Modular assembly type welding tooling process for extra-high voltage iron tower

Abstract

The application relates to the technical field of welding processing, in particular to a modular assembly type welding tool process for an extra-high voltage iron tower. The process comprises three stages: a macroscopic disturbance position searching and assembly stress elimination stage, a position-parameter collaborative control stage in a welding process and a closed-loop feedback and rigidity model updating stage. The three stages solve the problems of large tool assembly stress, poor weld forming and model lag existing in the prior art, realize the whole-process closed-loop control logic of "macroscopic disturbance position searching-position parameter collaboration-closed-loop rigidity updating", and significantly improve the dimensional accuracy and structural reliability of the modular welding of the extra-high voltage iron tower.

Description

Technical Field

[0001] This application relates to the field of welding processing technology, and in particular to a modular assembly welding fixture process for ultra-high voltage transmission towers. Background Technology

[0002] Ultra-high voltage (UHV) transmission towers are the core supporting structures in power transmission lines, and their manufacturing typically employs modular prefabrication and on-site assembly and welding. In the factory, the tower is disassembled into several modules (such as tower leg modules, tower body modules, crossarm modules, etc.), which are then individually fabricated and welded together using welding fixtures. The quality of modular welding directly affects the tower's verticality, the tightness of joint fit, and the uniformity of load-bearing capacity.

[0003] Currently, the welding process for ultra-high voltage (UHV) tower modules mainly employs the following method: the module is hoisted onto a fixture base, clamped into a preset position using a clamping mechanism, and then welded by a welding robot according to a fixed trajectory and parameters. However, this process has the following technical problems: First, geometric deviations are unavoidable during the rolling, cutting, and transportation of tower modules, such as straightness of angle steel, end face perpendicularity errors, and local burrs. When the modules are hoisted into the tooling, the forced clamping of the clamping mechanism causes forced elastic deformation of the modules, resulting in residual assembly stress. This stress will be superimposed on the welding thermal stress during the welding process, causing cracks, undercut, or increased angular deformation in the weld area, affecting the overall load-bearing performance of the tower.

[0004] Secondly, the gaps in modular butt welds are often uneven, potentially exhibiting large gaps or misalignment in certain areas. When using fixed welding parameters (current, voltage, wire feed speed, welding speed), large gap areas are prone to insufficient penetration due to insufficient heat input, or burn-through due to excessive heat input; conversely, reducing parameters in areas with normal gaps reduces welding efficiency. Current technology lacks a method to adaptively adjust welding energy input and filler volume based on the local condition of the weld.

[0005] Third, welding thermal deformation has a cumulative effect, and the mechanical properties (such as yield strength and coefficient of thermal expansion) of different modules fluctuate. Traditional processes use fixed reverse deformation amounts or welding sequences, which are difficult to adapt to actual thermal deformation conditions, resulting in inconsistent forming accuracy of multiple welds, and even requiring post-weld straightening.

[0006] Therefore, there is an urgent need for an intelligent welding tooling process that can eliminate assembly stress, adapt to changes in weld gap, and learn the laws of thermal deformation. Summary of the Invention

[0007] To address the problems of high assembly stress, poor weld formation, and model lag in the aforementioned background technology, this invention provides a modular assembly welding fixture process for ultra-high voltage transmission towers.

[0008] The present invention adopts the following technical solution: According to a first aspect of the embodiments of this specification, a modular assembly welding fixture process for ultra-high voltage transmission towers is provided, the process comprising the following three stages: Phase 1: Macroscopic Disturbance Positioning and Assembly Stress Elimination Phase: After hoisting the first and second tower modules onto the welding fixture, the clamping mechanism of the welding fixture is controlled to apply macroscopic disturbances to the modules. During the application of the macroscopic disturbances, the reaction force or torque signal on the clamping mechanism is monitored in real time. When it is determined that there is an asymmetry in assembly stress based on the reaction force or torque signal, the position or clamping force of the clamping mechanism is adjusted until the assembly stress is eliminated. Phase Two: Position-Parameter Coordinated Control Phase During Welding: The welding process is initiated, and sensors continuously monitor the weld's condition. When a large gap or misalignment is detected, it is determined to be a defective weld edge, automatically reducing the welding heat input and increasing the wire feed. When an ideal butt joint is detected, it is determined to be a weld center state, restoring high-efficiency welding parameters. The welding robot's trajectory and the welding power supply's parameters are dynamically adjusted synchronously based on the weld's condition. Phase 3: Closed-loop feedback and stiffness model update phase: During the welding process, welding parameter fluctuation data and deformation feedback data of the welding fixture are collected in real time; when it is determined that the thermal deformation of a certain tower module is abnormal based on the welding parameter fluctuation data and / or the deformation feedback data, the welding sequence of subsequent welds and the preset deformation compensation amount are dynamically adjusted to update the stiffness model of the welding fixture.

[0009] Furthermore, the determination of the existence of assembly stress asymmetry includes: Acquire the reaction force or torque signals of the clamping mechanism in both positive and negative directions of the macroscopic disturbance; Calculate the degree of asymmetry of the reaction force or torque signals in the two opposite directions; When the degree of asymmetry exceeds a preset threshold, it is determined that there is an asymmetry in assembly stress. Adjusting the position or clamping force of the clamping mechanism includes: determining the interference direction based on the sign of the degree of asymmetry, and adjusting the actual position of the clamping mechanism or the tower module in a direction opposite to the interference direction.

[0010] Furthermore, the reduction of welding heat input includes: reducing welding speed, reducing welding current and / or reducing welding voltage; the increase of wire feed includes: increasing wire feed speed; the high-efficiency welding parameters include: welding heat input higher than the reduced welding heat input, and wire feed lower than the increased wire feed.

[0011] Furthermore, the trajectory movement of the welding robot and the welding parameters of the welding power source are dynamically adjusted synchronously according to the weld condition, including: A central controller is provided, which is communicatively connected to the motion controller of the welding robot and the power controller of the welding power source; The central controller synchronously generates trajectory control commands and welding parameter control commands based on the weld seam status detected by the sensors. The motion controller controls the movement of the end effector of the welding robot according to the trajectory control command, and the power controller controls the output parameters of the welding power supply according to the welding parameter control command.

[0012] Furthermore, the determination of abnormal thermal deformation of a certain tower module includes: The actual heat input at the current welding position is calculated based on the real-time collected fluctuations in welding current and / or welding voltage. When the deviation between the actual heat input and the theoretical heat input exceeds a preset heat input threshold, or when the rate of change of the deformation feedback data of the welding fixture exceeds a preset deformation rate threshold, the amount of thermal deformation is determined to be abnormal.

[0013] Furthermore, updating the stiffness model of the welding fixture includes: The correction value of the deformation compensation amount corresponding to the abnormal thermal deformation amount is used as the feedback amount; The preset stiffness parameters of the welding fixture are updated by weighted fusion or Kalman filtering using the feedback quantity to obtain an updated stiffness model. The updated stiffness model is used to determine the preset deformation compensation amount of the next tower module to be welded.

