An intelligent welding control method for an automated pipe welding workstation

By optimizing the selection of support coordinate points and scanning the weld seam with a laser sensor, a continuous trajectory segment is constructed and execution deviations are eliminated. This solves the problems of discontinuous trajectory connection and inconsistent process execution in the welding process in the existing technology, and improves the stability and collaborative accuracy of the welding process.

CN123103237APending Publication Date: 2026-11-17JIANGSU HONGSHI INTELLIGENT TECHNOLOGY CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611479165.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-09-22
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

The existing control methods for automated pipeline welding workstations lack a real-time status verification mechanism, resulting in discontinuous trajectory connections and inconsistent process execution during the welding process. In particular, it is difficult to correct in a timely manner when there are differences in the support contact state, which affects the stability of the posture and the consistency of the equipment's actions.

Method used

By acquiring support coordinate points, filtering and optimizing support layout, combining laser sensor scanning of weld seams, constructing continuous trajectory segments, and eliminating execution deviations through multi-device status comparison, the timing consistency of welding torch movement and wire feeding behavior is achieved.

Benefits of technology

It improves the continuity of the welding process and the reliability of trajectory execution, reduces the risk of attitude deviation caused by uneven support, and ensures the stability and collaborative accuracy of the welding process.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123103237A_ABST
    Figure CN123103237A_ABST
Patent Text Reader

Abstract

This invention relates to the field of intelligent welding control technology, specifically to an intelligent welding control method for an automatic pipeline welding workstation. The method involves generating a support arrangement set by selecting coordinates based on pipe fitting parameters, evaluating the contact fit state, and optimizing the support execution sequence. After positioning, a continuous weld trajectory is spliced ​​out by laser scanning and angle comparison. A process path is generated by extracting the welding torch posture and wire feeding action based on the trajectory. The operating states of multiple devices are matched, and inconsistent sections are eliminated, thereby continuously generating an intelligent welding control result sequence. This invention optimizes the fit relationship by selecting support coordinates and determining contact differences through multiple rounds, thereby reducing the risk of posture deviation. A continuous trajectory is constructed based on the consistency of weld point direction. Combined with posture change constraint action matching, the welding torch and wire feeding timing are made consistent. Cross-comparison of multiple device states eliminates execution deviations, improving collaborative accuracy and connection stability, and enhancing welding continuity and trajectory reliability.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of intelligent welding control technology, and in particular to an intelligent welding control method for an automatic pipeline welding workstation. Background Technology

[0002] The field of intelligent welding control technology encompasses technologies related to automated control and intelligent adjustment in industrial welding processes. It involves attitude control, trajectory control, and process parameter control of pipes, fittings, and welded structures during the welding process. It mainly includes pipe assembly positioning control, weld seam identification and tracking control, welding current, voltage, and travel speed control, as well as multi-state information acquisition and collaborative control during the welding process. This technology is widely used in pipe prefabrication, pressure pipe manufacturing, and industrial equipment welding operations. Its core lies in achieving continuous and stable operation of the welding process through the linkage control between the spatial position relationship of the welding and the process parameters.

[0003] One of the intelligent welding control methods for an automatic pipeline welding workstation refers to a control method used in an automatic pipeline welding workstation. This method covers pipe fitting assembly positioning control, weld position identification control, welding process parameter matching control, and welding execution path control. Specifically, it uses pipe fitting size information input and support position adjustment for pipe fitting assembly positioning control, uses bevel edge image acquisition and weld feature extraction for weld position identification control, uses the correspondence between TIG root pass welding parameters and MIG cap pass welding parameters for welding process parameter matching control, and uses welding torch motion trajectory coordinate generation and execution command output for welding execution path control.

[0004] Existing control methods focus on pre-setting support positions and fixed matching relationships, lacking continuous identification of differences in support contact states. When the pipe is subjected to changes in force or there are slight differences in the outer diameter, local instability in fit can easily occur. For example, when multiple support points are bearing loads simultaneously, local suspension phenomena are difficult to detect in time, thus affecting the overall posture stability. At the same time, the consistency of equipment actions during welding depends on preset logic and lacks a verification mechanism based on real-time operating status. When there are deviations in the welding torch movement rhythm, wire feeding behavior, or displacement movement, it is difficult to correct them in time, which can easily lead to discontinuous trajectory connection and inconsistent process execution. Summary of the Invention

[0005] To address the technical problems existing in the prior art, this invention provides an intelligent welding control method for an automated pipeline welding workstation. The technical solution is as follows: An intelligent welding control method for an automated pipeline welding workstation includes the following steps: S1: Obtain the straight pipe length, end clearance distance and roller frame X-axis movable position record, subtract the end clearance distance to obtain the effective support range, filter the position item by item, check the joint clamping spacing, match and filter the Z-axis target height, combine the support coordinate points, and generate the support layout coordinate set. S2: Based on the set of coordinates of the support arrangement, select the combination that meets the quantity in the X direction, write the X-direction position value and the Z-direction target height, read the contact point and the fitting state, compare the adjacent contact positions, mark the fitting change section, replace the adjacent support coordinate points one by one, sort by quantity, and generate the support execution position sequence; S3: Based on the support execution position sequence, drive the roller support assembly to the position. After the pipe support is positioned, the laser sensor continuously scans the annular weld, records the spatial position of the weld, generates direction vectors according to adjacent points and compares the included angles, retains the continuous connection relationship and splices the local trajectory segments to generate the continuous running trajectory of the weld. S4: Based on the continuous operation trajectory of the weld, read the local trajectory segment and splicing sequence, obtain the welding gun movement record, wire feeding action record and welding gun posture record corresponding to the weld spatial point along the splicing sequence, mark the posture change segment, limit the matching range and continue the corresponding action to generate the welding operation process path. S5: Based on the welding operation process path, retrieve the status of the welding robot, C-type positioner and wire feeding mechanism, match the welding torch movement position and positioner rotation position according to the same posture change segment, compare the wire feeding action status and movement sequence, eliminate inconsistent execution segments, and generate intelligent welding control result sequence.

