Steel pipe multi-point synchronous welding control method and system

Abstract

The invention discloses a steel pipe multi-point synchronous welding control method and system, and relates to the technical field of welding control. The method comprises the steps that a main controller generates a synchronous motion instruction according to a preset welding path plan and sends the synchronous motion instruction to a plurality of plasma arc welding execution units; molten pool image information, temperature distribution information and arc voltage and current information of corresponding welding spots are obtained in real time, and a multi-source monitoring data network is constructed; a built-in welding process optimization model is called, synchronous welding strategy optimization is executed, and a process parameter correction instruction is sent; and on the basis of the technological parameter correction instruction, all the welding execution units are controlled to conduct technological parameter adjustment with cooperative constraint, and the relative motion relation and thermal process coordination between all the welding spots are maintained while the synchronous welding operation is executed. The technical problem that in the prior art, during multi-point welding, accurate synchronous control over all welding points is difficult to achieve, and consequently the welding quality is unstable is solved, and the technical effect of improving the multi-point welding stability and the welding seam forming quality is achieved.

Description

Technical Field

[0001] This invention relates to the field of welding control technology, specifically to a method and system for controlling multi-point synchronous welding of steel pipes. Background Technology

[0002] Steel pipe structures are widely used in pressure vessels, power transmission and distribution networks, engineering machinery, and large steel structure manufacturing. Their welding quality directly affects the load-bearing capacity and service reliability of the structure. For steel pipe assemblies with multiple welding points operating simultaneously, existing welding processes typically employ multiple welding torches to perform welding operations separately. However, each welding unit relies on independent control, lacking unified coordination of key process parameters such as welding path, welding current, gas flow rate, and welding speed. During simultaneous multi-point welding, uneven heat input, significant differences in weld formation, and even localized thermal distortion accumulation can easily occur between welding points. Furthermore, traditional systems often rely on single sensors for welding status monitoring, failing to capture multi-source information such as molten pool images, temperature fields, and arc electrical parameters in real time. This makes it difficult for the system to perform coordinated control and adaptive parameter adjustment during welding, compromising welding consistency and stability. Summary of the Invention

[0003] This application provides a method and system for controlling multi-point synchronous welding of steel pipes, which solves the technical problem in the prior art that it is difficult to achieve precise synchronous control of each welding point during multi-point welding, resulting in unstable welding quality.

[0004] The first aspect of this application provides a method for controlling multi-point synchronous welding of steel pipes, the method comprising: The main controller generates synchronous motion commands based on a preset welding path plan and sends them to multiple plasma arc welding execution units. Through the visual sensors, temperature sensors, and arc electrical parameter acquisition modules associated with each welding execution unit, it acquires real-time image information of the molten pool, temperature distribution information, and arc voltage and current information of the corresponding weld point, constructing a multi-source monitoring data network. Based on this multi-source monitoring data network, the main controller receives and fuses the multi-source monitoring data, calls the built-in welding process optimization model, executes synchronous welding strategy optimization, and sends process parameter correction commands. These process parameters include at least welding current, plasma gas flow rate, and welding speed. Based on these process parameter correction commands, the main controller controls each welding execution unit to adjust process parameters with collaborative constraints, maintaining the relative motion relationship and thermal process coordination between weld points while performing synchronous welding operations.

[0005] A second aspect of this application provides a multi-point synchronous welding control system for steel pipes, the system comprising: Path planning component: The main controller generates synchronous motion commands based on the preset welding path plan and sends them to multiple plasma arc welding execution units; Data acquisition component: Through the vision sensor, temperature sensor, and arc electrical parameter acquisition module associated with each welding execution unit, the molten pool image information, temperature distribution information, and arc voltage and current information of the corresponding weld point are acquired in real time to construct a multi-source monitoring data network; Strategy optimization component: Based on the multi-source monitoring data network, the main controller receives and integrates the multi-source monitoring data, calls the built-in welding process optimization model, executes synchronous welding strategy optimization, and sends process parameter correction commands, wherein the process parameters include at least welding current, plasma gas flow rate, and welding speed; Welding control component: Based on the process parameter correction commands, the main controller controls each welding execution unit to adjust the process parameters with collaborative constraints, maintaining the relative motion relationship and thermal process coordination between each weld point while performing synchronous welding operations.

[0006] One or more technical solutions provided in this application have at least the following technical effects or advantages: First, the main controller generates synchronous motion commands based on a preset welding path plan and sends them to multiple plasma arc welding execution units. Next, through the vision sensors, temperature sensors, and arc electrical parameter acquisition modules associated with each welding execution unit, real-time image information of the molten pool, temperature distribution information, and arc voltage and current information of the corresponding weld point are acquired, constructing a multi-source monitoring data network. Then, based on the multi-source monitoring data network, the main controller receives and fuses the multi-source monitoring data, calls the built-in welding process optimization model, executes synchronous welding strategy optimization, and sends process parameter correction commands. These process parameters include at least welding current, plasma gas flow rate, and welding speed. Finally, based on the process parameter correction commands, the main controller controls each welding execution unit to adjust process parameters with collaborative constraints, maintaining the relative motion relationship and thermal process coordination between weld points while performing synchronous welding operations. This solves the technical problem in existing technologies where precise synchronous control of weld points during multi-point welding is difficult, leading to unstable welding quality, and achieves the technical effect of improving the stability of multi-point welding and the quality of weld formation. Attached Figure Description

[0007] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0008] Figure 1 A schematic flowchart of a multi-point synchronous welding control method for steel pipes provided in this application embodiment; Figure 2This is a schematic diagram of a multi-point synchronous welding control system for steel pipes provided in an embodiment of this application.

[0009] Figure labeling: Path planning component 11, data acquisition component 12, strategy optimization component 13, welding control component 14. Detailed Implementation

[0010] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below.