[0014] Furthermore, the macroscopic disturbance is a small oscillation or vibration applied by the clamping mechanism to the tower module in a specific direction, the amplitude of the small oscillation or vibration is within the elastic deformation range of the tower module, and the scale of the macroscopic disturbance is on the order of meters.

[0015] Furthermore, the sensor includes a vision sensor and a force sensor; the large gap or misalignment state includes a weld gap width greater than a preset width threshold, or a misalignment amount between two tower modules greater than a preset misalignment threshold.

[0016] Furthermore, reducing welding heat input and increasing wire feed includes switching to a cold metal transition welding mode; restoring high-efficiency welding parameters includes switching to a conventional pulse welding mode.

[0017] Furthermore, in the second stage, the welding parameters also include oscillation parameters, which include oscillation amplitude, oscillation frequency, and oscillation trajectory; when the weld is determined to be in an edge defect state, the oscillation amplitude is reduced; when the weld is determined to be in a center state, the preset oscillation amplitude is restored.

[0018] According to a second aspect of the embodiments of this specification, a modular assembly welding fixture system for ultra-high voltage transmission towers is provided. The system is used to perform the aforementioned modular assembly welding fixture process, and the system includes: The welding fixture body includes a fixture base and a plurality of clamping mechanisms disposed on the fixture base, the clamping mechanisms being used to fix the iron tower module to be welded; The drive unit, integrated on the clamping mechanism, is used to apply macroscopic disturbances to the tower module; The detection unit includes a vision sensor, a force sensor and / or a torque sensor, used to detect the weld condition, reaction force or torque signal and tooling deformation feedback data in real time. Welding execution unit, including welding robot and welding power source; The control unit is communicatively connected to the drive unit, the detection unit, the welding robot, and the welding power supply, and is configured to execute the modular assembly welding fixture process.

[0019] This application also provides a computer-readable storage medium that enables a processor to implement the above-described method through program instructions, thereby achieving a higher degree of automation and reliability.

[0020] Compared with the prior art, the beneficial effects of the present invention include at least the following: This application introduces a macroscopic disturbance positioning mechanism during the module assembly stage. By applying minute oscillations and monitoring torque asymmetry, it actively eliminates assembly stress caused by module geometric deviations and hoisting positioning errors, allowing welding to begin when the module is under no additional stress. This not only avoids residual stress caused by forced assembly but also improves the consistency of weld formation and the overall stress uniformity of the tower.

[0021] This application achieves differentiated energy input between the poor state of the weld edge and the ideal state of the center during the welding stage: when a large gap or misalignment is detected, the heat input is automatically reduced and the wire feed is increased (e.g., switching to cold metal transition mode) to prevent burn-through or lack of fusion; when an ideal butt joint is detected, high-efficiency parameters are restored. By synchronously adjusting the robot trajectory and welding parameters through a central controller, position-energy coordinated control is achieved, ensuring both the welding quality of the edge area and maintaining the welding efficiency of the center area.

[0022] This application uses real-time data collection of welding parameter fluctuations and tooling deformation feedback to online identify abnormal thermal deformation and dynamically adjust the welding sequence and preset deformation compensation of subsequent welds, effectively updating the tooling stiffness model. This closed-loop adaptive mechanism enables the process to adapt to fluctuations in the mechanical properties of different modules, significantly improving the consistency of forming accuracy among multiple welds.

[0023] This application further introduces a deformation compensation model and subsequent feedback updates, which continuously optimizes the model parameters by using the difference between the actual forming coordinates after welding and the preset coordinates. This allows the determination of welding sequence and compensation amount to evolve from fixed empirical values ​​to data-driven self-learning values, resulting in continuous improvement in welding accuracy over long-term operation. Attached Figure Description

[0024] Figure 1 This is a schematic flowchart of a modular assembly welding fixture process for ultra-high voltage transmission towers according to an embodiment of the present invention. Figure 2 This is a schematic diagram of the logic flow of stage one, "macroscopic disturbance location search," according to an embodiment of the present invention. Figure 3 This is a schematic diagram of the logic flow of stage two "position-parameter coordinated control" according to an embodiment of the present invention; Figure 4 This is a schematic diagram of the logic flow of stage three, "closed-loop feedback and stiffness model update," according to an embodiment of the present invention. Detailed Implementation

[0025] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0026] The terminology used in this application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. The singular forms “a,” “the,” and “the” used in this application and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any or all possible combinations of one or more of the associated listed items.

[0027] It should be understood that although the terms first, second, third, etc., may be used in this application to describe various information, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, without departing from the scope of this application, first information may also be referred to as second information, and similarly, second information may also be referred to as first information. Depending on the context, the word "if" as used herein may be interpreted as "when," "when," or "in response to determination."

[0028] In the embodiments of the present invention, it should be noted that: Ultra-high voltage (UHV) transmission tower modules refer to the prefabricated steel structure units that make up UHV transmission towers. They are typically welded from angle steel, steel plates, and other profiles, and include tower leg modules, tower body modules, crossarm modules, etc. A single module can be 6 to 14 meters long and weigh several tons to more than ten tons.

[0029] Modular assembly welding fixtures refer to specialized process equipment used to position, clamp, and weld multiple tower modules together. They typically include a fixture base, clamping mechanism, positioning device, and welding execution mechanism (such as a welding robot).

[0030] In this application, macroscopic disturbance is distinguished from the microscopic oscillation (millimeter-level, millisecond-level) of the galvanometer in handheld laser welding equipment. Macroscopic disturbance refers to the low-frequency, small-amplitude reciprocating oscillation or vibration applied to the tower module by the drive unit of the clamping mechanism. Its spatial scale is on the order of meters (displacement amplitude ±2mm to ±10mm), and its temporal scale is on the order of seconds, with the amplitude limited to the elastic deformation range of the module material. Its physical significance lies in: stimulating the mechanical interaction between the module and the clamping mechanism through active disturbance, thereby detecting and eliminating assembly stress.

[0031] Assembly stress asymmetry refers to the unequal external constraint reaction forces experienced by a tower module in the feeding direction (or weld normal direction) and opposite directions when the module is forcibly placed in a certain position due to geometric deviations (such as straightness or end face tilt) or external constraints (such as block interference or burr jamming). This results in forced elastic deformation within the module. This application quantifies the degree of this asymmetry by monitoring the difference in torque feedback signals from the clamping mechanism in the positive and negative disturbance directions.

[0032] In this application, position-parameter coordinated control specifically refers to using the state of the weld (imperfect edge state or center state) as the source of synchronous control commands. The central controller simultaneously sends adjustment commands to the motion controller of the welding robot and the power controller of the welding power supply to achieve dynamic coordination of "reducing energy input when the weld position is poor and restoring high-efficiency input when the weld position is good".

[0033] In this application, the stiffness model refers to the equivalent mathematical model describing the elastic deformation of the welding fixture, tower module, and welding actuator under stress. This model is used to calculate the required preset deformation compensation (such as reverse deformation) based on the expected thermal deformation. This application updates the stiffness model parameters online by collecting welding parameter fluctuations and fixture deformation feedback data in real time, adapting it to the actual mechanical properties of the current module.