[0006] The beneficial effects of the technical solutions provided in the embodiments of the present invention include at least the following: In this invention, by performing multiple rounds of screening on the support coordinates and introducing contact state difference judgment, the dynamic optimization of the fit between the support arrangement and the outer surface of the pipe is achieved, reducing the risk of attitude deviation caused by uneven support. At the same time, a continuous trajectory segment is constructed based on the consistency of the spatial position direction of the weld, and the action matching range is constrained by the attitude change segment, so that the movement of the welding gun and the wire feeding behavior are kept in chronological order. Then, by cross-comparing the operating status of multiple devices, the execution deviation segment is eliminated, improving the coordination accuracy and action connection stability of each execution link, thereby strengthening the continuity of the welding process and the reliability of trajectory execution. Attached Figure Description

[0007] Figure 1 This is a flowchart of the method of the present invention; Figure 2 This is a flowchart illustrating the acquisition process of S1 in this invention; Figure 3 This is a flowchart illustrating the acquisition process of S2 in this invention; Figure 4 This is a flowchart illustrating the acquisition process of S3 in this invention; Figure 5 This is a flowchart illustrating the acquisition process of S4 in this invention; Figure 6 This is a flowchart of the acquisition process for S5 of the present invention. Detailed Implementation

[0008] The following detailed explanation illustrates the specific implementation methods: The basic implementation examples are as follows: Figure 1 The following is illustrated: An intelligent welding control method for an automatic pipeline welding workstation, comprising the following steps: S1: Obtain the straight pipe length, end clearance distance, and X-axis movable position record of the roller frame. Subtract the end clearance distance from the straight pipe length to obtain the effective support range. Substitute each X-axis movable position into the effective support range for filtering to form a position correspondence sequence. Obtain the joint clamping spacing record. Compare adjacent positions in the position correspondence sequence in turn and eliminate positions that do not meet the joint clamping spacing to form a spacing qualified position sequence. Obtain the straight pipe outer diameter and Z-axis height correspondence record. Match the Z-axis target height item by item according to the spacing qualified position sequence to form a height correspondence sequence. Substitute the height correspondence sequence into the height adjustment range for filtering to eliminate positions that exceed the height adjustment range to form a height qualified position sequence. Combine the X-axis position value in the height qualified position sequence with the corresponding Z-axis target height to form a support coordinate point and generate a support layout coordinate set. Please see Figure 2 Specifically, S1 includes: S11: Obtain the straight pipe length, end clearance distance, and X-axis movable position record of the roller frame. Subtract the end clearance distances from the straight pipe length to define the effective support interval. Substitute the X-axis movable position records of the roller frame into the effective support interval filter, retaining the positions within each interval to generate a position-corresponding sequence; details are as follows: The process involves acquiring the straight pipe length, end clearance distance, and X-axis movable position of the roller frame. An industrial-grade laser scanning rangefinder is used to physically scan the straight pipe to be processed, and the geometric measurement data reveals a straight pipe length of 12000 mm. The welding process parameter database in the MAS system is retrieved, showing an end clearance distance of 500 mm. The MAS system, a manufacturing automation management system, stores or receives workpiece information, pipe specifications, support parameters, and welding process parameters, and transmits the corresponding parameters to the workstation control system. An absolute photoelectric encoder mounted on the bottom guide rail is used to collect real-time feedback on the frame's movement position. This ensures that at least three roller support assemblies are installed along the straight pipe length direction at the workstation. Each roller support assembly is equipped with an X-axis walking drive mechanism and a Z-axis lifting drive mechanism, and can independently adjust according to the corresponding support coordinates. For ease of illustration, the first, second, and third roller support components in their working state are extracted here for specific deduction. These at least three roller support components can move independently along the X-axis guide rail and rise and fall independently along the Z-axis. This yields five initial candidate X-axis movable positions along a continuously adjustable guide rail for position combination selection of the above components: 300mm, 2000mm, 6000mm, 10000mm, and 11800mm. The guide rail range supports subsequent dynamic generation of other continuous candidate positions based on translation step length calculations. Subtracting the sum of the end clearance distances of 1000mm (the sum of the first and last end clearance distances) from the straight pipe length of 12000mm, the left boundary value of the effective support interval is calculated to be 500mm and the right boundary value to be 11500mm. The effective support interval is thus constructed as [500, ...].

[11500] , the physical significance of this interval is to avoid mechanical interference between the roller frame and the heat-affected zone of the welded pipe joints at both ends. Each initial candidate movable position in the X-direction is extracted sequentially and compared with the boundary value of the effective support interval. The first movable position in the X-direction (300mm) is determined to be less than the left boundary value of the effective support interval (500mm), indicating that the position is within the interference zone. It is determined to be an outside position and is eliminated. The second movable position in the X-direction (2000mm) is determined to be greater than or equal to 500mm and less than or equal to 11500mm, indicating that it is within the interval and is retained. The third movable position in the X-direction (6000mm) is determined to be greater than or equal to 500mm and... If the value is less than or equal to 11500mm, it is considered within the range and retained. If the fourth movable X-axis position is 10000mm, which is greater than or equal to 500mm and less than or equal to 11500mm, it is also considered within the range and retained. If the fifth movable X-axis position is 11800mm, which is greater than the right boundary value of the effective support range of 11500mm, it means that the position exceeds the effective support range of the pipe end and is considered outside the range, thus being rejected. The retained initial movable X-axis positions of 2000mm, 6000mm, and 10000mm within the safe support range are arranged in ascending order of value in a one-dimensional array to generate a position-corresponding sequence.

[0009] S12: Call the position correspondence sequence to obtain the joint clamping spacing record, compare adjacent positions in turn, remove positions that do not meet the joint clamping spacing, form a spacing qualified position sequence, obtain the straight pipe outer diameter and Z-direction height correspondence record, match the Z-direction target height item by item according to the spacing qualified position sequence, and generate a height correspondence sequence; specifically as follows: Call the position correspondence sequence to obtain the pipe fitting pre-assembly process data corresponding to the current workpiece identifier in the MAS system, read the minimum allowable arrangement spacing record of adjacent roller support components from the pipe fitting pre-assembly process data, which is greater than or equal to 1500mm. The minimum allowable arrangement spacing is used to avoid mechanical interference between adjacent roller support components and to avoid excessive concentration of multiple support components, which would lead to local load concentration; The straight pipe's self-weight deflection is limited by the maximum allowable support span determined subsequently based on the pipe fitting specifications. From the corresponding position sequence, adjacent positions of 2000mm and 6000mm are extracted sequentially and subtracted to obtain the first set of adjacent spacing values ​​of 4000mm. This 4000mm value is compared with the minimum allowable spacing of 1500mm. If 4000mm is greater than or equal to 1500mm, the minimum allowable spacing requirement is met, and positions 2000mm and 6000mm are retained. The next set of adjacent positions, 6000mm and 10000mm, is then extracted and subtracted to obtain the second set of adjacent spacing values ​​of 4000mm. A comparison is made between 0mm and the minimum allowable spacing of 1500mm. If 4000mm is greater than or equal to 1500mm, the minimum allowable spacing requirement is met, and positions 6000mm and 10000mm are reserved. All reserved positions are summarized to generate a qualified spacing position sequence [2000, 6000, 10000]. The pipe fitting specification table is retrieved to obtain the straight pipe outer diameter of 800mm and its corresponding Z-axis height. From the straight pipe outer diameter and Z-axis height correspondence records, the standard centroidal axis height corresponding to the straight pipe outer diameter of 800mm is found to be 600mm. This is based on a formula obtained by combining the roller assembly's own structural dimensions and the pre-calibrated mechanical zero-point deviation. Calculate the target height in the Z direction, in the formula This represents the absolute height that the roller support reference surface needs to be adjusted relative to the workstation mounting reference surface, and its dimension is millimeters. This represents the standard center of gravity height value to ensure the pipe fitting is level, and is taken as 600mm. This represents the outer diameter of the straight pipe, taken as 800 mm. Substituting the values ​​into the calculation yields... With a target height of 200mm, this result represents the absolute lifting height of the rollers to keep the pipe axis on the 600mm baseline, for each of the qualified position sequences [2000, 6000, 10000]. To the position and the calculated Combine each item with the target height value of 200mm to generate a height-corresponding sequence.