[0011] Example 1, as Figure 1 As shown, this application provides a method for controlling multi-point synchronous welding of steel pipes, wherein the method includes: The main controller generates synchronous motion commands based on the preset welding path plan and sends them to multiple plasma arc welding execution units.

[0012] In this embodiment, after acquiring the preset welding path output by the path planning module, the main controller discretizes the welding path, dividing it into several continuous path point sets according to the spatial curve characteristics of the weld trajectory. Each path point is associated with the desired welding torch posture, desired welding speed, and plasma arc welding start / stop timing information. The main controller further performs timing calculations on the path point sets based on the kinematic models of each welding execution unit, mechanism constraint parameters, and the synchronous motion relationships between multiple welding points. This generates a synchronous motion instruction set containing position, posture, and speed commands. The synchronous motion instruction set is timestamped using a unified time base to ensure that all welding execution units execute the corresponding motion trajectory within the same time window. Subsequently, the main controller sends the generated synchronous motion commands to multiple plasma arc welding execution units via a fieldbus or industrial Ethernet communication interface, enabling each execution unit to perform coordinated motion according to the synchronous commands.

[0013] Furthermore, before generating synchronous motion commands based on the preset welding path plan, the following steps are included: The weld area of ​​the steel pipe to be welded is scanned by a 3D vision scanning device to obtain point cloud data including the actual weld bevel shape, misalignment amount, and workpiece spatial pose. The point cloud data is matched and compared with the steel pipe design model to identify and extract the 3D trajectory of the centerline of the actual weld. Based on the extracted actual weld trajectory, the preset welding process database, and the multi-welding gun cooperative motion constraint parameters, a search is performed to obtain a welding path that meets the current actual working conditions, and the preset welding path plan is generated.

[0014] First, a 3D vision scanning device (such as a structured light scanner or laser line scanner) is used to perform a full-circumference scan of the weld area of ​​the steel pipe to be welded after it is mounted and positioned, acquiring high-density point cloud data covering the actual weld bevel morphology, misalignment amount, and workpiece spatial pose. After scanning, the main controller calls the point cloud preprocessing module to perform filtering, noise reduction, coordinate system alignment, and region segmentation operations on the original point cloud, obtaining an effective point cloud set representing the true geometric features of the weld area. Subsequently, the effective point cloud set is matched with the steel pipe design model in 3D, comparing the difference between the designed weld centerline position and the actual weld spatial morphology, and the 3D trajectory of the actual weld centerline is identified and extracted using a feature extraction algorithm. After obtaining the actual weld trajectory, the main controller performs multi-objective path search and feasibility assessment based on the weld trajectory, the standard process parameter range in the pre-set welding process database, and the cooperative motion constraint parameters configured for multiple welding execution units. During the search process, the main controller evaluates candidate paths one by one using constraints such as weld geometry, welding posture limitations, minimum safe distance between welding torches, and mechanism motion space. It then selects the optimal welding path that meets the current working conditions and generates a corresponding preset welding path plan. This preset welding path plan includes a sequence of welding trajectory points, welding torch posture parameters, welding speed reference, and coordinated start-stop timing information, providing a basis for the generation of subsequent synchronous motion commands.

[0015] Furthermore, based on the extracted actual weld trajectory, the preset welding process database, and the multi-welding torch cooperative motion constraint parameters, a search is performed to obtain a welding path that meets the current actual working conditions, including: Based on the geometric characteristics of the actual weld trajectory and the heat input requirements of the welding process, the trajectory is divided into straight line segments, circular arc segments, and characteristic transition segments. A set of welding process parameter constraints is parameterized and associated for each trajectory segment, including: the allowable welding speed range and the welding torch attitude angle range. Based on the welding process parameter constraints of each segment, and combined with the kinematic model and spatial layout constraint parameters of multiple welding execution units, collision-free cooperative motion simulation optimization is performed to determine the cooperative start-stop sequence and spatial path allocation of each welding execution unit, thereby obtaining the welding path.

[0016] The main controller first analyzes the spatial geometric characteristics of the actual weld trajectory. Based on the curvature of the weld centerline, the changes in bevel morphology, and the structural transition characteristics of adjacent areas, the entire weld trajectory is automatically segmented into straight segments, circular arc segments, and characteristic transition segments. Subsequently, for each trajectory segment, the main controller retrieves the corresponding process data template from the preset welding process database and parametrically associates a set of welding process parameter constraints with the trajectory segment. These welding process parameter constraints include at least the allowable welding speed range, welding torch attitude angle range, heat input limit value, and welding direction switching conditions for that trajectory segment, thereby ensuring that each path segment meets the requirements for thermal process control and forming stability.

[0017] After completing the process parameter constraint association, the main controller, based on the welding parameter constraints of each segment and combined with the kinematic models, attitude rotation range, reachable space of the mechanism, and spatial layout constraints such as the minimum safe distance between welding torches of multiple welding execution units, constructs a cooperative motion solution model to calculate the cooperative motion parameters that meet the working conditions. The main controller uses the cooperative motion solution model to perform collision-free cooperative motion simulation on multiple welding execution units, iteratively optimizing the working area division, welding cycle arrangement, and start / stop timing allocation of each welding execution unit in the corresponding trajectory segment through multiple iterations. By comprehensively evaluating indicators such as collision risk, mechanism limits, synchronization deviation, and process compliance, the main controller determines the cooperative start / stop timing and spatial path allocation scheme that meets the current working conditions, and generates a welding path that meets the actual welding conditions. This welding path is ultimately used as a preset welding path plan for the generation of subsequent synchronous motion commands.