[0034] This application provides a modular assembly welding fixture process for ultra-high voltage (UHV) transmission towers. UHV transmission towers are typically pre-assembled and welded in a factory from multiple large steel structure modules (such as tower leg modules, crossarm modules, etc.). Due to the enormous size (meters in diameter), heavy weight, and long butt welds of the modules, traditional rigid fixing and fixed-parameter welding methods are insufficient to guarantee weld quality and overall tower precision.

[0035] To address the issues of high tooling assembly stress, poor weld formation, and model lag in the aforementioned background technologies, this application proposes a full-process closed-loop control logic of "macroscopic disturbance positioning - position parameter coordination - closed-loop stiffness update".

[0036] Please see Figure 1-4 The welding fixture process provided in this application mainly includes the following three stages: Phase 1: Macroscopic Disturbance Positioning and Assembly Stress Elimination. The core of this phase is resolving the forced assembly issue after module hoisting. Unlike the traditional "lock-in upon positioning" mode, this application controls the clamping mechanism to apply minute macroscopic disturbances (such as slight swaying or vibration) to the meter-scale tower modules. By real-time monitoring of the reaction force or torque signal on the clamping mechanism, the degree of torque asymmetry in both directions is calculated. When the asymmetry exceeds a threshold, assembly stress interference is determined, and the position or clamping force of the clamping mechanism is adjusted accordingly until the module is in a free state without additional stress. This achieves "stress-free positioning" of large-scale components.

[0037] Phase Two: Position-Parameter Coordinated Control Phase During Welding. The core of this phase lies in addressing the adaptability issue of uneven weld gaps. After welding begins, the system uses visual and force sensors to monitor the weld status in real time. When an edge defect ("large gap or misalignment") is detected, the system automatically reduces the welding heat input (e.g., reducing current / speed) and increases the wire feed (or switches to CMT mode), while simultaneously adjusting the oscillation parameters. When a center defect ("ideal butt joint") is detected, high-efficiency welding parameters are restored. This strategy of synchronously and dynamically adjusting "position status (edge / center)" and "welding parameters (heat input / wire feed / oscillation)" ensures the stability and quality of the weld pool under different joint conditions.

[0038] Phase Three: Closed-Loop Feedback and Stiffness Model Update. The core of this phase is addressing the accuracy drift caused by accumulated thermal deformation. During welding, the system collects welding current and voltage fluctuations (reflecting heat input) and tooling deformation feedback data in real time. Once an abnormality in thermal deformation is detected (such as excessive heat input deviation or excessively rapid deformation rate), the system immediately and dynamically adjusts the welding sequence of subsequent welds and the preset deformation compensation amount. More importantly, the system uses the deformation compensation correction value from this weld, through weighted fusion or Kalman filtering algorithms, to update the preset stiffness parameters of the tooling, generating an updated stiffness model. This means that the tooling's "brain" can continuously "learn" and "evolve" as the welding process progresses, adapting to the laws of thermal deformation.

[0039] The technical solution provided in this application eliminates the assembly stress of meter-level modules through macroscopic disturbance positioning in stage one; solves the forming problem of large-gap welds through position-parameter collaborative control in stage two; and achieves adaptive evolution of the tooling model through closed-loop stiffness updating in stage three. The combination of these three aspects forms a complete "perception-decision-execution-learning" closed loop, significantly improving the dimensional accuracy and structural reliability of modular welding for ultra-high voltage towers.

[0040] Specifically, Phase 1: Macroscopic Disturbance Positioning and Assembly Stress Elimination Phase, includes: after hoisting the first and second tower modules onto the welding fixture, controlling the clamping mechanism of the welding fixture to apply macroscopic disturbances to the modules; during the application of the macroscopic disturbances, monitoring the reaction force or torque signal on the clamping mechanism in real time; when it is determined that there is an asymmetry in assembly stress based on the reaction force or torque signal, adjusting the position or clamping force of the clamping mechanism until the assembly stress is eliminated. The first tower module includes any one of the prefabricated steel structure units of the main leg section, tower body section, or crossarm section of an ultra-high voltage tower. The second tower module includes a mating section, connecting plate, or diagonal member matching the first tower module. Both are prefabricated in the factory from high-strength angle steel or steel plates, and the length of a single module is 6 to 14 meters, with a weight of several tons to more than ten tons. The two form a weld joint to be joined on the welding fixture.

[0041] The determination of assembly stress asymmetry includes: acquiring the reaction force or torque signals of the clamping mechanism in both directions of the macroscopic disturbance; calculating the degree of asymmetry of the reaction force or torque signals in both directions; determining that assembly stress asymmetry exists when the degree of asymmetry exceeds a preset threshold; and adjusting the position or clamping force of the clamping mechanism includes: determining the interference direction according to the sign of the degree of asymmetry, and adjusting the actual position of the clamping mechanism or the tower module in the opposite direction to the interference direction.

[0042] Please continue reading. Figure 1 and Figure 2 The core of Phase One lies in solving the "forced assembly" problem of large-scale modules (such as tower legs and crossarms) of UHV transmission towers after hoisting and positioning. Traditional rigid fixtures often lock the modules directly, resulting in huge residual assembly stress inside the modules. This stress is very likely to cause cracks or deformation under the subsequent welding thermal cycle. To solve this problem, this application proposes a stress-free positioning mechanism based on "active disturbance-force feedback-adaptive adjustment".

[0043] In this embodiment, the specific implementation process of Phase One is as follows: After the first tower module (e.g., the main section of the tower leg) and the second tower module (e.g., the inclined or butt joint section of the tower leg) are hoisted to the preset position of the welding fixture, the clamping mechanism does not immediately perform a rigid locking action, but enters a "flexible positioning" mode. The control unit first controls the drive unit integrated on the clamping mechanism to apply a small macroscopic disturbance to the tower module. This macroscopic disturbance is not random vibration, but a low-frequency reciprocating oscillation or micro-amplitude vibration along a specific direction (e.g., the weld normal or tangential direction). Its displacement amplitude is usually controlled within the range of ±2mm to ±10mm, and is strictly limited within the elastic deformation range of the module material to ensure that it does not cause damage to the base material.

[0044] During the application of macroscopic disturbances, the detection unit (including a torque sensor and a six-dimensional force sensor installed at the joints of the clamping mechanism) collects the reaction force or torque signals experienced by the clamping mechanism in real time and at high frequency. The system in this embodiment does not simply monitor the magnitude of the force, but focuses on analyzing the symmetry of the force. Specifically, the system acquires the reaction force or torque values ​​of the module at the two extreme positions of forward and reverse disturbances, and calculates the degree of asymmetry between them (e.g., by calculating the absolute value of the difference |F). 正 -F 反 | or ratio difference). In this embodiment, the operating principle is as follows: when the module is in a completely free and stress-free state, its motion resistance (manifested as reaction force) in both positive and negative directions should be basically symmetrical; conversely, if there are geometric deviations (such as the angle steel end face not being perpendicular) or local interference (such as burr jamming), the module's movement in a certain direction will be more constrained, resulting in a significant asymmetry in the reaction force or torque signal. Therefore, the degree of asymmetry becomes a quantitative indicator for measuring the existence of assembly stress.