[0010] S13: Based on the height correspondence sequence, call the height adjustment range, sequentially substitute the Z-axis target height into the height adjustment range for filtering, eliminate positions exceeding the height adjustment range, and form a sequence of qualified height positions. Combine the X-axis position values ​​in the qualified height position sequence with the corresponding Z-axis target height to form support coordinate points, generating a support layout coordinate set; as detailed below: Based on the height correspondence sequence, the electromechanical hardware parameter table of the roller frame equipment is retrieved to obtain the height adjustment range of its Z-axis lead screw, which is 150mm to 450mm. The Z-axis target height value of 200mm is extracted sequentially from the height correspondence sequence for each position. This 200mm is compared with the physical lower limit of the height adjustment range (150mm) and the upper limit (450mm). If 200mm is greater than or equal to 150mm and less than or equal to 450mm, the Z-axis target height is determined to be within the mechanical stroke height adjustment range of the equipment, and a hold action is executed. If the calculated Z-axis target height value for a certain position is 100mm, it is determined that 100mm is less than the lower limit of 150mm, meaning that the height requirement is lower than the minimum bottoming limit of the mechanical structure. Therefore, it is determined to be outside the height adjustment range, and an error is reported. And the rejection action: if there is another position where the calculated Z-direction target height value is 500mm, it is determined that 500mm is greater than the upper limit value of 450mm, which means that the height requirement exceeds the maximum lifting stroke of the lead screw. It is determined that the height adjustment range is exceeded and the rejection action is performed. All positions that pass the mechanical limit screening are summarized to form a height qualified position sequence. The X-direction position values ​​[2000, 6000, 10000] retained in the height qualified position sequence are mapped and merged with their corresponding physical feasible Z-direction target height value of 200mm in two-dimensional coordinates in space, and combined into the first support coordinate point (2000, 200mm), the second support coordinate point (6000, 200mm) and the third support coordinate point (10000, 200mm), generating the initial support layout coordinate set. S2: Based on the support coordinate points of the support layout coordinate set, select the support coordinate points that meet the support quantity requirements in ascending order of X-direction position value as the current support position combination. Write the X-direction position value and Z-direction target height into the roller support assembly. When the roller support assembly is in place, read the distribution state of its contact points with the outer surface of the pipe, determine the contact point position as the contact position, and obtain the corresponding fitting state. Compare the fitting state of adjacent contact positions. Mark the fitting change section where the fitting state is inconsistent or the fitting difference exceeds the preset threshold, and count the number. Extract the adjacent support coordinate points before and after each support coordinate point in the current combination in ascending order of X-direction position value and replace them one by one. After each replacement, rewrite the corresponding X-direction position value and Z-direction target height, and recount the number of fitting change sections. Sort by the number of fitting change sections in ascending order, select the support position combination corresponding to the first position in the sort, and generate the support execution position sequence. Please see Figure 3Specifically, S2 includes: S21: Based on the support coordinate points of the support layout coordinate set, read the support coordinate points in ascending order of X-axis position values, compare the reading order with the support quantity requirements item by item, select the positions that meet the support quantity requirements, retain the corresponding coordinate point arrangement, record the support coordinate point sequence number, write the X-axis position value and Z-axis target height into the roller support assembly, and generate the current support position combination after the roller support assembly is in place; as follows: Based on the support coordinate points of the initial support layout coordinate set, the X-axis position values ​​corresponding to the first support coordinate point (2000, 200mm), the second support coordinate point (6000, 200mm), and the third support coordinate point (10000, 200mm) in the initial support layout coordinate set are read. These values ​​are then extracted as 2000mm, 6000mm, and 10000mm, respectively. These three values ​​are sorted in ascending order from smallest to largest, resulting in the following read order: first item (2000, 200mm), second item (6000, 200mm), and third item (10000, 200mm). The pipe support parameter table in the MAS system is retrieved. Based on the straight pipe length, outer diameter, wall thickness, material elastic modulus, and allowable deflection, the maximum allowable support span is determined according to the preset support span calculation rules. The required support quantity value corresponding to the current pipe specification is read as 3 based on the effective support interval length. The current cumulative value of the read items is then compared with the required support quantity value of 3. The system performs a row-by-row comparison to determine the size of each item. When the first item is read, the cumulative item count is 1, so it is determined that 1 is less than 3. When the second item is read, the cumulative item count is 2, so it is determined that 2 is less than 3. When the third item is read, the cumulative item count is 3, so it is determined that 3 equals 3. The reading action is stopped, and the currently read items are truncated. The spatial arrangement of the three coordinate points is preserved. The first support coordinate point is marked and recorded as 1, the second support coordinate point as 2, and the third support coordinate point as 3. The coordinate parameters are written into the PLC position register of the corresponding roller support component via the industrial Ethernet control bus. The main control system performs a position closed-loop monitoring action. The real-time X-axis position and Z-axis height feedback data of each roller support component are continuously read through a high-precision absolute encoder. The limit switches set at the end of the stroke monitor whether each drive mechanism exceeds the allowable stroke. When the absolute value of the difference between the position feedback data and the target coordinate value is less than the control dead zone of 0.1mm, the roller support component is determined to be physically in place, and the current support position combination is generated.

