calculate the temperature rise rate and the plastic deformation growth rate, predict the subsequent bending deformation trend of the pipeline, and determine the deformation level and temperature anomaly level. S5. Graded closed-loop adaptive control: Based on the grade determined in step S4, the welding current, arc voltage and pipe rotation speed are adjusted in a linked manner. S6. Gradual parameter recovery: When the weld temperature drops to a safe range and the straightness deformation rate approaches zero, the parameters are gradually restored to the standard process parameters in multiple cycles. S7. Cyclic closed-loop monitoring and control: Repeat steps S3 to S6 until the welding process is completed; S8. Data storage and traceability: Stores data throughout the entire process and supports exporting and uploading.
[0013] Furthermore, in step S5, when there is moderate or severe deformation or temperature deviation, a coupled correction strategy of reducing welding heat input and reducing pipe rotation speed is adopted; in step S3, the acquisition stage adopts a composite algorithm of time-domain mean filtering and frequency-domain vibration filtering to filter out detection noise caused by flux dust scattering, rotating mechanical vibration and instantaneous arc thermal expansion, and distinguish between elastic instantaneous offset and permanent plastic deformation.
[0014] The beneficial effects of this invention are: This invention integrates pipeline straightness deformation and weld temperature into a unified closed-loop control framework. By establishing a three-dimensional absolute reference axis before welding and locking it throughout the welding process, and using a multi-section orthogonal laser array to dynamically acquire radial offset of the pipe wall, combined with a composite filtering algorithm to eliminate interference factors such as rotational eccentricity, mechanical vibration, and instantaneous thermal expansion, the system can accurately identify permanent plastic deformation and its development trend, providing a reliable data foundation for parameter adjustment. This invention also adds a narrow-band filter to the front of the infrared temperature measuring lens to shield against arc light interference, and simultaneously deploys an annular air curtain nozzle. During welding, a low-pressure air curtain is continuously blown to isolate dust and spatter. Every 30 seconds, a high-pressure pulse air blowing mechanism is automatically triggered, along with a silicone dust scraper ring to wipe the optical window. This active protection mechanism ensures long-term stable operation of the sensor in the submerged arc welding field, making online monitoring possible from the laboratory to the production line. The deformation-temperature coupled mathematical model of this invention incorporates parameters such as real-time straightness deviation, temperature rise rate, pipe linear expansion coefficient, wall thickness, and pipe length into a unified calculation framework. It predicts deformation and triggers parameter adjustments one to two welding cycles in advance, achieving preventative intervention rather than post-event compensation. The four-level control strategy covers different operating conditions from minor to severe anomalies. In cases of moderate to severe anomalies, it simultaneously reduces welding heat input and pipe rotation speed, both reducing heat supply and extending heat dissipation time, thus blocking the chain reaction path of high-temperature accumulation inducing plastic bending. The system also integrates multiple interlocking protections for sensor failure, over-temperature / over-range, and molten pool collapse. In critical conditions, it automatically cuts off power and decelerates to stop the machine, while simultaneously locking fault data snapshots, improving the safety of equipment and pipelines. Attached Figure Description
[0015] Figure 1 This is a flowchart of the multi-parameter adaptive control of the present invention; Figure 2 This is a schematic diagram of the system structure principle of the present invention; Figure 3 This is a schematic diagram of the system hardware layout structure of the present invention; Figure 4 This is a cross-sectional view of the lens active sweep protection module of the present invention; Figure 5 This is a schematic diagram illustrating the pipeline reference calibration principle of the present invention; Figure 6 This is a schematic diagram of the deformation-temperature coupling control logic of the present invention; Figure 7 This is a schematic diagram of the fault interlocking protection logic of the present invention; Figure 8 This is a schematic diagram of the structure of Embodiment 4 of the present invention.
[0016] In the diagram: 1. Submerged arc welding machine; 2. Variable frequency speed control motor; 3. Chuck; 4. Roller frame; 5. Infrared temperature probe; 501. Optical window; 502. Narrow band filter; 503. Annular air curtain nozzle; 504. Air source pressure regulating valve; 505. Pulse timing controller; 506. Silicone dust scraper ring; 507. Laser light shield; 6. Laser displacement sensor array; 7. Calibration point; 8. Cross laser reference emitting device; 9. Laser receiver; 10. Attitude locking bracket; 11. Clamp; 12. Grating displacement sensor. Detailed Implementation
[0017] The following is in conjunction with the appendix Figure 1-8 Specific embodiments of the present invention will be described in detail below. It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of the invention. Example
[0018] like Figure 1 , Figure 2 As shown, the present invention provides an adaptive welding control system for submerged arc welding of rotating pipes based on dual monitoring of straightness and weld temperature. It mainly includes a pipe reference calibration module, a pipe straightness dynamic detection module, a weld infrared real-time temperature measurement module, a molten pool and deposited metal visual monitoring module, a multi-parameter data fusion analysis module, a threshold judgment module, a submerged arc welding current and voltage adjustment module, a pipe rotation speed frequency conversion control module, an early warning storage and traceability module, and a lens active purging protection module.
[0019] like Figure 2 As shown, the submerged arc welding machine 1 is fixed to one side of the welding station. The pipe to be welded is supported by a roller frame 4 or a chuck 3, and is driven by a variable frequency speed-regulating motor 2 to rotate the pipe at a uniform speed around its own axis, completing the multi-layer, multi-pass submerged arc welding operation of the circumferential seam. During operation, the system collects multi-dimensional data in real time through various sensor detection modules. After analysis and processing by the main control module, control commands are output to drive the execution module to complete the dynamic adaptive adjustment of welding parameters.