[0045] When the system determines that the asymmetry exceeds a preset threshold, it immediately identifies the existence of assembly stress asymmetry. At this point, the system accurately determines the direction of interference based on the sign of the asymmetry (i.e., which side has a greater force). For example, if the reaction force during a forward disturbance is significantly greater than that during a reverse disturbance, it indicates that the module is interfering in the forward direction. Subsequently, the control unit instructs the fine-tuning unit of the clamping mechanism or tooling base to make a slight translation or rotation adjustment of the actual position of the clamping mechanism or tower module in the opposite direction to the interference direction. This process is repeated until the reaction force or torque signals in both directions tend to balance, and the asymmetry falls back below the threshold. At this point, the module is considered to be in a stress-free state, and the clamping mechanism can then perform the final locking action.

[0046] Example: In the actual assembly of UHV tower leg modules, assuming the first module is made of Q420 angle steel and the second module is a connecting plate, due to transportation deformation, the end face of the connecting plate has a 3mm verticality deviation. If forcibly clamped directly, a tensile stress of up to 150MPa will be generated at the root of the connecting plate. After adopting the process described in this first stage, when a macroscopic disturbance of ±5mm is applied to the clamping mechanism, the sensor detects a positive torque M. 正 Reverse torque M 反 The error was 25%. Based on this, the system determined that there was interference in the positive direction and automatically controlled the hydraulic push rod to finely adjust the connecting plate module 1.5mm in the negative direction and tilt it by 0.5°. During the second disturbance test, the torque asymmetry was reduced to within 3%. At this time, the locking fixture was used for welding. Post-weld inspection showed that there were no obvious stress cracks in the weld area, and the geometric dimensional accuracy was improved by 40%.

[0047] As can be seen, this embodiment, by introducing a macroscopic disturbance positioning mechanism, transforms "forced compression" into "flexible alignment," thereby eliminating the source of assembly stress from a physical perspective and laying a solid foundation for subsequent high-quality welding.

[0048] Specifically, Phase Two: the position-parameter coordinated control phase during the welding process, includes: starting the welding process and using sensors to detect the weld's state in real time; when a large gap or misalignment is detected, it is determined to be a defective weld edge, automatically reducing the welding heat input and increasing the wire feed; when a weld is detected to be in an ideal butt joint state, it is determined to be a weld center state, restoring high-efficiency welding parameters; wherein, the trajectory movement of the welding robot and the welding parameters of the welding power supply are synchronously and dynamically adjusted according to the weld state. The sensors include vision sensors and force sensors; the large gap or misalignment state includes a weld gap width greater than a preset width threshold, or a misalignment between two tower modules greater than a preset misalignment threshold. Reducing the welding heat input and increasing the wire feed includes: switching to a cold metal transition welding mode; restoring high-efficiency welding parameters includes: switching to a conventional pulse welding mode.

[0049] The reduction of welding heat input includes: reducing welding speed, reducing welding current and / or reducing welding voltage; the increase of wire feed includes: increasing wire feed speed; the high-efficiency welding parameters include: welding heat input higher than the reduced welding heat input, and wire feed lower than the increased wire feed, and also include oscillation parameters, which include oscillation amplitude, oscillation frequency and oscillation trajectory; when the weld is determined to be in an edge defect state, the oscillation amplitude is reduced; when the weld is determined to be in a center state, the preset oscillation amplitude is restored.

[0050] The trajectory movement of the welding robot and the welding parameters of the welding power source are dynamically adjusted synchronously according to the weld condition, including: A central controller is provided, which is communicatively connected to the motion controller of the welding robot and the power controller of the welding power source; The central controller synchronously generates trajectory control commands and welding parameter control commands based on the weld seam status detected by the sensors. The motion controller controls the movement of the end effector of the welding robot according to the trajectory control command, and the power controller controls the output parameters of the welding power supply according to the welding parameter control command.

[0051] After successfully eliminating assembly stress between tower modules in Phase One and placing the modules in a stress-free state, the welding process officially began. However, eliminating assembly stress cannot completely solve the welding quality problem—due to module manufacturing tolerances, cutting bevel deviations, and residual errors in hoisting and positioning, the gap width and misalignment along the weld seam are often uneven. In the actual production of UHV tower modules, on a weld seam several meters long, the gap may vary from 0.2mm (ideal butt joint) to 2.5mm (large gap), and even local misalignment may occur. If uniform welding parameters are used, the large gap area is prone to burn-through or depression due to insufficient molten pool filling, while using too low parameters in the ideal butt joint area will reduce efficiency. Therefore, the core task of Phase Two is to further achieve dynamic and coordinated control of weld seam position and welding energy based on the "stress-free benchmark" provided in Phase One.

[0052] Please continue reading. Figure 1 and Figure 3 Phase Two is automatically triggered upon completion of Phase One, requiring no manual intervention. This embodiment provides a synchronous dynamic adjustment architecture based on a central controller: the central controller is communicatively connected to both the motion controller of the welding robot and the power controller of the welding power source. Based on the weld seam status detected by sensors, the central controller synchronously generates trajectory control commands and welding parameter control commands; the motion controller controls the movement of the welding robot's end effector according to the trajectory control commands, and the power controller controls the output parameters of the welding power source according to the welding parameter control commands. This architecture ensures real-time linkage between position (such as trajectory, speed, and oscillation) and parameters (such as current, voltage, wire feed, and welding mode).

[0053] The specific implementation process of Phase Two is as follows: The welding robot starts and moves the welding torch along a preset trajectory, while a sensor system monitors the weld condition in real time. The sensors include vision sensors (e.g., laser structured light cameras) and force sensors (e.g., contact weld tracking sensors). The vision sensor scans an area 10mm to 30mm in front of the weld at a frequency of at least 50Hz to obtain the weld gap width G and misalignment Δh; the force sensor assists in detecting bevel geometry changes through a contact probe or arc sensing method. The central controller classifies the weld condition in real time according to preset thresholds. When G ≤ 0.5mm and Δh ≤ 0.5mm, it is considered to be in "center state" (ideal butt joint). At this time, the weld geometry is good, and high-efficiency welding parameters can be used to ensure production cycle time.

[0054] When G>1.5mm or Δh>1.0mm, it is judged as "imperfect edge condition". At this time, the weld is in unfavorable conditions of large gap or misalignment. The heat input must be reduced and the filler volume increased to prevent defects such as burn-through, lack of fusion or undercut.

[0055] It should be noted that the above thresholds (0.5mm, 1.5mm, 1.0mm) are example values. In actual applications, they can be calibrated and adjusted according to the module material (such as Q420, Q460 high-strength steel) and plate thickness.

[0056] When the central controller determines that the edge is in a defective state, it immediately performs the following multi-dimensional collaborative adjustment actions: (1) Reduce welding heat input: The central controller sends instructions to the power controller to reduce the welding current (e.g., from 280A to 180A) and / or reduce the welding voltage; at the same time, it sends instructions to the motion controller to reduce the welding speed (e.g., from 0.8m / min to 0.4m / min). The reduction in heat input directly reduces the width and depth of the molten pool, avoiding the loss of liquid metal in large gaps.