[0018] The construction of the cooperative motion solution model includes: establishing corresponding kinematic models based on the joint type, link size, and installation posture of each welding execution unit, and establishing a one-to-one mapping relationship between the joint space of each welding execution unit and the working space of the welding torch end near the weld seam of the steel pipe; on this basis, using the end posture, linear velocity and angular velocity, and start-stop time of each welding execution unit on each trajectory segment as decision variables, introducing multi-welding torch cooperative motion constraint parameters, and constructing a constraint set including trajectory tracking constraints, upper and lower limits of joint angle and joint velocity constraints, boundary constraints of the working space of each welding execution unit, and minimum safe distance constraints between welding torches; at the same time, multi-welding point synchronization deviation, overall cycle time, welding torch motion smoothness, and working space utilization rate are used as optimization objectives or comprehensive cost functions to form a cooperative motion solution model.

[0019] The collision-free cooperative motion simulation optimization includes: constructing a virtual simulation environment corresponding to the actual welding station in the main controller; importing the 3D model of the steel pipe workpiece, the tooling fixture model, and the mechanism models of multiple welding execution units into the simulation environment; and performing spatial positioning according to the actual installation posture; based on the candidate motion parameters of each welding execution unit obtained from the cooperative motion solution model, discretizing the welding path in time; and calculating the spatial position and attitude of the end of each welding execution unit and the mechanism linkage in each time step according to the preset time step. In each time step, the main controller calls the collision detection module to calculate the minimum distance between welding execution units, and between welding execution units and the steel pipe workpiece and tooling fixture. When any distance is detected to be less than the safety threshold or geometric overlap occurs, the current motion parameters are marked as a collision state, and the corresponding start / stop sequence, path allocation, or attitude trajectory is adjusted. Through multiple rounds of iterative simulation and parameter adjustment, under the premise of satisfying joint motion constraints and welding process parameter constraints, the number of collisions and minimum spacing violations are gradually reduced until no collision events occur within the simulation cycle, and the multi-weld point synchronization deviation, motion smoothness and cycle efficiency indicators all meet the preset requirements, and finally the collision-free cooperative motion optimization results for actual execution are obtained.

[0020] By using the vision sensor, temperature sensor, and arc electrical parameter acquisition module associated with each welding execution unit, the system can acquire real-time image information of the molten pool, temperature distribution information, and arc voltage and current information of the corresponding weld point, thus constructing a multi-source monitoring data network.

[0021] In this embodiment, each plasma arc welding execution unit is equipped with a high-speed vision sensor, an infrared temperature sensor, and an arc electrical parameter acquisition module at the end of its welding torch. The high-speed vision sensor captures image sequences of the weld pool area at a preset sampling frequency to reflect the molten pool boundary morphology, droplet transition behavior, and back-dragging angle changes. The infrared temperature sensor scans and measures the real-time temperature distribution around the weld point to obtain temperature field characteristics such as the peak temperature of the weld metal, the area of ​​the high-temperature region, and the high-temperature dwell time. The arc electrical parameter acquisition module synchronously acquires arc voltage and current waveforms through a high-speed sampling circuit to characterize the arc energy input and arc column stability. The data acquired by the above three types of sensors are transmitted back to the main controller via fieldbus or industrial Ethernet using a unified time reference. The main controller performs timestamp synchronization and format standardization processing on the monitoring data from different sources, and forms a monitoring data package containing multi-dimensional monitoring information through data structured encapsulation. All monitoring data packages are organized according to weld point number and time sequence to form a multi-source monitoring data network, realizing real-time, multi-dimensional monitoring of the welding status of each welding execution unit.

[0022] Based on the multi-source monitoring data network, the main controller receives and integrates the multi-source monitoring data, calls the built-in welding process optimization model, executes synchronous welding strategy optimization, and sends process parameter correction instructions, wherein the process parameters include at least welding current, plasma gas flow rate, and welding speed.

[0023] In this embodiment, the main controller receives batches of molten pool image information, temperature distribution information, and arc electrical parameters from multiple welding execution units at a preset synchronization clock cycle. It aligns data from different sources using a unified timestamp and fuses them into a multi-dimensional state vector containing molten pool morphology features, temperature field features, and arc energy features. The main controller inputs the fused state vector into a built-in welding process optimization model. Based on the current welding state deviation of each weld point, the heat input balance index, and the synchronization control objective, the welding process optimization model calculates corresponding action vectors. These action vectors include suggestions for adjusting process parameters such as welding current, plasma gas flow rate, and welding speed. The main controller generates process parameter correction instructions based on the action vectors and sends these instructions to each welding execution unit according to the weld point number. This enables real-time process adjustment during synchronous welding operations, thereby improving the synchronization of multi-weld-point welding and the stability of weld quality.

[0024] Furthermore, the main controller receives and fuses multi-source monitoring data, invokes the built-in welding process optimization model, executes synchronous welding strategy optimization, and sends process parameter correction instructions, including: The main controller, according to a preset synchronization clock, timestamps and encapsulates the molten pool image information, temperature distribution information, and arc voltage and current information to construct a time-series data frame with a unified structure. It normalizes each data item in the time-series data frame and extracts its features. The extracted features are then concatenated into vectors to construct a high-order feature tensor describing the instantaneous welding state of each weld point. This high-order feature tensor is used as input to the welding process optimization model. Based on the current welding state, and combined with the output action vector and reward feedback of the internal policy network, the model performs optimization to determine the process parameter strategy that satisfies the synchronous welding objective, and outputs the working parameter optimization strategy. Finally, it sends a process parameter correction command based on the working parameter optimization strategy.