[0020] The pipeline reference calibration module consists of a central horizontal cross laser reference transmitter 8, pipeline head and tail laser receivers 9, a high-precision attitude locking bracket 10, a coordinate fitting calculation unit, and a reference parameter storage and locking unit. Before welding, the pipeline reference calibration operation is performed first. After the pipeline to be welded is hoisted and positioned on the roller frame 4, the roller frame 4 is driven to zero and reset. The central horizontal cross laser reference transmitter 8 is then activated, emitting a collimated cross laser towards the head and tail of the pipeline. The laser receivers 9 at both ends of the pipeline receive the laser signals and complete the initial alignment calibration, ensuring that the laser reference line roughly coincides with the theoretical axis of the pipeline. The high-precision attitude locking bracket 10 remains fixed throughout the calibration and subsequent welding process, providing a stable installation foundation for the system that is not affected by equipment vibration, pipeline thermal expansion, or installation eccentricity.
[0021] Subsequently, the drive roller frame 4 rotates the pipe around its own axis at a uniform speed for one revolution. Every 1° of rotation, the system automatically collects radial distance data between the outer surface of the pipe wall and the laser reference. After the pipe completes one revolution, a total of 360 sets of circumferential distance data are collected. The coordinate fitting calculation unit receives all the above distance data and uses the least squares method to fit and calculate the spatial reference axis of the pipe's outer contour. Calibration and verification are completed at four verification points 77 at 0°, 90°, 180°, and 270° along the pipe's circumference. This fitting process automatically eliminates initial errors caused by the parallelism deviation of the roller frame 4, the initial bending of the pipe, and installation eccentricity, outputting the standard three-dimensional absolute reference axis coordinates of the pipe under no-welding-heat-input conditions. After calibration, the reference parameter storage and locking unit locks and stores the fitted three-dimensional absolute reference axis coordinate data. Throughout the entire welding process, the reference parameters remain fixed and are not refreshed, unaffected by welding thermal expansion, equipment vibration, or other factors that could cause drift.
[0022] The pipeline straightness dynamic detection module is equipped with at least three sets of laser displacement sensor arrays 6 fixedly arranged along the pipeline axis. Each sensor array contains two sets of orthogonally arranged two-dimensional laser displacement sensors, corresponding to the X and Y directions respectively, to achieve two-dimensional synchronous detection of the radial offset of the pipe wall. Each sensor frame is independently fixed on the ground or a dedicated support and does not rotate with the pipeline. The sensor detection end face faces the outer wall of the pipeline to ensure no detection blind spots during continuous pipeline rotation. During welding, the pipeline rotates at a uniform speed, and the laser displacement sensors at each cross-section continuously collect radial distance data from the outer surface of the pipe wall to the sensor at sampling intervals of 1°. After the pipeline completes one full rotation, 360 sets of radial distance measurement data for that cross-section are obtained. The data collected synchronously at each cross-section is transmitted in real time to the multi-parameter data fusion analysis module for processing.
[0023] It should be noted that the pipeline is subject to various interferences during rotary welding, including the pipeline's own elliptic eccentricity, periodic vibrations caused by rotation, mechanical vibrations of the roller frame 4, and instantaneous thermal expansion elastic deformation caused by welding heat input. These interferences all cause fluctuations in the radial offset of the pipe wall, but are not true permanent plastic bending deformation. To accurately extract the permanent plastic bending deformation data, the pipeline straightness dynamic detection module incorporates a composite filtering algorithm combining time-domain averaging and frequency-domain vibration filtering. Specifically, the periodic average of 360 sets of radial distance measurement data collected during the pipeline's full rotation is taken to filter out the periodic vibration interference caused by the pipeline's ellipticity and rotational eccentricity. Based on this, a frequency-domain filtering algorithm is used to further separate the high-frequency noise component caused by mechanical vibration and the low-frequency component caused by instantaneous thermal expansion elastic deformation. After the above composite filtering process, the slowly accumulating spatial offset of the pipeline axis is extracted, which is the permanent plastic bending deformation caused by welding heat accumulation.
[0024] The multi-parameter data fusion analysis module uses the least squares method to refit the current spatial position of the pipe axis based on the pipe wall radial offset data obtained after filtering and processing of each section. This is then compared with the three-dimensional absolute reference axis locked before welding to calculate the real-time straightness error in mm / m, the bending vector direction, and the deformation growth rate in mm / s. This processing ensures that the system can accurately distinguish between elastic thermal deformation, mechanical vibration, and permanent plastic bending, avoiding misjudging instantaneous elastic offset as plastic deformation and triggering unnecessary parameter adjustments.
[0025] The weld seam infrared real-time temperature measurement module is equipped with a colorimetric infrared temperature probe 5, which is positioned near the head of the submerged arc welding machine. The optical window 501 of the temperature probe is aligned with the weld seam area. During the welding process, this module simultaneously collects the peak temperature of the molten pool, the temperature of the heat-affected zone, and the interpass temperature during multi-layer, multi-pass welding, monitoring the heat accumulation status in real time and identifying abnormal conditions such as excessive temperature, excessively rapid temperature rise, and localized heat concentration. The submerged arc welding environment contains a large amount of flux dust, high-temperature metal spatter, and strong visible arc radiation. The optical window 501 of conventional infrared temperature measurement equipment is easily blocked by dust or damaged by spatter, and the strong arc light can severely interfere with the temperature measurement accuracy. Therefore, this invention includes an active lens cleaning protection module at the lens end of the weld seam infrared real-time temperature measurement module. Figure 4As shown, the lens active purge protection module consists of an annular high-pressure air curtain nozzle, an air source pressure regulating valve 504, a pulse timing controller 505, a silicone rotating dust scraper ring, a narrow-band infrared filter, and a laser light-shielding plate 507. The annular high-pressure air curtain nozzle is fitted around the optical window 501 of the infrared temperature measuring probe 5. The air source pressure regulating valve 504 is connected to an external compressed air source and regulates the air supply pressure. During the welding process, the air curtain nozzle continuously sprays a low-pressure air curtain, forming a stable gas barrier in front of the optical window 501, effectively isolating flux dust and high-temperature metal spatter particles.