[0011] S22: Call the current support position combination, read the contact point distribution between the roller support assembly and the outer surface of the pipe, determine the contact point as the contact position, obtain the corresponding bonding state of the contact position, compare adjacent bonding states according to the contact position arrangement order, mark the bonding change segment where the bonding state is inconsistent or the bonding difference exceeds the bonding difference threshold, retain the segment order, count the number of marked segments, and obtain the number of bonding change segments; as detailed below: The current support position combination is invoked, and the actual contact state between the roller and the outer surface of the 800mm diameter straight pipe is read using an array of thin-film piezoresistive pressure sensors with multiple numbered sensing units installed on the metal wheel surface of each roller support assembly and an LVDT displacement sensor on the base side. Based on the sensing unit number that generates the effective pressure signal and the current X and Z positions of the roller assembly, the corresponding calibration coordinates are converted to the contact positions in the workstation coordinate system. The pressure fit state corresponding to each contact position is obtained. The fit state is represented by the logical Boolean values ​​1 and 0 after A / D conversion by the pressure sensor, where 1 represents a tight contact state where the contact stress detected by the corresponding pressure sensing unit is not less than a preset fit stress threshold of 5MPa, and 0 represents a gap separation state where the detected contact stress is less than 5MPa. The allowable fit difference threshold is obtained from the process parameter table as 0. The fit state values ​​of the first contact position and the adjacent second contact position are extracted and subtracted along the arrangement of contact positions from small to large values ​​in the X direction. If the fit state value of the first contact position is 1... The second contact position has a fit state value of 1. The absolute value of the difference between the two is 0. The 0 is compared with the fit difference threshold 0. If 0 equals 0, the fit state is consistent and the difference does not exceed the threshold. This means that the pipe has good cylindricity and uniform stress in this section, so it is not marked. The fit state value 1 of the second contact position is extracted and subtracted from the fit state value 0 of the adjacent third contact position. The absolute value of the difference between the two is 1. The 1 is compared with the fit difference threshold 0. If 1 is greater than 0, the fit state is inconsistent and the fit difference exceeds the fit difference threshold. This means that the pipe has a local ellipticity deviation or the roller is not fully lifted in this section. The first fit change section is marked at this abnormal node. The order of the section occurrence is preserved. The accumulator value is changed from 0 to 1. The comparison of subsequent adjacent contact positions continues until the traversal is completed. For example, fit state changes are formed between the second and third contact positions and between the fifth and sixth contact positions. The number of all marked stress uneven sections is counted. The initial number of fit change sections is 2.

[0012] S23: Based on the number of fitting change segments, call the current support position combination, extract the adjacent support coordinate points before and after each support coordinate point in the current combination according to the X-direction position value from smallest to largest, and replace them one by one. After each replacement, rewrite the X-direction position value and Z-direction target height, recount the number of fitting change segments, sort them according to the number of fitting change segments from smallest to largest, select the support position combination corresponding to the first sorted segment, and generate the support execution position sequence; as detailed below: Based on the number of bonding change sections, the current support position combination is called. The X-axis position values ​​(2000mm, 6000mm, 10000mm) of each support coordinate point are extracted from the current combination and arranged in ascending order. Based on the X-axis position of each roller support component, a support influence range is established for each roller support component. Except for the last support influence range, each support influence range adopts a left-closed, right-open form. When the center position of a bonding change section is exactly located at the boundary point of adjacent influence ranges, it is assigned to the influence range corresponding to the roller support component with the larger X-axis position value. The midpoint of the X-axis position of adjacent roller support components is used as the boundary point of the support influence range. For example, the influence range of the first support component can be represented as... The influence range of the second support component can be represented as follows: The influence range of the third support component can be represented as follows: Extract the center position of the bonding change section. When the center position of the bonding change section falls within the support influence range corresponding to a certain roller support component, associate the bonding change section with the corresponding roller support component. Sequentially lock each support coordinate point and extract its adjacent support coordinate points before and after in the initial support arrangement coordinate set. If the center position of the bonding change section corresponding to the second support coordinate point (6000, 200mm) in the current support position combination falls within its influence range [4000, 8000), that is, the support coordinate point is within the marked and associated bonding change section. Extract its previous adjacent point in the set as (2000, 200mm) and its next adjacent point as (10000, 200mm). By adjusting the force... The constant point (6000, 200mm) is translated in the opposite X direction of the continuous guide rail with a step size of 2000mm, generating a translation point with an X position of 4000mm. The X position of 4000mm generated by the translation is then substituted into the effective support range, the minimum spacing between adjacent supports, and the height adjustment range for verification. Since 4000mm is within the effective support range [500, 11500] and the spacing between it and the reserved points 2000mm and 10000mm meets the minimum allowable arrangement spacing condition of greater than or equal to 1500mm, and the corresponding Z-direction height is valid, after verification, it is written into the spare support coordinate set, generating a spare coordinate point (4000, 200mm) that is not occupied by the current combination and is safe and compliant. The system calls upon the safety backup coordinate point (4000, 200mm) in the backup support coordinate set for item-by-item physical replacement. The second support coordinate point (6000, 200mm) with uneven stress is replaced with the backup coordinate point (4000, 200mm) to avoid the pipe deformation zone. After each coordinate replacement, the updated X-axis position value and Z-axis target height are re-sent to the servo drive controller of the corresponding roller support assembly. The drive roller frame moves along the guide rail to the new coordinates and rereads the pressure sensor contact data. The system calculates the absolute value of the contact state difference between adjacent contact positions and compares it with a threshold, re-counting the current global contact change segment count as 1. The system then uses the third support coordinate point (10000, 200mm) to... A translation calculation with a step size of 2000mm is performed in the opposite direction of the X-axis along the continuous guide rail to obtain an X-axis position of 8000mm. It is determined that 8000mm is within the effective support range [500, 11500], and the distance between it and the adjacent point position 6000mm (not replaced) in the original combination is 2000mm. It is determined that 2000mm is greater than or equal to the minimum allowable arrangement spacing of 1500mm. Combined with its corresponding legal Z-axis target height of 200mm within the range of 150mm to 450mm, another safe backup coordinate point (8000, 200mm) is verified and written into the backup support coordinate set. This safe backup coordinate point (8000, 200mm) is extracted to replace the third support coordinate point (10000, ...).(200mm) Generate other backup plans, and re-execute the corresponding movement and pressure status reading. Compare and count again to find that the number of bonding change sections caused by the change in support span of the other backup plan is 3. Collect and summarize the initial statistical value 2 of the initial combination, the re-statistical value 1 after the replacement plan, and the value 3 of the other backup plan. Sort each support position combination in ascending order according to the corresponding bonding change section value. Compare the value of each sorted position and select the support position combination with the smallest bonding abnormality value of 1 at the first position of the sort. Generate the support execution position sequence. S3: Based on the support position combination in the support execution position sequence, extract the X-axis position value and Z-axis target height of each support coordinate point, and drive the roller support assembly to the corresponding support position. After the pipe is positioned, call the laser sensor to perform continuous scanning along the annular weld, and record the weld space points in the order of acquisition during the scanning process. Connect the adjacent weld space points in sequence to obtain the point direction vector. Compare the angle between the previous direction vector and the next direction vector in sequence. When the angle change is within the preset continuous range of direction, retain the connection relationship between the corresponding adjacent weld space points. Connect the weld space points in the order of the retained connection relationship to obtain a local trajectory segment. Splice the local trajectory segments whose endpoint distance meets the preset splicing distance in the order of acquisition to generate the continuous running trajectory of the weld. Please see Figure 4 Specifically, S3 includes: S31: Based on the support position combinations in the support execution position sequence, extract the X-direction position value and Z-direction target height of each support coordinate point, and write them into the roller support assembly according to the arrangement order of the support coordinate points. Drive the roller support assembly to reach the corresponding support position. After the pipe fitting completes the support positioning, generate the support positioning coordinate sequence; specifically as follows: Based on the support position combinations in the support execution position sequence, extract the X-direction position value and Z-direction target height value of each support coordinate point in the optimal force-bearing support position combination corresponding to the support execution position sequence. The first item X-direction position value 2000mm and Z-direction target height value 200mm, the second item X-direction position value 4000mm and Z-direction target height value 200mm, and the third item X-direction position value 10000mm and Z-direction target height value 200mm are extracted respectively, according to the topological arrangement order of the first item, the second item, and the third item. The system writes the position control registers of the first, second, and third roller support assemblies via fieldbus. The main control system sends high-level start signals to the servo drivers of each roller support assembly, driving the X and Z servo motors of the first roller support assembly to move to coordinates (2000, 200mm), the second roller support assembly to coordinates (4000, 200mm), and the third roller support assembly to coordinates (10000, 200mm). It reads the feedback from the absolute encoders of the Z-axis lifting drive shafts of each roller support assembly and calls the array-type thin-film piezoresistive pressure sensors installed inside each roller support assembly to read the pressure response values ​​of each sensing unit. Based on the calibration relationship of each sensing unit and the effective pressure-bearing area, the pressure response values ​​are converted into contact stress. Retrieve the support process parameters from the MAS system to obtain the lower limit of contact support stress. and upper limit of contact support stress The contact support stress detected in each roller support assembly Substitute into the formula Make a judgment when the contact support stress is lower than When it is determined that the roller assembly has not formed a stable support, the contact support stress is higher than the specified value. If a risk of local indentation in the thin-walled straight pipe is detected, and the feedback contact stress value is within the range of [5MPa, 15MPa] and the Z-axis position deviation is within the preset position tolerance range, then the straight pipe fitting with an outer diameter of 800mm and a length of 12000mm is determined to have achieved stable positioning of the straight pipe axial position, center height, and support posture. The actual absolute encoder values ​​of each roller support assembly are read and summarized to generate a support positioning coordinate sequence. S32: The support positioning coordinate sequence is called, and the laser sensor performs continuous scanning along the annular weld, recording the weld space points according to the acquisition sequence. The weld space points are matched with the acquisition sequence number, arranged according to the acquisition sequence number, and the acquisition order of adjacent points is compared to obtain the weld space point sequence; specifically as follows: The system invokes the support positioning coordinate sequence and sends trigger pulses to a line laser profile sensor based on triangulation principles, mounted on the end flange of a six-axis welding robot. This controls the welding robot to carry the laser sensor and perform a continuous spatial circular arc scan along the edge of an 800mm diameter straight pipe bevel weld. Following a fixed 10ms acquisition sequence from the underlying hardware, the sensor's internal DSP digital signal processor performs real-time calculations and invokes the pre-calibrated first transformation matrix from the laser sensor coordinate system to the robot's end-effector coordinate system and the second transformation matrix to the robot's base coordinate system, according to the coordinate transformation formula: This is converted into absolute coordinate data of each scan point in the robot's base coordinate system in three-dimensional space, where... For the weld space point in the laser sensor coordinate system, Let be the first transformation matrix. This is the second transformation matrix. To obtain the absolute coordinates of the weld points in the robot's base coordinate system, the global weld space points are acquired. The coordinates of the first acquired weld space point (100.0, 200.0, 300.0) are uniquely mapped to the initial acquisition sequence number 1. The coordinates of the second acquired weld space point (100.2, 200.5, 300.1) are mapped to the acquisition sequence number 2, and so on. All thousands of acquired weld space points are mapped to their respective cumulative acquisition sequences, and then sorted in ascending order based on the integer values ​​of the acquisition sequences. The column operation arranges the weld space point matrix from index 1 to index N, which has a fixed linear topology. It then calls the acquisition index of adjacent spatial points and performs a subtraction operation, calculating the difference between the subsequent index and the preceding index. It checks if the difference is always equal to 1. If the difference is equal to 1, it means that no frame loss occurred in the sensor's sampling data packet; if the difference is greater than 1, it means that the laser sensor was obstructed or experienced network packet loss during scanning. Abnormal isolated points are removed at these breakpoints. When only one sampling index is missing, the midpoint interpolation formula is applied using the two adjacent valid time-series coordinates before and after the breakpoint. Three-dimensional coordinate repair is performed. When the number of consecutive missing points exceeds the preset value, interpolation is not performed, and the corresponding position is determined as an unrepairable trajectory breakpoint. After comparing the acquisition sequence of adjacent points and confirming their temporal continuity, the spatial point sequence of the weld is obtained.