[0057] (2) Increase wire feed rate: The central controller sends a command to the controller of the wire feeding mechanism (which can be integrated into the power controller) to increase the wire feeding speed (e.g., from 5 m / min to 8 m / min). The increased filler metal can compensate for the lack of metal in the large gap area and prevent the formation of dents or burn-through.

[0058] (3) Switching welding mode: The central controller sends a mode switching command to the power controller to switch the welding mode from conventional pulse welding to cold metal transfer (CMT) welding mode. In CMT mode, the welding current is almost zero through mechanical retraction during short-circuit transfer, resulting in extremely low heat input, which is particularly suitable for welding large gaps or thin plates.

[0059] (4) Adjusting swing parameters: The central controller sends swing parameter adjustment instructions to the motion controller to reduce the swing amplitude (e.g., from ±2mm to ±1mm), and can appropriately adjust the swing frequency or swing trajectory (e.g., change the circular swing to a straight swing) to reduce the swing width and concentrate the heat at the root of the bevel.

[0060] When the central controller determines that the state is in the center, it restores the preset high-efficiency welding parameters: the welding mode is switched from CMT back to the conventional pulse welding mode, the higher welding current (280A), the higher welding speed (0.8m / min), the lower wire feed speed (5m / min), and the preset oscillation amplitude (±2mm) and oscillation trajectory are restored.

[0061] The core mechanism of synchronous dynamic adjustment lies in the fact that the central controller simultaneously generates trajectory control commands (including position, speed, oscillation amplitude, oscillation frequency, and oscillation trajectory) and welding parameter control commands (including current, voltage, wire feed speed, and welding mode) within the same control cycle based on the same sensor signal (visual + force fusion). These two commands are sent to the motion controller and power controller respectively via a real-time industrial bus (such as EtherCAT or Profinet), with a time synchronization accuracy better than 10ms, ensuring that the welding parameters have been pre-switched to the correct position when the welding torch reaches the defective area. Furthermore, since the switching between CMT mode and conventional pulse mode requires a response from the power supply's internal state machine, the central controller sends a mode pre-switching command approximately 50ms in advance to ensure that the mode is stable when entering the defective area.

[0062] Phase Two maps the "position-power synergy" concept from the microscopic (laser galvanometer oscillation) to the macroscopic (weld trajectory and welding parameters), and further expands it into a multi-parameter synergy of "position-mode-oscillation". In handheld laser welding, the laser beam power is reduced at the edge of the oscillation trajectory to prevent edge overheating; in this application, at the "edge defective state" (large gap / misalignment), the weld simultaneously reduces heat input, increases wire feed, switches to CMT low heat input mode, and reduces oscillation amplitude, achieving multiple protections for the edge region. Meanwhile, the conventional pulse mode and high-efficiency parameters are restored in the center state, ensuring welding efficiency.

[0063] Continuing the example from Phase One: In the aforementioned tower leg module welding, after the disturbance and positioning in Phase One, the assembly stress between the connecting plate and the main material has been eliminated. After welding commenced, the visual sensor scanned along the weld seam and found that: the gap in the first 300mm area of ​​the weld seam was 0.3mm (center state), the gap in the middle 200mm area increased to 2.0mm due to beveling deviation (edge ​​defect state), and the gap in the last 400mm area returned to 0.4mm. The force sensor simultaneously confirmed that the misalignment reached 1.2mm in the middle area.

[0064] At the same moment the central controller detects a defective state at the edge of the intermediate region (within a detection cycle of 20ms), it simultaneously generates two commands: Send the following message to the power controller: "Current reduced to 160A, voltage reduced to 22V, wire feed speed increased to 9m / min, switch to CMT mode"; Send the following to the motion controller: "Welding speed reduced to 0.3m / min, oscillation amplitude reduced to ±0.5mm, oscillation frequency reduced from 3Hz to 2Hz, oscillation trajectory changed from circular to straight line."

[0065] Since CMT mode switching requires approximately 30ms of setup time, the central controller sends a pre-switching command 50ms in advance (i.e., when the welding torch is approximately 2.5mm from the start of the defective area) to ensure that CMT mode is already stably operating when the welding torch enters the 2.0mm gap area. The two commands are sent via the EtherCAT bus, and the measured timing deviation is less than 5ms.

[0066] Post-weld, the middle 200mm area was cross-sectionally inspected and subjected to X-ray flaw detection. The results showed good penetration, no burn-through, no lack of fusion, and no porosity. The weld reinforcement was uniform (0.5mm~0.8mm). Furthermore, due to the reduced oscillation amplitude, the width of the heat-affected zone decreased from the conventional 3mm to 1.5mm, effectively reducing overheating of the base material. In contrast, the comparative specimens welded using traditional fixed parameters (280A, 0.8m / min, 5m / min wire feed, conventional pulse mode, ±2mm oscillation) showed obvious burn-through holes and undercut defects in the same large gap area.

[0067] Meanwhile, since the assembly stress was eliminated in the first stage, the weld gap did not change during the welding process, and the sensor detection results were highly consistent with the actual state, further ensuring the accuracy of the judgment in the second stage.

[0068] Specifically, Phase 3: Closed-loop feedback and stiffness model update phase: During the welding process, welding parameter fluctuation data and deformation feedback data of the welding fixture are collected in real time; when it is determined that the thermal deformation of a certain tower module is abnormal based on the welding parameter fluctuation data and / or the deformation feedback data, the welding sequence of subsequent welds and the preset deformation compensation amount are dynamically adjusted to update the stiffness model of the welding fixture.

[0069] Determining an abnormal thermal deformation of a certain tower module includes: calculating the actual heat input at the current welding position based on the real-time collected fluctuations in welding current and / or welding voltage; determining an abnormal thermal deformation when the deviation between the actual heat input and the theoretical heat input exceeds a preset heat input threshold, or when the rate of change of the deformation feedback data of the welding fixture exceeds a preset deformation rate threshold.

[0070] Updating the stiffness model of the welding fixture includes: using the correction value of the deformation compensation amount corresponding to the abnormal thermal deformation amount as a feedback amount; using the feedback amount to perform weighted fusion or Kalman filtering to update the preset stiffness parameters of the welding fixture to obtain the updated stiffness model, which is used to determine the preset deformation compensation amount of the next tower module to be welded.

[0071] In this embodiment, the above-mentioned execution parameters are adjusted in real time through a pre-built deformation compensation model; the process also includes: after welding is completed, obtaining the actual forming coordinates of the weld; updating the pre-built deformation compensation model according to the difference between the actual forming coordinates and the preset coordinates, wherein the deformation compensation model is used to determine the correction amount of the welding sequence using the real-time thermal deformation amount.