[0026] Based on this, the pulse timing controller 505 triggers high-pressure pulse blowing at a frequency of once every 30 seconds, instantly enhancing the air curtain blowing force and blowing away the fine dust adhering to the surface of the optical window 501. At the same time, the silicone rotating dust scraper ring rotates and wipes the surface of the optical window 501 in conjunction with the pulse blowing, further removing stubborn deposits. Through the active blowing method that combines continuous low-pressure air curtain with periodic high-pressure pulse blowing, the optical window 501 is kept clean during long-term welding operations.
[0027] The optical window 501 of the infrared temperature probe 5 is equipped with a narrowband infrared filter in the 1500nm to 2200nm wavelength range. This filter can effectively shield against strong visible light arc interference generated during submerged arc welding, allowing only infrared thermal radiation signals of specific wavelengths to pass through, thereby ensuring the long-term stability and accuracy of infrared temperature measurement under high temperature, high dust, and strong arc conditions. The optical window 501 of the laser displacement sensor is also equipped with a laser light shield 507 to shield the laser ranging signal from interference by arc light.
[0028] The molten pool and deposited metal visual monitoring module is equipped with a high-speed industrial camera. The camera lens is also equipped with the aforementioned active lens cleaning protection structure to ensure the cleanliness of the optical window 501 in high-dust environments. The industrial camera is positioned appropriately behind the first head of the submerged arc welding machine to capture high-definition images of the molten pool and deposited metal area in real time from a suitable angle. During the welding process, the high-speed industrial camera continuously acquires visual data such as the width of the molten pool, the shape of the molten pool contour, the depth of the weld, and the flatness of the deposited metal. The acquired image data is transmitted in real time to the multi-parameter data fusion analysis module, where image processing algorithms automatically identify forming defects such as incomplete fusion, weld collapse, and metal accumulation. The molten pool morphology data, as an auxiliary judgment basis, is input into the fusion analysis module along with straightness deformation data, temperature data, and welding electrical parameter data to help the system comprehensively determine the degree of matching between the current welding heat input and the actual state of the pipeline.
[0029] The multi-parameter data fusion and analysis module is the core computing unit for achieving adaptive closed-loop control in this invention. This module incorporates a deformation-temperature coupling mathematical model to fuse four types of core monitoring data: the actual plastic straightness deformation and deformation growth rate output by the pipeline straightness dynamic detection module; the real-time weld temperature and temperature rise rate output by the weld infrared real-time temperature measurement module; the molten pool morphology and weld formation quality data output by the molten pool and deposited metal visual monitoring module; and the real-time welding current and voltage waveform data fed back from the welding power source.
[0030] The input variables of the deformation-temperature coupled mathematical model include real-time straightness deviation ΔL, deformation growth rate vL, peak molten pool temperature, interpass temperature, temperature rise rate vT, pipe wall thickness δ, pipe linear expansion coefficient α, and total pipe length L. The core prediction formula of the model is: ΔL_pred=ΔL+k1·vT·δ·α·L+k2·vL, In the formula, ΔL_pred is the predicted deformation, ΔL is the real-time straightness deviation, vT is the temperature rise rate, δ is the pipe wall thickness, α is the pipe linear expansion coefficient, L is the total length of the pipe, and k1 and k2 are correction coefficients calibrated by welding tests of the same material.
[0031] The model works by using welding heat input to induce a localized temperature rise in the pipeline, causing the pipe to expand and undergo elastic deformation. When the heat input accumulates and the temperature exceeds the pipe's yield temperature, the elastic deformation transforms into permanent plastic deformation. The model calculates the thermal expansion trend using the real-time temperature rise rate vT and the pipe's linear expansion coefficient α, converting this into axial deformation components based on the pipe wall thickness δ and total length L. This is then superimposed with the current real-time straightness deviation ΔL and deformation growth rate vL to predict the pipeline's bending deformation trend within a 5-20 second time window. The k1 and k2 coefficients are calibrated using standard welding tests on different materials to correct for discrepancies between theoretical calculations and actual operating conditions.
[0032] The multi-parameter data fusion analysis module, based on the predicted deformation output from the model, combined with current temperature data, molten pool morphology data, and welding electrical parameter data, comprehensively determines the abnormality level of the current welding state. Specifically, the system classifies four deformation levels and four temperature abnormality levels based on the predicted deformation amount and the degree to which the real-time temperature exceeds the threshold, providing a quantitative basis for the subsequent threshold judgment module and control execution module. Because the model can complete the prediction and output adjustment commands 1 to 2 welding cycles before the actual deformation occurs, the system can achieve preventive parameter intervention, rather than passive correction after deformation occurs.
[0033] The threshold judgment module has a pre-stored process database covering various materials such as carbon steel, chromium-molybdenum heat-resistant steel, and stainless steel. It also stores temperature safety thresholds and straightness deformation safety thresholds for different pipe diameters and wall thicknesses. Users can retrieve the corresponding threshold parameters and standard welding process parameters with a single click through the system interface based on the specific pipe material grade, diameter, and wall thickness of the pipe to be welded. The threshold judgment module matches a corresponding four-level linkage control strategy based on the comprehensive anomaly level output by the multi-parameter data fusion analysis module. Each level of control strategy has quantified adjustment ranges for welding current, arc voltage, and pipe rotation speed.