[0025] The main controller performs timestamp alignment processing on the molten pool image information, temperature distribution information, and arc voltage and current information from each welding execution unit according to a preset synchronization clock cycle. It then formats and encapsulates the aligned multi-source data to construct a time-series data frame with a fixed field structure and uniform data length, ensuring consistent input of multi-source monitoring data in subsequent optimization models. Subsequently, the main controller normalizes the data in the time-series data frame, including normalizing the brightness values ​​of the molten pool image feature matrix, scaling the temperature distribution sequence, and standardizing the amplitude of the arc electrical parameter curve, to eliminate the influence of differences in data dimensions and dimensions. After normalization, the main controller calls a preset feature extraction algorithm to extract multi-dimensional features such as molten pool boundary width, trailing angle, peak temperature, high-temperature zone area, and arc stability indicators. These features are then concatenated in a fixed order to generate a high-order feature tensor describing the instantaneous welding state of each weld point.

[0026] The main controller inputs a high-order feature tensor into a welding process optimization model, which can be constructed based on a deep neural network structure and includes a policy network and a value evaluation unit. The policy network outputs an action vector representing the direction of process adjustment based on the input welding state features, and the value evaluation unit evaluates the action vector according to the model's internal reward feedback mechanism. Under the combined effect of the action vector and reward feedback, the main controller determines the process parameter strategy that satisfies the goal of simultaneous welding of multiple weld points through optimization and outputs an optimized working parameter strategy for real-time process adjustment.

[0027] The main controller generates process parameter correction instructions based on the working parameter optimization strategy. The process parameter correction instructions include at least updated set values ​​for welding current, plasma gas flow rate and welding speed, and are sent to the corresponding welding execution units to realize dynamic process adjustment based on real-time monitoring status.

[0028] The welding process optimization model is used to evaluate the welding status of each weld point in real time based on the fused multi-source monitoring data during multi-point synchronous welding, and output process parameter adjustment strategies suitable for the current welding conditions.

[0029] The welding process optimization model can be constructed based on a deep neural network structure. Specifically, this deep neural network consists of an input feature layer, a feature encoding layer, a policy inference layer, and a process parameter output layer. The input feature layer receives high-order feature tensors representing the instantaneous welding state of the weld point. The feature encoding layer uses convolutional neural networks, long short-term memory networks, or self-attention encoding modules to perform cross-dimensional fusion of molten pool image features, temperature field features, and arc electrical parameter features to extract deep feature representations reflecting the evolution trend of the welding state. Based on the extracted deep features, the policy inference layer generates action vectors for adjusting process parameters through a multi-layer fully connected network or reinforcement learning policy network. The process parameter output layer maps and solves the welding current, plasma gas flow rate, and welding speed according to the action vectors to generate a process parameter strategy that meets the current welding synchronization and thermal balance requirements. Simultaneously, a value evaluation network can be configured within the model to construct a reward function based on indicators such as weld stability, heat input balance, and molten pool morphology consistency, adaptively optimizing the policy inference process. This allows the model to continuously improve the accuracy and stability of process adjustment decisions under continuous online monitoring and feedback.

[0030] Furthermore, various data features are extracted, including: extracting the molten pool width and trailing angle features from the molten pool image; extracting the peak temperature and residence time in the high-temperature zone features from the temperature distribution; and extracting the energy concentration features from the arc electrical parameters.

[0031] The main controller performs image feature extraction processing on the timestamped and normalized molten pool image information. Specifically, it identifies the boundary contour of the molten pool region using an edge detection algorithm or a deep learning image segmentation network, calculates the molten pool width based on the pixel distance between the left and right endpoints of the boundary and camera calibration parameters, and determines the molten pool trailing angle feature based on the difference in the front and rear edge positions of the molten pool morphology to characterize the flow and extension behavior of the molten metal region, thereby reflecting the welding stability.

[0032] For temperature distribution information, the peak temperature point is identified from the two-dimensional or one-dimensional temperature distribution curve output by the temperature sensor to obtain the peak temperature characteristics of the weld point; and the residence time characteristics of the high temperature zone are calculated by detecting the length of time the temperature exceeds the preset high temperature threshold in the temperature curve, so as to evaluate the degree of influence of heat input on the surrounding materials.

[0033] For the electric arc electrical parameter data, the main controller performs waveform root mean square calculation, frequency domain analysis, or energy density estimation on the instantaneous waveforms of the collected arc voltage and current to obtain the energy concentration characteristics that characterize the stability of the arc energy distribution; the higher the energy concentration, the more stable the current physical morphology of the arc and the more controllable the melting depth.

[0034] Furthermore, sending process parameter correction instructions based on the aforementioned operating parameter optimization strategy includes: A hierarchical control structure is adopted to send process parameter correction instructions. Based on the overall state of the weld joint, the basic parameter settings for maintaining the collaborative benchmark are calculated, and collaborative instructions for each execution unit are generated and sent to each welding execution unit. According to the state deviation of the weld joint, the parameter compensation value for eliminating local deviations is calculated, and local compensation instructions are generated and sent to the corresponding welding execution unit. Each welding execution unit superimposes the received collaborative instructions and local compensation instructions as the final process parameter settings to be executed.

[0035] The main controller employs a hierarchical control structure to generate and issue process parameter correction commands. First, based on the overall state characteristics of the current multi-weld-point welding process, including heat input uniformity, weld-point synchronization indices, and global molten pool stability indices, the main controller calculates basic parameter settings to maintain the collaborative baseline of multiple welding execution units. These settings include uniform welding current reference values, plasma gas flow rate reference values, and synchronous welding speed reference values. The main controller generates collaborative commands based on these basic parameter settings and issues them to all welding execution units to ensure the synchronization and consistency of the overall welding process. Subsequently, based on the analysis results of local state deviations at each weld point in the working parameter optimization strategy, such as molten pool width deviations, local energy input deviations, or abnormal temperature field distributions, the main controller calculates parameter compensation values ​​to eliminate local deviations. These compensation values ​​are used to locally fine-tune the basic parameter settings, thereby forming local compensation commands acting on individual welding execution units. These commands are then issued by the main controller to the corresponding welding execution units to improve the local welding state and achieve dynamic stability control. After receiving the coordination command and local compensation command from the main controller, each welding execution unit performs command superposition processing on the two types of commands. That is, the coordination command is used as the baseline setting and the compensation command is used as the incremental adjustment. The process parameters such as welding current, plasma gas flow rate and welding speed are adjusted in real time to form the process parameter setting values ​​used to drive the actual welding process, thereby ensuring that the multi-weld-point welding process can achieve local adaptive control under global coordination.