[0013] S33: Based on the weld spatial point sequence, consecutively acquired adjacent weld spatial points are connected sequentially to obtain point direction vectors. The angle between the previous and subsequent direction vectors is compared, and the connection relationships within the preset continuous direction range are retained. Weld spatial points are connected to obtain local trajectory segments. Local trajectory segments whose endpoint distances meet the preset splicing distance requirements are spliced ​​in the order of acquisition to generate a continuous weld running trajectory; specifically as follows: Based on the weld spatial point sequence, adjacent three-dimensional weld spatial points are sequentially extracted and coordinate subtraction is performed. The coordinates of the previous point (100.0, 200.0, 300.0) are subtracted from the coordinates of the next point (100.2, 200.5, 300.1) to obtain the vector difference of each small spatial segment. The vector difference is then substituted into the vector standardization operation to calculate the point direction vector of the line connecting the two points. The previous direction vector calculated for the previous adjacent point pair is then used. The next direction vector calculated from the next adjacent point pair Formula for calculating the angle between common input vectors The formula performs dot product and modulus division operations. Represents a unit direction vector in three-dimensional space. Representing the change in angle with units of degrees, the change in angle between adjacent direction vectors is calculated. Retrieve the process parameter database to obtain the preset angle threshold range [0°, 5°] for a continuous direction, and then calculate the angle change value. The magnitude of the change in the angle between adjacent direction vectors is compared with the boundary value of the included angle threshold interval. If the angle change is 1.5°, and it falls within the range of [0°, 5°], then the angle change is considered to be within the preset continuous range of the physical curvature smoothness. The connection between corresponding adjacent weld spatial points is preserved. If the angle change value... If the angle is 8.0°, and is considered greater than 5°, exceeding the preset continuous directional range, this location is marked as a candidate directional change point. A preset number of weld space points before and after the candidate directional change point are then read. If the directional change exists only at a single point or a few isolated points, and the corresponding laser reflection intensity or point position confidence is below a preset threshold, it is considered scanning noise and the corresponding connection is disconnected. If the directional change remains consistent across multiple consecutive points, it is considered a true weld geometry change and the corresponding connection is retained. Based on the retained smooth connection, adjacent weld space points that meet the continuity requirements are sequentially connected end-to-end to obtain local trajectory segments. The spatial coordinate values ​​of the endpoints of each local trajectory segment are extracted and substituted into the Euclidean distance formula. In the formula, the square root and square root operations are performed. It represents linear distance in space and has the dimension of millimeters. Representing the endpoint coordinates, the calculated endpoint distance between the partially broken trajectory segments is 1.2mm. The preset splicing distance requirement is retrieved from the welding control system parameters and is less than or equal to 2.0mm. The endpoint distance of 1.2mm is compared with the preset splicing distance requirement of 2.0mm. Since 1.2mm is less than or equal to 2.0mm, the endpoint distance of this minor fracture meets the preset splicing distance requirement for a smooth transition. Following the time stamp order of data collection from each local trajectory segment, and considering that the endpoint distances of adjacent local trajectory segments meet the preset splicing distance requirement, the original spatial coordinates of the two local trajectory segments remain unchanged. The endpoint of the previous local trajectory segment is used as the reference point. and the starting point of the next local trajectory segment As boundary points, and based on the unit tangent vectors corresponding to the two boundary points. Construct the transition interpolation curve using the cubic Hermite interpolation formula: Generate interpolation paths, where The tangential control length is determined based on the distance between the endpoints of adjacent trajectory segments. Distance between endpoints To preset the transition ratio coefficient, the generated transition interpolation curve is inserted between two local trajectory segments to perform interpolation and continuation operations, so that the previous local trajectory segment, the transition interpolation curve and the next local trajectory segment are connected in sequence, realizing positional continuity and tangential continuity between adjacent local trajectory segments, and generating a continuous weld movement trajectory.