[0034] The first level is a slight abnormality control strategy: when the straightness deviation does not exceed 0.1 mm / m and the temperature exceeds the threshold by no more than 50℃, it is judged as a slight abnormality. The system controls the submerged arc welding current and voltage adjustment module to slightly reduce the welding current by 5% to 10%, and maintain the arc voltage and pipeline speed unchanged, thereby suppressing the continuous upward trend of heat input. The second level is a moderate anomaly control strategy: when the straightness deviation is between 0.1 mm / m and 0.3 mm / m, and the temperature exceeds the threshold between 50℃ and 120℃, it is judged as a moderate anomaly. The system controls to reduce the welding current by 10% to 25%, reduce the arc voltage by 3% to 8%, and simultaneously reduce the pipe rotation speed by 5% to 15% through the pipe rotation speed frequency conversion control module. This coupled control strategy reduces the welding heat input while extending the circumferential heat dissipation time of the pipe, balancing the arc heat accumulation on one side, and correcting the bending trend from a mechanistic perspective. The third level is the severe anomaly control strategy: when the straightness deviation exceeds 0.3mm / m or the temperature exceeds the threshold by more than 120℃, it is judged as a severe anomaly. The system controls the reduction of welding current by 25% to 40%, the reduction of arc voltage by 8% to 15%, and the reduction of pipeline speed by 10% to 20%. At the same time, the first-level audible and visual alarm is triggered to remind the operator to pay attention to the abnormal status.
[0035] The fourth level is the fusion-insufficiency control strategy: when the width or depth of the weld pool is lower than the process threshold, it is determined that the fusion is insufficient. The system controls a slight increase in welding current and arc voltage, with the increase controlled within 5%, and fine-tunes the pipeline speed to moderately increase the welding heat input to ensure the fusion quality of the weld.
[0036] Once the weld temperature returns to a safe range and the rate of increase in straightness deformation approaches zero, the system does not immediately restore the standard parameters. Instead, it gradually recovers to the standard process parameters over 3 to 5 adjustment cycles, with each parameter adjustment not exceeding 10%. This gradual recovery method avoids drastic fluctuations in weld formation caused by sudden parameter changes, ensuring a smooth transition in welding quality.
[0037] The submerged arc welding current and voltage regulation module connects to the control terminal of the welding power supply. It receives graded regulation commands from the multi-parameter data fusion analysis module via the threshold judgment module, and continuously and linearly adjusts the output current and arc voltage of the welding power supply. This module uses analog or digital control interfaces, enabling stepless smooth adjustment and avoiding the impact of step-like parameter changes on the stability of the welding process.
[0038] The frequency conversion control module for pipe rotation speed is connected to the frequency conversion speed control motor 2 that drives the roller frame 4 or chuck 3. After receiving the graded adjustment command, it adjusts the motor speed through the frequency converter, thereby changing the rotation speed of the pipe. The adjustment of the pipe speed is coordinated with the adjustment of welding current and voltage to achieve dynamic matching between welding heat input and pipe heat dissipation efficiency.
[0039] Under moderate and severe abnormal operating conditions, the system employs a coupled correction strategy that combines reducing welding heat input with reducing pipe rotation speed. Reducing welding current and voltage directly decreases the welding heat input per unit time, while reducing pipe rotation speed prolongs the residence time of the same pipe section in the non-heated area, increasing circumferential heat dissipation time. The two work synergistically to reduce heat input at the source and enhance heat dissipation, effectively balancing the circumferential temperature distribution of the pipe and suppressing axial plastic bending deformation caused by continuous heating on one side.
[0040] The early warning storage and traceability module features tiered audible and visual early warning output, fault interlock shutdown protection, and full-process data storage and traceability. For tiered early warning, the module determines the level of abnormality based on thresholds and outputs corresponding audible and visual warning signals: minor abnormalities trigger a suggestive warning, moderate abnormalities trigger a cautionary warning, and severe abnormalities trigger a level one audible and visual alarm, reminding on-site operators to monitor equipment status. Furthermore, the early warning storage and traceability module has built-in fault interlock logic. The system automatically triggers interlock shutdown protection when the following conditions occur: loss or abnormality of laser displacement sensor or infrared temperature sensor signals; severe deformation exceeding the limit for more than 10 seconds; severe molten pool collapse exceeding the threshold; and other critical conditions determined by the system. During interlock shutdown, the module automatically cuts off the welding power output and controls the pipeline to decelerate and stop via the pipeline rotation speed frequency conversion control module. Simultaneously, it locks all current sensor data, welding parameters, and molten pool image snapshots, and displays equipment maintenance suggestions in a pop-up window on the system interface.
[0041] The early warning storage and traceability module continuously stores the following data throughout the welding process: pipeline reference calibration data; straightness deformation time-series curves for each monitored section; peak temperature of the weld pool, temperature of the heat-affected zone, and interpass temperature time-series curves; visual monitoring images of the weld pool and deposited metal; real-time welding current and voltage waveform data; and records of various levels of abnormal alarms and control actions. This data can be exported locally as report files or uploaded to the factory's MES system cloud server via an industrial Ethernet interface. The system can automatically generate welding quality traceability reports that comply with the NB / T47014 welding procedure qualification and pressure vessel special equipment inspection digital traceability specifications.