[0036] Based on the process parameter correction instructions, each welding execution unit is controlled to adjust the process parameters with collaborative constraints, so as to maintain the relative motion relationship and thermal process coordination between each weld point while performing synchronous welding operations.

[0037] After generating the process parameter correction command, the command, which includes updated settings such as welding current, plasma gas flow rate, and welding speed, is distributed to each welding execution unit. Upon receiving the command, each welding execution unit's internal servo control module synchronously adjusts the welding torch posture controller, wire (powder) feeder, and plasma gas flow regulator based on the final set values ​​composed of the collaborative reference parameters and local compensation parameters. This ensures that each execution unit completes the coordinated update of process parameters under the same time reference.

[0038] While adjusting process parameters, the main controller, based on preset welding path planning and synchronous motion commands, performs position and speed linkage control on each welding execution unit, ensuring they perform synchronous welding operations along a predetermined trajectory, thus maintaining the relative motion relationship between multiple weld points. Through dynamic correction and synchronous control of welding current, gas flow rate, and welding speed, the instantaneous heat input and molten pool evolution behavior of each weld point are constrained within a set range, keeping the thermal processes of multiple weld points coordinated and consistent, avoiding uneven penetration, weld formation differences, or accumulation of local thermal distortion caused by local parameter drift. Ultimately, through process parameter adjustment and synchronous motion control under collaborative constraints, each welding execution unit achieves process consistency, molten pool stability, and heat input balance during multi-point welding, thereby ensuring consistent and stable overall weld formation quality and structural performance.

[0039] Furthermore, it also includes: Based on the process parameter correction command, the updated multi-source detection data after parameter adjustment is tracked and executed; the heat input balance index among all weld points is calculated based on the updated multi-source detection data; based on the heat input balance index, with minimizing the balance index as the optimization objective, and combined with the molten pool image features of each weld point, the process parameter adjustment amount for each welding execution unit is calculated; the tracking and adjustment steps are repeated until welding is completed, wherein the heat input balance index is stabilized within a preset threshold, achieving adaptive balance between welding heat input and molten pool behavior, so as to ensure consistent welding quality of steel pipe joint welds in each welding execution unit.

[0040] After the main controller issues process parameter correction commands and each welding execution unit completes the corresponding process parameter adjustments, the main controller continues to collect updated molten pool image information, temperature distribution information, and arc electrical parameters in real time based on the multi-source monitoring data network, forming new multi-source detection data after the process adjustment. The main controller processes the updated multi-source detection data, extracting features such as molten pool width, high-temperature zone dwell time, and arc energy concentration, and calculates the instantaneous heat input value for each weld point accordingly. Subsequently, the main controller calculates the heat input balance index among all weld points based on the instantaneous heat input values ​​of each weld point. The heat input balance index can be the variance, range, or normalized deviation of the instantaneous heat input values, used to quantify the consistency of heat input among different weld points. The smaller the heat input balance index, the more coordinated and stable the thermal process among multiple weld points.

[0041] After acquiring the heat input balance index, the main controller optimizes by minimizing this index. It then analyzes this index in conjunction with the molten pool image features of each weld point (including molten pool width, penetration tendency, and trailing angle changes). Using a built-in process parameter optimization model, it calculates the process parameter adjustments for each welding execution unit. These adjustments include at least increases / decreases in welding current, fine-tuning of plasma gas flow rate, and corrections to welding speed. Based on these adjustments, the main controller generates new process parameter correction commands and reissues them to each welding execution unit according to the aforementioned collaborative constraint mechanism. Through continuous iterative execution of the above tracking and adjustment steps, the main controller gradually converges the heat input deviation of multiple weld points to a preset threshold range, ensuring the heat input balance index remains stable and does not exceed the set threshold. This achieves adaptive balance control of heat input and molten pool behavior at each weld point during the welding process. Ultimately, different welding execution units maintain consistent molten pool stability and weld formation characteristics throughout the entire welding cycle, ensuring consistent overall welding quality of multi-point weld joints on the steel pipe.

[0042] Furthermore, based on updated multi-source detection data, the thermal input uniformity index among all solder joints is calculated, including: Based on updated multi-source detection data, the instantaneous heat input value of each solder joint is calculated in real time; based on the instantaneous heat input value of each solder joint, the variance or range of the instantaneous heat input values ​​of all solder joints is calculated as an indicator to measure the heat input balance.

[0043] After receiving updated multi-source detection data uploaded by each welding execution unit, the main controller calculates the instantaneous heat input value of each weld point in real time based on the molten pool image features, temperature distribution information, and arc electrical parameters. Specifically, the main controller uses data such as welding current, arc voltage, and welding speed, combined with the energy input formula, to obtain the instantaneous heat input value corresponding to the current moment.