[0014] S4: Based on the local trajectory segments and their splicing sequence in the continuous operation trajectory of the weld, read the weld space points in the local trajectory segments segment by segment along the splicing sequence, and read the corresponding welding gun movement record, wire feeding action record and welding gun posture record according to the order of the weld space points. Compare the welding gun posture records corresponding to adjacent weld space points in turn, and mark the posture change segment when the posture angle difference or posture offset exceeds the preset posture threshold. Define the matching range by the start and end weld space points of the posture change segment, and match the welding gun movement action and wire feeding action according to the order of the weld space points within the defined range. After completing the action matching of each posture change segment, continue the corresponding actions of each posture change segment along the splicing sequence of the local trajectory segments to generate the welding operation process path. Please see Figure 5 Specifically, S4 includes: S41: Based on the local trajectory segments and their splicing sequence in the continuous weld seam operation trajectory, read the weld seam spatial points within each local trajectory segment segment segment by segment along the splicing sequence. Read the corresponding welding torch movement record, wire feeding action record, and welding torch posture record in the order of the weld seam spatial points. Bind these three types of records to the weld seam spatial points item by item to generate a point action record sequence; specifically as follows: Based on the local trajectory segments and their splicing order in the continuous operation trajectory of the weld, the main control unit retrieves the first and second local trajectory segments according to the pre-calculated splicing order, reads the weld spatial point data stored in the memory structure of each local trajectory segment segment, and extracts the coordinates of the first weld spatial point. The coordinates of (100.0, 200.0, 300.0) and the spatial coordinates of the second weld seam are given. The coordinates are (100.2, 200.5, 300.1). Since the laser scanning points are acquired at 10ms cycles, actual control points need to be generated during the welding execution phase. Following the system sequence and based on the current pipe fitting model and bevel type, a pre-stored standard process action template is retrieved from the qualified welding process template library of the MAS system. The first movement record of the welding torch, matching the first weld seam spatial point according to the arc length ratio, is read. The corresponding TCP end linear motion speed value is 5mm / s. The corresponding first action record of the wire feeding mechanism is read, and the wire feeding motor speed mapped to the wire feeding speed value is 2. With the wire feed speed at 0.5 m / min and the wire feed start relay signal value at 1, read the first posture record of the welding gun calculated by the positive kinematics of each joint of the robot. The Euler angle deflection values ​​corresponding to this record are 0 degrees on the X-axis, 45 degrees on the Y-axis, and 90 degrees on the Z-axis. Read the second movement record of the welding gun that matches the second weld seam spatial position. The linear motion speed value corresponding to this record is 5 mm / s. Read the corresponding second action record of the wire feed. The wire feed speed value is maintained at 2.5 m / min and the signal value is 1. Read the corresponding second posture record of the welding gun. The Euler angle deflection values ​​change to 0 degrees on the X-axis, 47 degrees on the Y-axis, and 90 degrees on the Z-axis. Based on the formula for generating the TCP target position of the welding torch in For the rotation matrix corresponding to the target posture of the welding torch, The preset compensation vector for the welding torch TCP relative to the weld seam spatial points is used to calculate the corresponding target position coordinates of the first welding torch TCP (100.0, 200.0, 300.0) and the second welding torch TCP (100.2, 200.5, 300.1). During the welding execution phase, the target execution time is recalculated based on the distance between adjacent TCP target positions and the target movement speed. The execution time of adjacent target points is calculated according to... The welding torch trajectory is determined and resampled according to the underlying control cycle of the equipment. The extracted welding torch moving speed record, wire feeding drive action record, welding torch three-dimensional posture record and the corresponding resampled welding torch TCP target position coordinates are mapped and bound one-to-one based on spatial point index and cumulative arc length identifier. The TCP coordinates (100.0, 200.0, 300.0), speed 5mm / s, wire feeding speed 2.5m / min, wire feeding signal 1, Euler angle [0, 45, 90] are packaged and written into the first data structure row of the underlying database. The TCP coordinates (100.2, 200.5, 300.1) and corresponding attribute combination are written into the second data structure row. This process is repeated to complete the multi-source heterogeneous data association of all micro-interpolation points and generate a point action record sequence.