[0042] This invention's system incorporates dedicated process databases for three main types of pipes: carbon steel, chromium-molybdenum heat-resistant steel, and stainless steel. Each material type is further subdivided into different grades, corresponding to and storing temperature safety thresholds, straightness deformation safety thresholds, and process parameters such as standard welding current, arc voltage, and rated pipe rotation speed under different pipe diameters and wall thicknesses. After selecting the pipe grade and inputting the pipe diameter and wall thickness through the system's interface, the system automatically retrieves the matching threshold parameters and standard process parameters to complete the initial setup, enabling one-click switching.
[0043] The mounting brackets for the laser displacement sensor array 6 and the infrared temperature probe 5 of this invention feature an adjustable structure, allowing for adjustments to the sensor detection distance and working angle according to different pipe diameters. This adapts to straight pipes with diameters ranging from φ50mm to φ1500mm and wall thicknesses from 8mm to 120mm, as well as circumferential submerged arc welding of pipe transition sections. For pipe transition sections, the system can also achieve effective detection and closed-loop control by adjusting the sensor placement and calibration parameters at each monitoring section. Example
[0044] This embodiment, in conjunction with a specific welding process, provides a detailed explanation of the implementation process of the adaptive welding control method for rotating pipe submerged arc welding based on dual monitoring of straightness and weld temperature.
[0045] S1. Pre-welding reference calibration and process initialization: After the pipe to be welded is positioned on the roller frame, the pre-welding reference calibration and process initialization operation is first performed. The pipe reference calibration module is activated, the roller frame is reset to zero, and the pipe rotates uniformly for one revolution without welding heat input. As mentioned earlier, the outer wall distance measurement data is collected every 1° of rotation. The three-dimensional absolute reference axis is fitted using the least squares method, and the reference parameters are locked. The operator selects the pipe material grade, inputs the pipe diameter and wall thickness through the system interface. The system automatically retrieves the corresponding data from the built-in process database based on the above information, presets four-level temperature thresholds and four-level straightness deformation thresholds, and initializes the standard welding current, arc voltage and pipe rated rotation speed. At the same time, the lens active purge protection module is activated, and the continuous low-pressure air curtain is opened to protect the optical windows of the infrared temperature probe and laser displacement sensor from dust.
[0046] S2. Welding Start: After calibration and initialization, start the welding program. The submerged arc welding machine remains stationary, and the roller frame or chuck drives the pipeline to rotate uniformly at the rated rotation speed. The submerged arc welding head ignites the arc at the beginning position of the circumferential weld, and the multi-layer, multi-pass submerged arc welding operation of the circumferential weld begins. The system enters the real-time monitoring and adaptive control operation state.
[0047] S3. Full-Process Synchronous Multi-Parameter Acquisition: Throughout the welding process, each detection module synchronously performs data acquisition tasks. At least three sets of laser displacement sensor arrays arranged along the pipe axis acquire radial offset data of the pipe wall at each cross-section in real time at a sampling interval of 1°. After processing by a composite filtering algorithm, the pipe axial straightness deformation data is output. The weld seam infrared real-time temperature measurement module synchronously acquires peak temperature of the molten pool, temperature of the heat-affected zone, and interpass temperature data. The high-speed industrial camera of the molten pool and deposited metal vision monitoring module continuously acquires images of the molten pool morphology and weld formation status. Real-time welding current and voltage waveform data fed back by the welding power supply are synchronously transmitted to the main control module. The above four types of data are synchronously marked with a unified timestamp to ensure the time consistency of subsequent fusion analysis. During the acquisition process, the composite algorithm of time-domain mean filtering and frequency-domain vibration filtering runs continuously to filter out detection noise caused by flux dust scattering, rotating mechanical vibration, and instantaneous arc thermal expansion, accurately distinguishing between elastic instantaneous offset and permanent plastic deformation, and avoiding the system from erroneously adjusting welding parameters due to noise interference.
[0048] S4. Multi-parameter Coupled Modeling and Predictive Analysis: The multi-parameter data fusion analysis module receives all the data collected in step S3 and inputs it into the deformation-temperature coupled mathematical model for calculation. Based on the real-time straightness deviation ΔL, deformation growth rate vL, temperature rise rate vT, pipe wall thickness δ, pipe linear expansion coefficient α, and total pipe length L, the model calculates the predicted deformation ΔL_pred according to ΔL_pred = ΔL + k1·vT·δ·α·L + k2·vL. The module comprehensively considers the predicted deformation, the degree to which the current temperature exceeds the threshold, and the molten pool morphology identification results to determine the overall anomaly level.
[0049] S5. Graded Closed-Loop Adaptive Control: Based on the comprehensive abnormality level determined in step S4, the threshold judgment module matches the corresponding control strategy and outputs graded adjustment instructions. The submerged arc welding current and voltage adjustment module and the pipeline rotation speed frequency conversion control module synchronously receive the instructions and execute the adjustments. Under slightly abnormal conditions, the system slightly reduces the welding current by 5% to 10% while maintaining the arc voltage and pipeline speed unchanged. Under moderately abnormal conditions, the system synchronously executes coupled control by reducing the current by 10% to 25%, reducing the voltage by 3% to 8%, and reducing the pipeline speed by 5% to 15%. Under severely abnormal conditions, the system executes large-scale control by reducing the current by 25% to 40%, reducing the voltage by 8% to 15%, and reducing the pipeline speed by 10% to 20%, and triggers a first-level audible and visual alarm. If the system identifies insufficient fusion formation as indicated by the weld pool width or weld depth being lower than the process threshold, it slightly increases the welding current and voltage by up to 5% each and fine-tunes the pipeline speed to moderately increase the welding heat input.