[0044] After obtaining the instantaneous heat input values ​​of all solder joints, the main controller summarizes these instantaneous heat input values ​​according to the solder joint number and calculates the heat input balance index among the solder joints based on the set balance evaluation rules. The balance index can reflect the degree of heat input fluctuation between different solder joints by calculating the variance of the instantaneous heat input values ​​of each solder joint; alternatively, it can be used to measure the degree of heat input dispersion by calculating the range of the instantaneous heat input values, i.e., the difference between the maximum and minimum heat input. A smaller balance index indicates that the heat input of multiple solder joints is more consistent; conversely, a larger index indicates that further process parameter compensation and adjustment are needed.

[0045] Furthermore, based on the process parameter correction command, controlling each welding execution unit to adjust process parameters with cooperative constraints includes: Welding execution units are positioned based on process parameter correction instructions. When the process parameter correction instructions are for correcting the working parameters of a local weld point, the coupling interference amount on the molten pool stability, arc morphology, and local heat input of adjacent weld points is predicted based on the thermo-mechanical coupling relationship between the molten pools of the steel pipe weld points. According to the coupling interference amount, the main controller generates a feedforward compensation instruction to cancel the interference. The feedforward compensation instruction is used to cancel the interference amount and maintain the original coordinated thermal process and synchronous motion relationship. Based on the feedforward compensation instruction and the process parameter correction instructions, each welding execution unit is controlled to adjust the corresponding process parameters.

[0046] The main controller first parses the received process parameter correction command and determines the location range of the welding execution units involved in the process parameter adjustment based on the weld point number, correction command type, and adjustment amount. When the process parameter correction command is for the working parameters of a local weld point, the main controller performs local operating condition analysis for that weld point, calls the built-in thermo-mechanical coupling prediction model, and predicts the coupling interference that this local adjustment may cause to the molten pool stability, arc morphology, and local heat input of adjacent weld points based on the thermal conductivity characteristics of the steel pipe structure, the geometric relationship of the welding path, the mutual influence law between molten pools, and the current heat input state of each weld point. The coupling interference may include indicators such as temperature field disturbance of adjacent weld points, molten pool morphology change trend, and arc energy distribution shift.

[0047] After obtaining the coupling interference, the main controller generates feedforward compensation instructions to counteract the coupling interference based on the cooperative motion target and thermal process coordination requirements of adjacent weld points through a compensation algorithm. These feedforward compensation instructions include process compensation amounts for adjacent welding execution units, such as current fine-tuning, gas flow fine-tuning, or welding speed increments. These are used to counteract potential instability caused by local parameter corrections, thereby ensuring that the molten pool behavior, arc morphology, and heat input of adjacent weld points remain within normal ranges, and maintaining the established cooperative thermal process and synchronous motion relationship throughout the overall welding process.

[0048] Finally, the main controller fuses the feedforward compensation command and the process parameter correction command, and sends them to the corresponding welding execution units. Each welding execution unit adjusts the corresponding process parameters according to the received command. By performing process compensation and synchronous adjustment, it achieves adaptive correction of local deviations and dynamic maintenance of overall coordination, thereby ensuring process stability and weld formation consistency during multi-point synchronous welding.

[0049] In summary, the embodiments of this application have at least the following technical effects: First, the main controller generates synchronous motion commands based on a preset welding path plan and sends them to multiple plasma arc welding execution units. Next, through the vision sensors, temperature sensors, and arc electrical parameter acquisition modules associated with each welding execution unit, real-time image information of the molten pool, temperature distribution information, and arc voltage and current information of the corresponding weld point are acquired, constructing a multi-source monitoring data network. Then, based on the multi-source monitoring data network, the main controller receives and fuses the multi-source monitoring data, calls the built-in welding process optimization model, executes synchronous welding strategy optimization, and sends process parameter correction commands. These process parameters include at least welding current, plasma gas flow rate, and welding speed. Finally, based on the process parameter correction commands, the main controller controls each welding execution unit to adjust process parameters with collaborative constraints, maintaining the relative motion relationship and thermal process coordination between weld points while performing synchronous welding operations. This solves the technical problem in existing technologies where precise synchronous control of weld points during multi-point welding is difficult, leading to unstable welding quality, and achieves the technical effect of improving the stability of multi-point welding and the quality of weld formation.

[0050] Example 2, based on the same inventive concept as the multi-point synchronous welding control method for steel pipes in the previous examples, such as... Figure 2 As shown, this application provides a multi-point synchronous welding control system for steel pipes, wherein the system includes: Path planning component 11: The main controller generates synchronous motion commands based on the preset welding path planning and sends them to multiple plasma arc welding execution units; Data acquisition component 12: Through the visual sensor, temperature sensor, and arc electrical parameter acquisition module associated with each welding execution unit, it acquires the molten pool image information, temperature distribution information, arc voltage and current information of the corresponding weld point in real time, and constructs a multi-source monitoring data network; Strategy optimization component 13: Based on the multi-source monitoring data network, the main controller receives and integrates the multi-source monitoring data, calls the built-in welding process optimization model, executes synchronous welding strategy optimization, and sends process parameter correction commands, wherein the process parameters include at least welding current, plasma gas flow rate, and welding speed; Welding control component 14: Based on the process parameter correction commands, it controls each welding execution unit to adjust the process parameters with collaborative constraints, maintaining the relative motion relationship and thermal process coordination between each weld point while performing synchronous welding operations.

[0051] Furthermore, the path planning component 11 is used to perform the following methods: The weld area of ​​the steel pipe to be welded is scanned by a 3D vision scanning device to obtain point cloud data including the actual weld bevel shape, misalignment amount, and workpiece spatial pose. The point cloud data is matched and compared with the steel pipe design model to identify and extract the 3D trajectory of the centerline of the actual weld. Based on the extracted actual weld trajectory, the preset welding process database, and the multi-welding gun cooperative motion constraint parameters, a search is performed to obtain a welding path that meets the current actual working conditions, and the preset welding path plan is generated.