[0015] S42: Call the point action record sequence, compare the welding gun attitude records corresponding to adjacent weld seam spatial points in turn, read the attitude angle difference and attitude offset, compare the attitude angle difference and attitude offset with the preset attitude angle threshold and preset attitude offset threshold respectively, mark the corresponding segments where the position exceeds the threshold, and obtain the attitude change segment; as follows: The system retrieves the point-to-point action record sequence, sequentially extracts the welding torch attitude Euler angle record item and the welding torch TCP target position 3D coordinate item from two adjacent data structure rows, and performs a difference comparison operation. It retrieves the first attitude Euler angle [0,45,90] from the first data structure row and the second attitude Euler angle [0,47,90] from the second data structure row. It sets each Euler angle to be within a local continuous interval that does not cross the angular period boundary. The system then calls the space vector angle calculation module to directly perform the subtraction operation of the corresponding rotation angle. By subtracting the previous Y-axis angle value of 45 degrees from the subsequent Y-axis angle value of 47 degrees, the system calculates the corresponding Y-axis attitude angle difference to be 2 degrees. Finally, it extracts the first welding torch TCP target position coordinates corresponding to the first data structure row. (100.0, 200.0, 300.0) corresponds to the TCP target position coordinates of the second welding torch in the second data structure row. Substituting (100.2, 200.5, 300.1) into the three-dimensional spatial distance formula: The formula performs the square root operation on the sum of the squares of the differences between the coordinate components. Let represent the component values ​​of the two TCP target points in the world coordinate system, both with the dimension of millimeters. The axial absolute position deviation of the adjacent welding torch TCP target positions is calculated and determined as the welding torch position offset. Substitute these values ​​into the numerical solution: The calculated current welding torch position offset is approximately 0.547 mm, which means that the corresponding... The trajectory segment spatial increment over the time span is calculated by retrieving the preset welding process specification file, reading the attitude angle change threshold set for adjacent process template anchor points as 1.5 degrees, and reading the TCP position increment threshold set for adjacent process template anchor points as 0.50 mm. The TCP target positions of the first welding torch and the TCP target positions of the second welding torch are adjacent process template anchor points, and the target running time interval between the two is approximately 109.4 ms, which does not belong to the adjacent 10 ms underlying control cycle point. After identifying the attitude change segment, the trajectory, attitude, and motion parameters between anchor points are interpolated and resampled according to a 10ms device control cycle. The calculated attitude angle difference of 2 degrees is compared with the preset attitude angle threshold of 1.5 degrees. If 2 degrees is greater than 1.5 degrees, it means that the welding torch has deflected its attitude beyond the allowable range in this small trajectory segment, and the attitude angle difference is determined to exceed the threshold. The calculated welding torch position offset of 0.547mm is compared with the preset welding torch position offset threshold of 0.50mm. If 0.547mm is greater than 0.50mm, it means that the displacement increment between adjacent welding torch TCP target positions exceeds the preset position change range, and the welding torch position offset is determined to exceed the threshold. The system executes the abnormal curvature segment marking action, recording the starting index number at the spatial position corresponding to the current two data rows, until the difference between adjacent points is less than the corresponding threshold and the system returns to smoothness, and then recording the ending index number. These segments corresponding to the positions that exceed the threshold and require special kinematic interpolation compensation are retained, thus obtaining the attitude change segment. S43: Based on the attitude change segment, read the starting and ending weld spatial point positions to define the matching range. Match the welding torch movement and wire feeding actions according to the sequence of weld spatial points within the defined range. After completing the action matching for each attitude change segment, continue the corresponding actions for each attitude change segment along the splicing sequence of local trajectory segments to generate the welding operation process path; specifically as follows: Based on the attitude change segment, the first start and end weld space points (100.0, 200.0, 300.0) corresponding to the starting index number of the currently marked segment exceeding the allowable deflection range are read, and the second start and end weld space points (105.0, 202.0, 301.5) corresponding to the ending index number are read. Utilizing the timing logic relationship of the start and end points in the overall queue, a local spatial trajectory matching range strictly defined by index numbers from 1 to 25 is determined. Following the interpolation order of each weld space micro-point within this defined range, the pulse output frequency parameters of the underlying servo motion controller corresponding to each point are sequentially retrieved. The running step size of the welding robot's six-axis servo motors is hard-wired with the timing of the welding torch's movement to the specified spatial coordinates. The digital output of the control cabinet I / O board is then retrieved. In port status, the wire feeding current control drive signal, wire feeding speed control analog voltage, and the wire feeding action execution timing of the corresponding point are electrically delayed and matched to complete the cross-correlation of electromechanical actions within this attitude change segment with index numbers from 1 to 25. The next complex attitude change segment is extracted, and the above pulse and timing comparison and matching action is repeated. After completing the action matching of each individual attitude change segment, the continuous operation trajectory of the weld and its transition interpolation points generated in the previous steps are retrieved. According to the order of each point in the continuous operation trajectory, the multi-axis linkage action corresponding to the first attitude change segment, the action of the intermediate transition smooth segment, and the action corresponding to the second attitude change segment are mapped to the unified system time axis and seamlessly joined together to generate the welding operation process path.

[0016] S5: Based on the welding torch movement, wire feeding action, and action continuation relationship corresponding to each posture change segment in the welding operation process path, retrieve the operation status records of the welding robot, C-type positioner, and wire feeding mechanism. Read the welding torch movement position from the welding robot operation status record and the positioner rotation position from the C-type positioner operation status record. Match the two positions according to the same posture change segment. After the position matching is completed, extract the wire feeding action status of the wire feeding mechanism and compare it with the sequential relationship of the welding torch movement position in the same posture change segment. When the start, stop, or continuous interval of the wire feeding action is inconsistent with the sequential relationship of the welding torch movement position, it is marked as an inconsistent execution segment. After removing the inconsistent execution segments, continue the corresponding equipment actions according to the sequential order of the remaining posture change segments in the welding operation process path to generate an intelligent welding control result sequence.