[0050] S6. Gradual Parameter Recovery: The system continuously monitors the weld temperature and straightness deformation rate. When the weld temperature drops back to a safe range and the straightness deformation growth rate approaches zero, the system does not immediately restore to the standard process parameters. Instead, it gradually recovers in 3 to 5 adjustment cycles, with the parameter adjustment range of each adjustment cycle not exceeding 10% of the standard value, to avoid sudden parameter changes that could cause drastic fluctuations in weld formation.
[0051] S7. Circular closed-loop monitoring and control: The system continuously repeats the operation process from S3 to S6, maintaining real-time monitoring and dynamic control in each layer and each welding process until all the multi-layer and multi-pass welding processes of the entire circumferential seam are completed.
[0052] S8. Data Storage and Digital Traceability: Throughout the entire welding process, the early warning, storage, and traceability module continuously stores all data, including benchmark calibration data, straightness deformation time-series curves, temperature time-series curves, molten pool images, welding parameter waveforms, and abnormal alarm records. After welding is completed, the system allows operators to export complete welding quality reports via a local interface, or upload the data to the MES system cloud server via the industrial network, automatically generating report documents that comply with the NB / T47014 Welding Procedure Qualification and Pressure Vessel Special Equipment Inspection Digital Traceability Specifications.
[0053] S9. Emergency Interlock Handling for Faults: At any moment during the welding process, if the system detects sensor signal loss, severe deformation exceeding the standard for more than 10 seconds, severe collapse of the molten pool exceeding the threshold, or other critical conditions determined by the system, the early warning storage and traceability module will immediately trigger the fault interlock protection, automatically cut off the welding power output, control the pipeline to decelerate and stop, and simultaneously lock all sensor data and welding parameter snapshots at the time of the fault occurrence. A pop-up window on the system operation interface will provide equipment maintenance suggestions to ensure the safety of the equipment and pipelines. Example
[0054] The following example, using the submerged arc welding of circumferential seams in thick-walled 12Cr1MoV heat-resistant steel pipes for power plant boilers, further illustrates the specific implementation of this invention.
[0055] The pipe to be welded has an outer diameter of φ325mm, a wall thickness of 40mm, a total length of 5m, and is made of 12Cr1MoV heat-resistant steel. After the pipe is hoisted onto the roller frame, pre-welding calibration is performed. The pipe rotates 360°, and outer wall contour data is collected every 1° of rotation. The three-dimensional reference axis is fitted using the least squares method. The calibration process automatically identifies and eliminates the initial 0.12mm / m curvature of the pipe itself and locks the reference parameters. The operator selects the 12Cr1MoV heat-resistant steel material and inputs the pipe diameter and wall thickness through the system interface. The system automatically retrieves the dedicated process library for heat-resistant steel, presets the temperature safety threshold to 350℃, the straightness deformation threshold to 0.1mm / m, initializes the standard welding current to 480A, the arc voltage to 36V, and the rated rotation speed of the pipe to 0.12r / min. The lens active purge protection module activates continuous low-pressure air curtain protection and performs high-pressure pulse air blowing and dust scraping ring wiping every 30 seconds.
[0056] After welding was initiated, the pipeline rotated at a constant speed, and the submerged arc welding head sequentially completed multiple layers and multiple passes of welding, including the root pass, fill pass, and cap pass. When welding reached the third fill pass, the real-time infrared temperature measurement module detected an interpass temperature of 420℃, exceeding the safe temperature threshold of 70℃. Simultaneously, the pipeline straightness dynamic detection module detected a straightness deviation of 0.22mm / m, falling between 0.1mm / m and 0.3mm / m. The multi-parameter data fusion analysis module input the above data into the deformation-temperature coupled mathematical model. After calculating and predicting the deformation, the model comprehensively determined that the current operating condition was moderately abnormal.
[0057] The threshold judgment module matches the moderate anomaly control strategy and outputs adjustment commands. The submerged arc welding current and voltage adjustment module reduces the welding current from 480A to 360A and the arc voltage from 36V to 33V; the pipeline rotation speed frequency conversion control module reduces the pipeline speed from 0.12r / min to 0.098r / min. After the system continuously monitors in a closed loop for about 80 seconds, the weld interpass temperature gradually drops to 340℃, and the pipeline deformation growth rate approaches zero. Subsequently, the system gradually restores the welding current, arc voltage, and pipeline speed to standard process parameters in four adjustment cycles, with the adjustment range in each cycle not exceeding 10% of the standard value.
[0058] After the multi-layer welding of the entire circumferential seam was completed, the measured straightness of the pipe after welding was 0.07 mm / m, which is better than the preset threshold requirement of 0.1 mm / m, and the pipe showed no excessive axial bending. The weld was subjected to non-destructive testing and mechanical property inspection, and there were no defects such as hot cracks, coarse grains, lack of fusion, or slag inclusions. No flame straightening or weld repair was required, and the first-pass yield rate met the high standard requirements. The straightness deformation curve, temperature curve, molten pool image, and welding parameter data of the entire welding process are completely stored in the early warning storage and traceability module, and can be directly exported as traceability reports that meet the inspection requirements. Example
[0059] like Figure 8 As shown, multiple sets of grating displacement sensors 12 are circumferentially installed near both ends of the pipe weld using clamps 11. Each set of grating displacement sensors 12 is at the same horizontal level and arranged opposite to each other. The grating displacement sensors 12 rotate synchronously with the pipe, detecting changes in the straightness of the pipe during the welding process. In this embodiment, the submerged arc welding machine 1, infrared temperature probe 5, and other structures are the same as in Embodiment 1, and will not be described again.