[0072] After eliminating assembly stress in the modules in Phase 1 and achieving dynamic coordinated control of weld position and welding parameters in Phase 2, the welding process can proceed under relatively ideal conditions. However, welding UHV tower modules is a physical process accompanied by intense heat input and complex thermo-mechanical coupling. Even if assembly stress has been eliminated and welding parameters have been adjusted in real time according to the gap state, fluctuations in the module's own material properties (such as yield strength and coefficient of thermal expansion) along the length direction, changes in ambient temperature, and the cumulative effect of heat can still cause actual thermal deformation to deviate from the preset value. In traditional processes, the welding sequence and deformation compensation amount are usually fixed values ​​set offline, which cannot adapt to such fluctuations on a module-by-module and weld-by-weld basis, resulting in inconsistent forming accuracy of multiple welds. Therefore, the core task of Phase 3 is to introduce a closed-loop feedback mechanism based on Phase 1 and Phase 2, use real-time data collected during the welding process to judge abnormal thermal deformation online, dynamically adjust the welding sequence and preset deformation compensation amount of subsequent welds, and update the stiffness model of the tooling through weighted fusion or Kalman filtering; furthermore, after the welding is completed, the deformation compensation model is continuously optimized by using the difference between the actual forming coordinates and the preset coordinates, so that the process has self-learning and self-adaptive capabilities.

[0073] Please continue reading. Figure 1 and Figure 4 Phase 3 is executed in parallel with Phase 2. That is, data acquisition and analysis for Phase 3 are performed continuously during the welding process in Phase 2, and the model is updated after welding is completed. The specific implementation process is as follows: I. Data Collection When the welding robot performs the welding task in Phase Two, the control unit (the central controller shared with Phase One and Phase Two) collects two types of data in real time at a sampling frequency of no less than 100Hz: (1) Welding parameter fluctuation data: The actual welding current I(t) and welding voltage U(t) are obtained through the feedback interface of the welding power source. Combined with the welding speed v(t) in the Phase 2 command, the actual heat input P_real(t) = U(t)·I(t) / v(t) at the current welding position is calculated. At the same time, the theoretical heat input P_theory (i.e., the preset value in the central state of Phase 2, such as the heat input corresponding to 280A and 0.8m / min) is recorded.

[0074] (2) Deformation feedback data of welding fixture: High-precision displacement sensors (such as laser displacement gauges) or strain gauges are pre-installed at key stress points of the fixture base (such as the connection between the clamping mechanism and the base, and the middle of the support beam with a large span) to measure the elastic deformation D(t) and its rate of change dD / dt of the fixture at each welding moment in real time. These sensors directly reflect the force and deformation transmitted to the fixture by the module under the action of welding heat.

[0075] II. Judgment of Abnormal Thermal Deformation The central controller analyzes the above data in real time and determines whether the thermal deformation of the current tower module (or the current weld section) is abnormal according to the following rules: Condition A (Abnormal Thermal Input Deviation): Calculate the relative deviation δ = |P_real-P_theory| / P_theory between the actual thermal input P_real and the theoretical thermal input P_theory. If δ exceeds a preset thermal input threshold (e.g., 15%) and lasts for more than 1 second, the thermal deformation is considered abnormal. This condition reflects the deviation between the material's thermal response characteristics (such as thermal conductivity and specific heat capacity) and the preset model, which may be caused by batch fluctuations in the material or local defects leading to the accumulation of abnormal thermal input.

[0076] Condition B (Abnormal Deformation Rate): Monitor the rate of change of tooling deformation feedback data dD / dt. If dD / dt exceeds the preset deformation rate threshold (e.g., 0.2 mm / s), the thermal deformation is deemed abnormal. This condition reflects that the module is undergoing rapid thermal deformation, which may be due to improper welding sequence or sudden changes in heat dissipation conditions leading to heat concentration.

[0077] The two conditions above are in an "OR" logical relationship: either condition being met triggers the determination of abnormal thermal deformation. The specific value of the threshold can be determined through preliminary calibration tests based on the module size, material, and tooling stiffness, and is pre-stored in the control unit.

[0078] III. Dynamic Adjustment: Welding Sequence and Preset Deformation Compensation Amount Once an abnormal thermal deformation is detected, the central controller immediately performs the following two dynamic adjustments, which apply to the preset parameters of subsequent welds of the current module that have not yet been welded, as well as the next module to be welded in the same batch: (1) Dynamically adjust the welding sequence of subsequent welds: The central controller re-plans the welding sequence of the remaining welds based on the type and severity of thermal deformation anomalies. Strategies include: Skip welding: If the abnormality is manifested as localized high temperature, skip the adjacent high temperature zone and weld the far end weld first, giving the heat-affected zone sufficient cooling time (e.g., cool for 5 minutes before returning to welding).

[0079] Symmetrical welding: If the abnormality is manifested as overall bending deformation of the module, switch to a symmetrical welding sequence (e.g., weld the left side first, then the right side, alternating) to counteract asymmetrical thermal shrinkage.

[0080] Segmented welding: The long weld is divided into multiple short segments, and the heat input is distributed by using "step welding" or "segmented skip welding".

[0081] (2) Dynamically adjust the preset deformation compensation amount: The central controller calculates the correction value of the deformation compensation amount ΔC = K·(ΔD_actual-ΔD_target) based on the deviation between the actual deformation amount ΔD_actual observed in this thermal deformation anomaly and the expected deformation amount ΔD_target, where K is a proportional coefficient (which can be learned through historical data). This correction value is used to update the preset deformation compensation amount of subsequent welds in the current module (for example, the reverse deformation amount is increased from 2mm to 2.5mm).

[0082] IV. Update the stiffness model of the welding fixture The stiffness model is an equivalent mathematical model describing the elastic deformation of tooling, modules, and welding actuators under stress. Traditional tooling uses fixed stiffness parameters, which cannot adapt to stiffness drift caused by factors such as fixture wear and module size fluctuations during actual welding. This application achieves online stiffness model updates through a closed-loop feedback in stage three: The central controller uses the correction value ΔC of the deformation compensation corresponding to the above-mentioned abnormal thermal deformation as the feedback value, and updates the preset stiffness parameters of the welding fixture using a weighted fusion or Kalman filter algorithm.

[0083] Weighted fusion update method: Let the preset stiffness be K0 (i.e., the compensation force required per unit deformation). The equivalent stiffness obtained from this observation is K_obs = F / ΔD_actual (F can be obtained from a force sensor or calculated based on welding parameters). Then the updated stiffness is K_new = α·K_obs + (1-α)·K_old, where α is the learning rate (e.g., 0.3), and K_old is the stiffness value before the update. The weighted fusion method is simple to calculate and suitable for scenarios with high real-time requirements.

[0084] Kalman filtering update method: Stiffness parameters are used as state variables, and observed values ​​are used as measurement inputs. A minimum variance estimate is obtained through prediction-update iteration. Kalman filtering can effectively suppress measurement noise and is suitable for applications with significant sensor noise.

[0085] The updated stiffness model is stored in the control unit to determine the preset deformation compensation amount for the next tower module to be welded, thereby achieving continuous evolution by "learning parameters once a weld is made".