[0052] Furthermore, the path planning component 11 is used to perform the following methods: Based on the geometric characteristics of the actual weld trajectory and the heat input requirements of the welding process, the trajectory is divided into straight line segments, circular arc segments, and characteristic transition segments. A set of welding process parameter constraints is parameterized and associated for each trajectory segment, including: the allowable welding speed range and the welding torch attitude angle range. Based on the welding process parameter constraints of each segment, and combined with the kinematic model and spatial layout constraint parameters of multiple welding execution units, collision-free cooperative motion simulation optimization is performed to determine the cooperative start-stop sequence and spatial path allocation of each welding execution unit, thereby obtaining the welding path.

[0053] Furthermore, the strategy optimization component 13 is used to perform the following method: The main controller, according to a preset synchronization clock, timestamps and encapsulates the molten pool image information, temperature distribution information, and arc voltage and current information to construct a time-series data frame with a unified structure. It normalizes each data item in the time-series data frame and extracts its features. The extracted features are then concatenated into vectors to construct a high-order feature tensor describing the instantaneous welding state of each weld point. This high-order feature tensor is used as input to the welding process optimization model. Based on the current welding state, and combined with the output action vector and reward feedback of the internal policy network, the model performs optimization to determine the process parameter strategy that satisfies the synchronous welding objective, and outputs the working parameter optimization strategy. Finally, it sends a process parameter correction command based on the working parameter optimization strategy.

[0054] Furthermore, the strategy optimization component 13 is used to perform the following method: The molten pool width and trailing angle features are extracted from the molten pool image; the peak temperature and residence time in the high-temperature zone features are extracted from the temperature distribution; and the energy concentration features are extracted from the arc electrical parameters.

[0055] Furthermore, the strategy optimization component 13 is used to perform the following method: A hierarchical control structure is adopted to send process parameter correction instructions. Based on the overall state of the weld joint, the basic parameter settings for maintaining the collaborative benchmark are calculated, and collaborative instructions for each execution unit are generated and sent to each welding execution unit. According to the state deviation of the weld joint, the parameter compensation value for eliminating local deviations is calculated, and local compensation instructions are generated and sent to the corresponding welding execution unit. Each welding execution unit superimposes the received collaborative instructions and local compensation instructions as the final process parameter settings to be executed.

[0056] Furthermore, the welding control component 14 is used to perform the following methods: Based on the process parameter correction command, the updated multi-source detection data after parameter adjustment is tracked and executed; the heat input balance index among all weld points is calculated based on the updated multi-source detection data; based on the heat input balance index, with minimizing the balance index as the optimization objective, and combined with the molten pool image features of each weld point, the process parameter adjustment amount for each welding execution unit is calculated; the tracking and adjustment steps are repeated until welding is completed, wherein the heat input balance index is stabilized within a preset threshold, achieving adaptive balance between welding heat input and molten pool behavior, so as to ensure consistent welding quality of steel pipe joint welds in each welding execution unit.

[0057] Furthermore, the welding control component 14 is used to perform the following methods: Based on updated multi-source detection data, the instantaneous heat input value of each solder joint is calculated in real time; based on the instantaneous heat input value of each solder joint, the variance or range of the instantaneous heat input values ​​of all solder joints is calculated as an indicator to measure the heat input balance.

[0058] Furthermore, the welding control component 14 is used to perform the following methods: Welding execution units are positioned based on process parameter correction instructions. When the process parameter correction instructions are for correcting the working parameters of a local weld point, the coupling interference amount on the molten pool stability, arc morphology, and local heat input of adjacent weld points is predicted based on the thermo-mechanical coupling relationship between the molten pools of the steel pipe weld points. According to the coupling interference amount, the main controller generates a feedforward compensation instruction to cancel the interference. The feedforward compensation instruction is used to cancel the interference amount and maintain the original coordinated thermal process and synchronous motion relationship. Based on the feedforward compensation instruction and the process parameter correction instructions, each welding execution unit is controlled to adjust the corresponding process parameters.

[0059] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A method of controlling multi-point simultaneous welding of a steel pipe, characterized by, The method includes: The main controller generates synchronous motion commands based on the preset welding path plan and sends them to multiple plasma arc welding execution units; By using the vision sensor, temperature sensor and arc electrical parameter acquisition module associated with each welding execution unit, the molten pool image information, temperature distribution information and arc voltage and current information of the corresponding weld point are acquired in real time, and a multi-source monitoring data network is constructed. Based on the multi-source monitoring data network, the main controller receives and integrates the multi-source monitoring data, calls the built-in welding process optimization model, executes synchronous welding strategy optimization, and sends process parameter correction instructions, wherein the process parameters include at least welding current, plasma gas flow rate, and welding speed. Based on the process parameter correction instructions, each welding execution unit is controlled to adjust the process parameters with collaborative constraints, so as to maintain the relative motion relationship and thermal process coordination between each weld point while performing synchronous welding operations.

2. The steel pipe multi-point synchronous welding control method according to claim 1, characterized by, Before generating synchronous motion commands based on the preset welding path plan, the following steps are included: The weld area of ​​the steel pipe to be welded is scanned by a three-dimensional vision scanning device to obtain point cloud data including the actual bevel morphology of the weld, the amount of misalignment, and the spatial pose of the workpiece. The point cloud data is matched and compared with the steel pipe design model to identify and extract the three-dimensional trajectory of the centerline of the actual weld. Based on the extracted actual weld trajectory, the preset welding process database, and the multi-welding gun cooperative motion constraint parameters, a search is performed to obtain a welding path that meets the current actual working conditions, and the preset welding path plan is generated.