[0060] In this embodiment, the pipeline straightness dynamic detection module uses a grating displacement sensor 12 to replace or supplement the laser displacement sensor array 6 in embodiment one. For example... Figure 8 As shown, clamps 11 are installed near both ends of the pipe weld. Each clamp 11 is formed by two semi-circular clamps joined together by fastening bolts and tightly fixed to the outer wall of the pipe. The clamps 11 rotate synchronously with the pipe, thereby driving the grating displacement sensors 12 installed on them to rotate synchronously. Multiple sets of grating displacement sensors 12 are installed circumferentially on each clamp 11. Figure 8 The diagram shows two groups, upper and lower. The actual number can be flexibly set according to the pipe diameter, wall thickness, detection accuracy and deformation monitoring direction. The installation reference of each group of grating displacement sensors 12 is at the same level, and they are set opposite to each other along the pipe axis and / or circumference, so that the grating displacement sensors 12 set opposite to each other form a differential or relative displacement detection relationship.
[0061] The grating displacement sensor 12 can be an incremental grating ruler, an absolute grating ruler, a grating micrometer sensor, or a grating displacement sensor. Internally, it includes a scale grating and a reading head, utilizing moiré fringes or diffraction interference principles to convert relative displacement into electrical pulse signals or digital position codes. In a pair of grating displacement sensors 12 arranged opposite each other, one can serve as the reference side and the other as the measuring side; or both can respectively constitute the scale grating and the reading head of the grating pair. When the pipeline experiences axial bending, warping, or axis misalignment due to unilateral heat accumulation during welding, relative displacement occurs between the grating displacement sensors 12 at the same phase angle, the same level, or the circumferential relative position at both ends of the pipeline. The moiré fringes or diffraction interference signal of the grating pair changes accordingly, and the reading head outputs a corresponding displacement pulse or absolute position signal. The multi-parameter data fusion analysis module calculates the radial offset of each detected section of the pipeline based on the above signals and further outputs real-time straightness error, bending vector, and deformation growth rate.
[0062] In the pre-welding reference calibration stage, in addition to performing the cross laser reference calibration in Example 1, the grating displacement sensor 12 can also be initially calibrated. Specifically, with the pipeline driven to rotate one revolution under no welding heat input, each grating displacement sensor 12 collects initial relative displacement data at predetermined angular intervals, for example, once every 1° of rotation, obtaining initial displacement values at multiple points throughout the revolution. This initial displacement value is locked by the reference parameter storage and locking unit, serving as the zero-point reference for changes in pipeline straightness during subsequent welding. During welding, the grating displacement sensor 12 rotates synchronously with the pipeline and collects the relative displacement change in real time. The difference between the real-time displacement value and the initial displacement value is used as the radial offset change of the pipeline in that phase angle direction. By comparing the radial offset change at the same phase angle at both ends of the pipeline, the bending component in that direction can be calculated. By vector synthesis of the bending components at different phase angles, the bending vector direction of the pipeline's current axis and the real-time straightness error in mm / m can be obtained, and the deformation growth rate can be calculated by combining the timestamp.
[0063] The output signal of the grating displacement sensor 12 may include A / B phase quadrature pulses, zero-position pulses, or absolute position codes. After counting, direction identification, and phase calculation, the relative displacement is obtained. Its output data is connected to the pipeline straightness dynamic detection module and processed using the same time-domain mean filtering and frequency-domain vibration filtering composite algorithm as in Embodiment 1. This filters out interference such as periodic jumps of pipeline rotation, elliptical eccentricity, instantaneous thermal expansion elastic deformation, mechanical vibration, and flux dust scattering. It extracts slowly accumulating plastic permanent bending deformation data, avoiding misjudging instantaneous elastic displacement as plastic deformation and triggering unnecessary parameter adjustments.
[0064] The grating displacement sensor 12 can be set up simultaneously with the laser displacement sensor array 6 in Embodiment 1 to form redundant detection, or it can replace the laser displacement sensor array 6 independently and be uniformly fused and processed by the multi-parameter data fusion analysis module. The real-time straightness error, bending vector and deformation growth rate detected by the grating displacement sensor 12 are also input into the deformation-temperature coupled mathematical model to participate in the prediction of the subsequent bending deformation trend of the pipeline, the classification of deformation levels, and the graded linkage adjustment of welding current, arc voltage and pipeline rotation speed.
[0065] The above description is only a preferred embodiment of the present invention and does not limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.
Claims
1. An adaptive welding control system for rotary pipe submerged arc welding based on dual monitoring of straightness and weld temperature, characterized in that, include: The pipeline datum calibration module is used to construct the three-dimensional absolute datum axis of the pipeline before welding and output the initial datum coordinates of the pipeline. After calibration, the datum parameters are locked. The pipeline straightness dynamic detection module arranges at least three sets of laser displacement sensor arrays along the pipeline axis to collect the radial offset of the pipe wall in real time during the pipeline rotation welding process, and outputs real-time straightness error, bending vector and deformation growth rate. The real-time infrared temperature measurement module for weld seams is used to synchronously collect temperature data of the weld seam area. A visual monitoring module for the molten pool and deposited metal is used to collect molten pool morphology data in real time; The multi-parameter data fusion analysis module has a built-in deformation-temperature coupled mathematical model, which is used to fuse straightness error, temperature data, molten pool morphology data and welding electrical parameter data to predict the subsequent bending deformation trend of the pipeline and classify the deformation level and temperature anomaly level. The threshold judgment module has pre-stored temperature thresholds and straightness deformation thresholds corresponding to different pipe materials, pipe diameters, and wall thicknesses. The submerged arc welding current and voltage regulation module and the pipeline rotation speed frequency conversion control module are used to adjust the welding current, arc voltage and pipeline rotation speed in conjunction with the deformation level and temperature anomaly level. The early warning storage and traceability module is used to output early warning signals, execute fault interlock shutdown, and store full-process data.