[0086] V. Post-feedback update of the deformation compensation model Beyond the online updates in Phase Three described above, this application further constructs a higher-level deformation compensation model. This model is pre-built based on finite element simulation or historical welding data, taking real-time thermal deformation (e.g., module temperature, expansion, or deformation rate during welding) as input and outputting corrections to the welding sequence (e.g., weld numbers to be skipped, waiting time, or welding direction adjustments). During the welding execution in Phase Two, the central controller can use this deformation compensation model to determine the welding sequence corrections in real time, achieving more intelligent welding sequence planning.

[0087] More importantly, after each weld is completed (or the entire module is welded), this application acquires the actual forming coordinates of the weld (e.g., the three-dimensional coordinates of the weld toe centerline) using a visual measurement device (such as an industrial camera or laser tracker). The central controller compares the actual forming coordinates with the preset coordinates (theoretical coordinates in the design drawings) and calculates the difference ΔE = |actual coordinates - preset coordinates|. This difference reflects the overall accuracy deviation of the welding process and also implies the prediction error of the deformation compensation model.

[0088] The central controller updates the pre-built deformation compensation model based on the difference ΔE: for example, by using ΔE as a loss signal and adjusting the model's internal parameters (such as the weights of a neural network or the slope / intercept of a piecewise linear function) using gradient descent; or by incorporating the correction amount from the current observation into the model's experience base using a weighted average. Through continuous updates across multiple modules and weld seams, the deformation compensation model gradually approximates the true mapping relationship between "thermal deformation amount and welding sequence correction amount," making the adjustment of the welding sequence increasingly precise.

[0089] Phase Three includes online anomaly detection and real-time adjustment: Thermal deformation anomalies are determined using two independent dimensions—actual thermal input deviation and tooling deformation rate—to avoid misjudgment based on a single signal. Phase Three also includes adaptive updating of the stiffness model: The correction value of the deformation compensation is used as feedback, and the tooling stiffness parameters are updated using weighted fusion or Kalman filtering, making the deformation compensation calculations of subsequent modules more accurate. Furthermore, Phase Three includes offline / semi-offline evolution of the deformation compensation model: The difference between the actual post-weld forming coordinates and the preset coordinates is used to continuously optimize the deformation compensation model, allowing the correction amount for the welding sequence to evolve from a fixed empirical value to a data-driven self-learning value. Phase Three, through these three levels, forms a complete closed-loop learning system.

[0090] Phase 1 eliminates assembly stress, enabling the tooling deformation feedback data D(t) to accurately reflect welding thermal deformation rather than assembly stress release, thus improving the accuracy of anomaly detection. Phase 2 achieves position-parameter coordination, ensuring that fluctuations in the actual heat input P_real(t) primarily originate from material properties rather than parameter switching disturbances, making the heat input deviation threshold more statistically significant. The updated stiffness model output in Phase 3 can be fed back to the clamping force preset in Phase 1 (e.g., adjusting the initial clamping force based on module stiffness) and the welding parameter benchmark in Phase 2 (e.g., fine-tuning the heat input in the center state based on thermal deformation trends), forming a complete closed loop of "stress-free clamping → adaptive welding → self-learning compensation".

[0091] Continuing the examples from Phases 1 and 2: In the aforementioned tower leg module welding, after stress-free positioning in Phase 1 and adaptive welding in Phase 2, the first three weld seams were of good quality. When welding the fourth weld seam, the central controller monitored that: the welding current and voltage were stable, but the actual heat input P_real was 18% higher than the theoretical value (δ=0.18>0.15), lasting for 2.5 seconds; at the same time, the tooling deformation rate dD / dt reached 0.25mm / s (threshold 0.2mm / s). Both conditions A and B were triggered, and the central controller determined that the thermal deformation in this section was abnormal.

[0092] Subsequently, the central controller performs dynamic adjustments: Welding sequence adjustment: The fifth weld, which was originally planned to be welded continuously, was delayed and replaced by welding the small connecting weld at the far end first, allowing the current section to cool naturally for 5 minutes.

[0093] Deformation compensation adjustment: Based on the fact that the actual deformation is 0.6mm more than expected, the correction value ΔC = 0.8 × 0.6 = 0.48mm is calculated, and the preset reverse deformation amount of the subsequent fifth weld is increased from 2.0mm to 2.48mm.

[0094] Simultaneously, the central controller uses the observed data to update the tooling stiffness model through Kalman filtering: the original preset stiffness K0 = 2000 N / mm, the equivalent stiffness observed this time K_obs = 2150 N / mm, and after filtering, it is updated to K_new = 2080 N / mm. This updated stiffness model is stored and applied to the welding preset values ​​of the next tower leg module.

[0095] After the fourth weld was completed, the vision measurement device acquired the actual forming coordinates of the weld and found a difference of 0.9 mm from the preset coordinates (slightly larger than the expected 0.5 mm). The central controller used this difference ΔE = 0.4 mm (0.9-0.5) to update the deformation compensation model: adjusting the weight parameters related to the amount of thermal deformation in the model so that the output of the welding sequence correction is more accurate when encountering similar amounts of thermal deformation in the future. After several iterations of modules, the prediction error of the deformation compensation model gradually converged from the initial 0.8 mm to within 0.2 mm.

[0096] After all welds were completed, a coordinate measuring machine (CMM) measurement was performed on the tower leg module. The results showed that the flatness error of the four corner points was ±0.8mm, which is far better than the industry standard (±2mm), and the accuracy was improved by 68% compared with the comparison module without Phase 3 (error ±2.5mm). This demonstrates that Phase 1 provides a stress-free initial state, Phase 2 enables real-time adaptive adjustment of welding parameters, and Phase 3, through the judgment of thermal deformation anomalies, stiffness model updates, and feedback from the deformation compensation model, gives the process the ability to learn and evolve across multiple time scales. These three aspects are interconnected and together achieve high precision and high consistency in the welding of UHV tower modules.

[0097] This application embodiment further provides a modular assembly welding fixture system for ultra-high voltage transmission towers, the system being used to perform the modular assembly welding fixture process, the system comprising: The welding fixture body includes a fixture base and a plurality of clamping mechanisms disposed on the fixture base, the clamping mechanisms being used to fix the iron tower module to be welded; The drive unit, integrated on the clamping mechanism, is used to apply macroscopic disturbances to the tower module; The detection unit includes a vision sensor, a force sensor and / or a torque sensor, used to detect the weld condition, reaction force or torque signal and tooling deformation feedback data in real time. Welding execution unit, including welding robot and welding power source; The control unit is communicatively connected to the drive unit, the detection unit, the welding robot, and the welding power supply, and is configured to execute the modular assembly welding fixture process.

[0098] This specification also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method shown in any of the above embodiments.

[0099] Computer-readable media, including both permanent and non-permanent, removable and non-removable media, can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, disk storage, quantum memory, graphene-based storage media or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0100] The processing and logic described in this specification can be executed by one or more programmable computers that execute one or more computer programs to perform corresponding functions by operating on input data and generating output.

[0101] Computers suitable for executing computer programs include, for example, general-purpose and / or special-purpose microprocessors, or any other type of central processing unit. Typically, the central processing unit receives instructions and data from read-only memory and / or random access memory. The basic components of a computer include a central processing unit for implementing or executing instructions and one or more memory devices for storing instructions and data. Typically, a computer will also include one or more mass storage devices for storing data, such as magnetic disks, magneto-optical disks, or optical disks, or the computer will be operatively coupled to such mass storage devices to receive data from or transfer data to them, or both.