3. The method for controlling multi-point synchronous welding of steel pipes according to claim 2, characterized in that, Based on the extracted actual weld trajectory, the preset welding process database, and the multi-welding torch cooperative motion constraint parameters, a search is performed to obtain a welding path that meets the current actual working conditions, including: Based on the geometric characteristics of the actual weld trajectory and the heat input requirements of the welding process, the trajectory is divided into straight line segments, circular arc segments, and characteristic transition segments; For each trajectory segment, a set of welding process parameter constraints are parametrically associated, including: the allowable welding speed range and the welding torch attitude angle range; Based on the welding process parameters of each segment, and combined with the kinematic model and spatial layout constraint parameters of multiple welding execution units, collision-free cooperative motion simulation optimization is performed to determine the cooperative start-stop sequence and spatial path allocation of each welding execution unit, thereby obtaining the welding path.

4. The method for controlling multi-point synchronous welding of steel pipes according to claim 1, characterized in that, The main controller receives and fuses multi-source monitoring data, invokes the built-in welding process optimization model, executes synchronous welding strategy optimization, and sends process parameter correction instructions, including: The main controller performs timestamp alignment and data encapsulation on the molten pool image information, temperature distribution information, arc voltage and current information according to a preset synchronization clock, and constructs a time-series data frame with a unified structure. The data in the time-series data frame are normalized and the features of each data are extracted; The extracted data features are concatenated into vectors to construct a high-order feature tensor that describes the instantaneous welding state of each weld point; The high-order feature tensor of the instantaneous welding state of each weld point is used as the input of the welding process optimization model. Based on the current welding state, the model is optimized by combining the output action vector of the internal policy network and the reward feedback to determine the process parameter strategy that satisfies the synchronous welding goal and output the working parameter optimization strategy. Based on the aforementioned operating parameter optimization strategy, a process parameter correction instruction is sent.

5. The method for controlling multi-point synchronous welding of steel pipes according to claim 4, characterized in that, Extract various data features, including: extracting the molten pool width and trailing angle features from the molten pool image; extracting the peak temperature and residence time in the high-temperature zone features from the temperature distribution; and extracting the energy concentration features from the arc electrical parameters.

6. The method for controlling multi-point synchronous welding of steel pipes according to claim 5, characterized in that, Based on the aforementioned operating parameter optimization strategy, a process parameter correction instruction is sent, including: A hierarchical control structure is adopted to send process parameter correction instructions. Based on the overall state of the weld joint, the basic parameter settings for maintaining the collaborative benchmark are calculated, collaborative instructions for each execution unit are generated, and sent to each welding execution unit. Based on the state deviation of the weld joint, calculate the parameter compensation value to eliminate the local deviation, generate a local compensation command, and send it to the corresponding welding execution unit; Each welding execution unit superimposes the received coordination instructions and local compensation instructions to form the final process parameter setting value.

7. The method for controlling multi-point synchronous welding of steel pipes according to claim 1, characterized in that, Also includes: Based on the process parameter correction command, track and execute the updated multi-source detection data after the parameter adjustment; Calculate the heat input uniformity index among all solder joints based on the updated multi-source detection data; Based on the heat input uniformity index, with minimizing the uniformity index as the optimization objective, and combined with the molten pool image features of each weld point, the adjustment amount of process parameters for each welding execution unit is calculated. The tracking and adjustment steps are repeated until welding is completed. The heat input balance index is stabilized within the preset threshold to achieve adaptive balance between welding heat input and molten pool behavior, so as to ensure consistent welding quality of steel pipe joint welds in each welding execution unit.

8. The method for controlling multi-point synchronous welding of steel pipes according to claim 7, characterized in that, The thermal input uniformity index among all solder joints was calculated based on the updated multi-source detection data, including: Based on updated multi-source detection data, the instantaneous heat input value of each weld point is calculated in real time; Based on the instantaneous heat input value of each solder joint, calculate the variance or range of the instantaneous heat input values ​​of all solder joints as an indicator of heat input balance.

9. The method for controlling multi-point synchronous welding of steel pipes according to claim 1, characterized in that, Based on the process parameter correction command, control each welding execution unit to adjust process parameters with cooperative constraints, including: Welding execution unit positioning is performed based on process parameter correction instructions. When the process parameter correction instructions are for working parameter correction of local weld points, the coupling interference amount on the molten pool stability, arc morphology and local heat input of adjacent weld points is predicted based on the thermo-mechanical coupling relationship between the molten pools of the steel pipe weld points. Based on the amount of coupling interference, the main controller generates a feedforward compensation command to cancel the interference. The feedforward compensation command is used to cancel the amount of interference and maintain the original cooperative thermal process and synchronous motion relationship. Based on the feedforward compensation command and the process parameter correction command, each welding execution unit is controlled to adjust the corresponding process parameters.

10. A multi-point synchronous welding control system for steel pipes, characterized in that, The system is used to implement the multi-point synchronous welding control method for steel pipes according to any one of claims 1-9, the system comprising: Path planning component: The main controller generates synchronous motion commands based on the preset welding path planning and sends them to multiple plasma arc welding execution units; Data acquisition component: Through the vision sensor, temperature sensor and arc electrical parameter acquisition module associated with each welding execution unit, the molten pool image information, temperature distribution information and arc voltage and current information of the corresponding weld point are acquired in real time to build a multi-source monitoring data network; Strategy optimization component: Based on the multi-source monitoring data network, the main controller receives and integrates the multi-source monitoring data, calls the built-in welding process optimization model, performs synchronous welding strategy optimization, and sends process parameter correction instructions, wherein the process parameters include at least welding current, plasma gas flow rate, and welding speed. Welding control component: Based on the process parameter correction command, it controls each welding execution unit to adjust the process parameters with collaborative constraints, and maintains the relative motion relationship and thermal process coordination between each weld point while performing synchronous welding operations.

Images