2. The adaptive welding control system for rotating pipe submerged arc welding according to claim 1, characterized in that, The pipeline reference calibration module includes a cross laser reference emitting device, a laser receiver, an attitude locking bracket, a coordinate fitting calculation unit, and a reference parameter storage and locking unit. Before welding, the pipeline rotates one revolution and automatically collects the distance measurement data of the outer wall at each predetermined rotation angle. The pipeline standard reference axis is fitted by the least squares method, and the reference parameters are locked after calibration.
3. The adaptive welding control system for rotating pipe submerged arc welding according to claim 1, characterized in that, The pipeline straightness dynamic detection module incorporates a time-domain mean filtering and frequency-domain vibration composite filtering algorithm to filter out pipeline rotation periodic jumps, elliptical eccentricity, instantaneous thermal expansion elastic deformation, and mechanical vibration interference, and outputs plastic permanent bending deformation data.
4. The adaptive welding control system for rotating pipe submerged arc welding according to claim 1, characterized in that, The lens end of the weld seam infrared real-time temperature measurement module is equipped with an active lens cleaning and protection module. The active lens cleaning and protection module includes an annular air curtain nozzle, an air source pressure regulating valve, a pulse timing controller, a dust scraper ring, and a narrow band filter, which are used to continuously prevent dust, shield light, and resist spatter during the welding process.
5. The adaptive welding control system for rotating pipe submerged arc welding according to claim 1, characterized in that, The input variables of the deformation-temperature coupled mathematical model include real-time straightness deviation, deformation growth rate, peak temperature of the molten pool, interpass temperature, temperature rise rate, pipe wall thickness, pipe linear expansion coefficient, and total pipe length; the model pre-adjusts parameters by pre-setting the welding cycle based on the predicted deformation amount.
6. The adaptive welding control system for rotating pipe submerged arc welding according to claim 1, characterized in that, The threshold judgment module is divided into a four-level linkage control strategy, including slight abnormality, moderate abnormality, severe abnormality and insufficient fusion. Each level is set with welding current adjustment range, arc voltage adjustment range and pipeline rotation speed adjustment range. When the temperature and deformation recover to the safe range, they are gradually raised to the standard process parameters in multiple adjustment cycles, and the single parameter adjustment range does not exceed 10%.
7. The adaptive welding control system for rotating pipe submerged arc welding according to claim 1, characterized in that, The early warning storage and traceability module has built-in fault interlocking logic. When sensor signals are lost, deformation continues to exceed the predetermined time, or the molten pool collapses beyond the threshold, the welding power supply is automatically cut off and the pipeline is controlled to decelerate and stop.
8. The adaptive welding control system for rotating pipe submerged arc welding according to claim 1, characterized in that, The system has a built-in multi-material process database including carbon steel, chromium-molybdenum heat-resistant steel and stainless steel, and is suitable for circumferential submerged arc welding of straight pipes and bends with pipe diameters of φ50 to φ1500 mm and wall thicknesses of 8 to 120 mm.
9. An adaptive welding control method for submerged arc welding of rotating pipes based on dual monitoring of straightness and weld temperature, characterized in that, The control system according to any one of claims 1 to 8 is used to perform the following steps: S1. Pre-welding reference calibration and process initialization: Start the pipeline reference calibration module, rotate the pipeline one revolution to fit the three-dimensional absolute reference axis and lock the reference parameters; according to the pipe material, pipe diameter and wall thickness, retrieve the built-in process database, preset the temperature threshold and straightness deformation threshold, and initialize the standard welding current, arc voltage and pipeline rated rotation speed. S2. Welding Start: Drive the pipeline to rotate at a constant speed to carry out circumferential submerged arc welding; S3. Full-process synchronous multi-parameter acquisition: The laser sensor array acquires the axial straightness deformation data of the pipeline in real time, the infrared temperature measurement unit acquires the temperature data of the weld area in real time, and the visual monitoring unit acquires the molten pool morphology data and acquires the real-time welding current and voltage data in real time. S4. Multi-parameter coupled modeling and predictive analysis: Input the data collected in step S3 into the deformation-temperature coupled mathematical model, calculate the temperature rise rate and the plastic deformation growth rate, predict the subsequent bending deformation trend of the pipeline, and determine the deformation level and temperature anomaly level. S5. Graded closed-loop adaptive control: Based on the grade determined in step S4, the welding current, arc voltage and pipe rotation speed are adjusted in a linked manner. S6. Gradual parameter recovery: When the weld temperature drops to a safe range and the straightness deformation rate approaches zero, the parameters are gradually restored to the standard process parameters in multiple cycles. S7. Cyclic closed-loop monitoring and control: Repeat steps S3 to S6 until the welding process is completed; S8. Data storage and traceability: Stores data throughout the entire process and supports exporting and uploading.
10. The adaptive welding control method for rotating pipe submerged arc welding based on dual monitoring of straightness and weld temperature according to claim 9, characterized in that, In step S5, when there is moderate or severe deformation or temperature deviation, a coupled correction strategy of reducing welding heat input and reducing pipe rotation speed is adopted; in step S3, the acquisition stage adopts a composite algorithm of time-domain mean filtering and frequency-domain vibration filtering to filter out detection noise caused by flux dust scattering, rotating mechanical vibration and instantaneous arc thermal expansion, and distinguish between elastic instantaneous offset and permanent plastic deformation.