[0102] Computer-readable media suitable for storing computer program instructions and data include all forms of non-volatile memory, media, and memory devices, such as semiconductor memory devices (e.g., EPROM, EEPROM, and flash memory devices), magnetic disks (e.g., internal hard disks or removable disks), magneto-optical disks, and CD-ROM and DVD-ROM disks. Processors and memory may be supplemented by or incorporated into dedicated logic circuitry.

[0103] While this specification contains numerous specific implementation details, these should not be construed as limiting the scope of any invention or the scope of the claims, but rather are primarily intended to describe features of specific embodiments of a particular invention. Certain features described in the various embodiments herein may also be implemented in combination in a single embodiment. Conversely, various features described in a single embodiment may also be implemented separately in various embodiments or in any suitable sub-combination. Furthermore, while features may function in certain combinations as described above and even initially claimed in this way, one or more features from a claimed combination may be removed from that combination in some cases, and a claimed combination may refer to a sub-combination or a variation thereof.

[0104] Similarly, although the operations are depicted in a specific order in the accompanying drawings, this should not be construed as requiring these operations to be performed in the specific order shown or sequentially, or requiring all illustrated operations to be performed to achieve the desired result. In some cases, multitasking and parallel processing may be advantageous. Furthermore, the separation of various system modules and components in the above embodiments should not be construed as requiring such separation in all embodiments, and it should be understood that the described program components and systems can generally be integrated together in a single software product or packaged into multiple software products.

[0105] Thus, specific embodiments of the subject matter have been described. Other embodiments are within the scope of the appended claims. In some cases, the actions recited in the claims may be performed in a different order and still achieve the desired result. Furthermore, the processes depicted in the drawings are not necessarily shown in a specific order or sequence to achieve the desired result. In some implementations, multitasking and parallel processing may be advantageous.

[0106] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A modular assembly welding fixture process for ultra-high voltage transmission towers, characterized in that, The process includes the following three stages: Phase 1: Macroscopic Disturbance Positioning and Assembly Stress Elimination Phase: After hoisting the first and second tower modules onto the welding fixture, the clamping mechanism of the welding fixture is controlled to apply macroscopic disturbances to the first and second tower modules. During the application of the macroscopic disturbances, the reaction force or torque signal on the clamping mechanism is monitored in real time. When it is determined that there is an asymmetry in assembly stress based on the reaction force or torque signal, the position or clamping force of the clamping mechanism is adjusted until the assembly stress is eliminated. Phase Two: Position-Parameter Coordinated Control Phase During Welding: The welding process is initiated, and sensors continuously monitor the weld's condition. When a large gap or misalignment is detected, it is determined to be a defective weld edge, automatically reducing the welding heat input and increasing the wire feed. When an ideal butt joint is detected, it is determined to be a weld center state, restoring high-efficiency welding parameters. The welding robot's trajectory and the welding power supply's parameters are dynamically adjusted synchronously based on the weld's condition. Phase 3: Closed-loop feedback and stiffness model update phase: During the welding process, welding parameter fluctuation data and deformation feedback data of the welding fixture are collected in real time; when it is determined that the thermal deformation of a certain tower module is abnormal based on the welding parameter fluctuation data and / or the deformation feedback data, the welding sequence of subsequent welds and the preset deformation compensation amount are dynamically adjusted to update the stiffness model of the welding fixture.

2. The modular assembly welding fixture process according to claim 1, characterized in that, The determination that there is an asymmetric assembly stress includes: Acquire the reaction force or torque signals of the clamping mechanism in both positive and negative directions of the macroscopic disturbance; Calculate the degree of asymmetry of the reaction force or torque signals in the two opposite directions; When the degree of asymmetry exceeds a preset threshold, it is determined that there is an asymmetry in assembly stress. Adjusting the position or clamping force of the clamping mechanism includes: determining the interference direction based on the sign of the degree of asymmetry, and adjusting the actual position of the clamping mechanism or the tower module in a direction opposite to the interference direction.

3. The modular assembly welding fixture process according to claim 1, characterized in that, The reduction of welding heat input includes: reducing welding speed, reducing welding current and / or reducing welding voltage; the increase of wire feed includes: increasing wire feed speed; the high-efficiency welding parameters include: higher than the reduced welding heat input and lower than the increased wire feed.

4. The modular assembly welding fixture process according to claim 1, characterized in that, The trajectory movement of the welding robot and the welding parameters of the welding power source are dynamically adjusted synchronously according to the weld condition, including: A central controller is provided, which is communicatively connected to the motion controller of the welding robot and the power controller of the welding power source; The central controller synchronously generates trajectory control commands and welding parameter control commands based on the weld seam status detected by the sensors. The motion controller controls the movement of the end effector of the welding robot according to the trajectory control command, and the power controller controls the output parameters of the welding power supply according to the welding parameter control command.

5. The modular assembly welding fixture process according to claim 1, characterized in that, The determination of abnormal thermal deformation of a certain tower module includes: The actual heat input at the current welding position is calculated based on the real-time collected fluctuations in welding current and / or welding voltage. When the deviation between the actual heat input and the theoretical heat input exceeds a preset heat input threshold, or when the rate of change of the deformation feedback data of the welding fixture exceeds a preset deformation rate threshold, the amount of thermal deformation is determined to be abnormal.

6. The modular assembly welding fixture process according to claim 1, characterized in that, The updating of the stiffness model of the welding fixture includes: The correction value of the deformation compensation amount corresponding to the abnormal thermal deformation amount is used as the feedback amount; The preset stiffness parameters of the welding fixture are updated by weighted fusion or Kalman filtering using the feedback quantity to obtain an updated stiffness model. The updated stiffness model is used to determine the preset deformation compensation amount of the next tower module to be welded.

7. The modular assembly welding fixture process according to claim 1, characterized in that, The macroscopic disturbance is a small oscillation or vibration applied by the clamping mechanism to the tower module in a specific direction. The amplitude of the small oscillation or vibration is within the elastic deformation range of the tower module, and the scale of the macroscopic disturbance is on the order of meters.

8. The modular assembly welding fixture process according to claim 1, characterized in that, The sensors include a vision sensor and a force sensor; the large gap or misalignment state includes a weld gap width greater than a preset width threshold, or a misalignment amount between two tower modules greater than a preset misalignment threshold.

9. The modular assembly welding fixture process according to claim 1, characterized in that, The reduction of welding heat input and increase of wire feed include: switching to cold metal transition welding mode; the restoration of high-efficiency welding parameters includes: switching to conventional pulse welding mode.

10. The modular assembly welding fixture process according to claim 1, characterized in that, In the second stage, the welding parameters also include oscillation parameters, which include oscillation amplitude, oscillation frequency, and oscillation trajectory. When the weld is determined to be in an edge defect state, the oscillation amplitude is reduced; when the weld is determined to be in a center state, the preset oscillation amplitude is restored.

Images