[0017] Please see Figure 6 Specifically, in S5, this includes: S51: Based on the welding torch movement, wire feeding action, and action sequence relationships corresponding to each posture change segment in the welding operation process path, retrieve the operating status records of the welding robot, C-type positioner, and wire feeding mechanism, read the welding torch movement position and the positioner rotation position, perform position matching according to the same posture change segment, and generate an equipment position matching sequence; specifically as follows: Based on the welding torch movement, wire feeding, and action sequence relationships corresponding to each posture change segment in the welding process path, it is clarified that the welding robot and C-type positioner perform coordinated linkage in complex bevel or posture change segments, while in posture stable segments, the welding torch maintains a preset position and the C-type positioner drives the pipe rotation. The system retrieves the high-resolution servo encoder log files of the six-axis joint of the welding robot, the high-precision absolute photoelectric encoder records of the horizontal rotation spindle of the C-type positioner, and existing data via the internal EtherCAT digital industrial bus of the main control unit. The underlying Hall current sensor and speed feedback history log of the wire feeding mechanism are used to read the actual welding torch's first moving position coordinates (100.0, 200.0, 300.0) and second moving position coordinates (100.2, 200.5, 300.1) within the attitude change range corresponding to index numbers 1 to 25. Simultaneously, the first rotational physical position angle value (45.0 degrees) and the second rotational position angle value (45.5 degrees) of the C-type positioner at the same hardware sampling timestamp as the above two positions are read, locking the PLC time axis within the same attitude change range. The first representative timing association line is marked with the coordinates of the welding torch's first moving position (100.0, 200.0, 300.0) and its corresponding first rotation angle value of the positioner (45.0 degrees). The second representative timing association line is marked with the coordinates of the welding torch's second moving position (100.2, 200.5, 300.1) and its corresponding second rotation angle value of the positioner (45.5 ... third representative timing association line is marked with the coordinates of the welding torch's second moving position (100.2, 20 The intermediate state record formed by resampling is used to sequentially perform coordinate mapping calculations on all position data within this segment, including these twenty-five discrete sampling points, in a coordinated manner between the dual-axis robot and the positioner. All three-dimensional spatial movement positions and one-dimensional rotation angle positions are aligned using a data linked list matrix. The one-to-one correspondence in the time series is checked. The hardware timestamps of each adjacent data row are called and the difference subtraction operation is performed. The difference between the next timestamp and the previous timestamp is calculated. It is then determined whether the data time difference of each sampling point is within the preset clock tolerance range, such as 9.8ms to 10ms.Within 2ms, when the calculated difference is within the tolerance range, the corresponding data item is determined to be valid and synchronous data without delay and is retained. If the time difference calculation result of a damaged acquisition point is 50ms, the 50ms is compared with the upper limit tolerance. If the 50ms exceeds the tolerance, it means that the data of the acquisition point has lost its accurate spatial reference meaning due to severe transmission lag caused by electromagnetic interference on site. At this time, if there are valid synchronous state records before and after the abnormal record and the abnormal duration does not exceed the preset allowable duration, it is interpolated and corrected according to the valid records before and after. If the abnormal duration exceeds the preset allowable duration or there is no valid state record that can be used for interpolation, the corresponding data is marked as unusable data, and the corresponding trajectory segment is re-acquired or suspended. The abnormal record is not directly used for equipment collaborative control to prevent welding burn-through or welding defects caused by asynchrony. Finally, the isochronous mapping of the spatial geometric position relationship of each device in the same spatial stepping area and the same time sequence node is completed, and the device position matching sequence is generated. S52: Call the equipment position matching sequence, extract the wire feeding action status from the wire feeding mechanism operation status record, read the sequence of welding torch movement positions according to the same posture change segment, compare the start, stop, and continuous intervals of the wire feeding action with the sequence of welding torch movement positions in turn, mark the positions where the action relationship is inconsistent, and obtain the execution inconsistency segment; as follows: The device position matching sequence is invoked to extract the high-frequency wire feeding action state corresponding to the attitude change range from index 1 to 25 from the wire feeding mechanism's operating status record. The wire feeding action state is represented by digital signals 0 and 1, where 1 represents that the wire feeding motor has received a pulse and is in a rotating feed state, and 0 represents that the wire feeding motor is stopped and disconnected. The spatial coordinates of the welding torch's movement position are read sequentially along the time-series linked list structure of the device position matching sequence to determine the starting physical movement position coordinates of the welding torch as (100.0, 200.0, 300.0) and the stopping position coordinates as (105.0, 202.0, 301). 5) The extracted start-up transition moment when the wire feeding action state changes from 0 to 1, the stop moment when it changes from 1 to 0, and the continuous wire feeding interval where the wire feeding state remains at 1 are compared with the corresponding welding torch movement position using timing difference subtraction and logical value consistency comparison. The feedback value of the wire feeding action state when the welding torch movement position just reaches the arc starting point (100.0, 200.0, 300.0) is actually 0, indicating that the mechanical action of the wire feeder at the physical start position of the welding torch is lagging behind due to inertia and has not yet produced wire. The actual mechanical start-up moment when the wire feeding action state changes from 0 to 1 is extracted, corresponding to the welding torch's actual... The actual movement position coordinates are (100.2, 200.5, 300.1). Substituting these effective movement position coordinates with the welding torch command start position coordinates (100.0, 200.0, 300.0), we perform a square root operation on the sum of squares of the differences in the three-dimensional Euclidean distance formula. This calculates the spatial axial lag offset distance between the actual physical start time of the wire feeding action and the ideal starting point of the welding torch movement. The calculated lag offset distance due to mechanical delay is 0.547 mm. Retrieving the preset welding quality control process parameter table, we find the permissible arc-breaking action deviation distance threshold to be 0.10 mm. The calculated actual... The actual lag offset distance value of 0.547mm is compared with the action deviation distance threshold of 0.10mm. If 0.547mm is greater than 0.10mm, it is determined that the geometric correspondence between the start, stop, or continuous interval of the wire feeding action and the sequential position of the welding torch is seriously inconsistent. The segment between the current first welding torch position and the second welding torch position is marked with a timestamp abnormality label and marked as an inconsistent action relationship segment. The electromechanical action matching status of the subsequent adjacent positions is traversed and compared along the time axis until the current segment ends. All inconsistent nodes marked with collaborative delay are summarized to obtain the execution inconsistency segment.

[0018] S53: Based on the inconsistent execution segments, call the corresponding equipment actions for each posture change segment in the welding operation process path, remove the welding torch movement, wire feeding, and positioner rotation actions corresponding to the inconsistent execution segments, and continue the corresponding equipment actions in the welding operation process path according to the remaining posture change segments to generate an intelligent welding control result sequence; as detailed below. Based on the inconsistency section, the multi-source equipment action instruction set corresponding to each posture change section in the welding process path is retrieved from the temporary high-speed cache register of the central controller. The abnormal coordinated action control instruction stream marked as an inconsistency section in the welding path is locked. In the underlying logic of the multi-axis synchronous motion controller, instruction safety isolation and correction actions are performed. This is achieved by preventing the control instructions corresponding to the inconsistency section from entering the equipment execution buffer to block asynchronous pulse output. The linear motion speed value of the welding torch movement position instruction (5mm / s), the lagging wire feeding mechanism wire feeding speed control voltage signal (2.5m / min), and the rotational pulse output action of the C-type positioner, which are in the inconsistency section and pose a risk of arcing, are temporarily isolated from the execution sequence. This is based on the welding torch movement position and wire feeding... The deviation between actions is dynamically adjusted by adjusting the wire feed start and stop times, welding torch speed, section entry time, or pre-section waiting time to generate corrective equipment actions to replace the original inconsistent actions. When the inconsistent section still fails to meet the consistency requirements after a preset number of timing compensations, the section is stopped from being written into the formal execution sequence. A pause welding command and a manual review flag are generated, and the original weld trajectory and action parameters of the section are retained for replanning. The weld section is not crossed or deleted. When the corrected action relationship meets the preset consistency requirements, the corrective equipment actions are rewritten according to the original section position and replace the original isolated section. The system obtains a compliant sequence that has been corrected for abnormal data, is under a unified clock reference, and whose time deviation meets the preset timing synchronization tolerance requirements. The actions of the first effective attitude change segment, the second effective attitude change segment, and the transition smooth segment are calculated. The full-time index number of each normal segment in the overall welding control path is retrieved. Based on the absolute time sequence of each remaining attitude change segment in the welding process path from beginning to end, the time-domain spline smoothing algorithm for the equipment action parameters is invoked. Without altering the spatial positions of each point in the continuous weld trajectory, a cubic B-spline function is used. The control parameters such as the welding torch running speed, acceleration, C-type positioner angular velocity, and wire feeding speed before and after the breakpoint are smoothed in the time domain to ensure the continuity of speed and acceleration between adjacent equipment actions. Cubic B-spline smoothing is applied to the welding torch running speed, C-type positioner angular velocity, and wire feeding speed respectively. After smoothing, the speed upper limit, acceleration upper limit, equipment position correspondence, and time synchronization tolerance are re-verified. Only when all verification conditions are met is the smoothed equipment action written into the formal execution sequence. The robot servo control pulse train, positioner angular velocity feedforward command, and wire feeding start / stop digital hardwired signal of the first effective attitude change segment are sequentially connected and merged with the multi-equipment action command stream corresponding to the second effective attitude change segment based on a unified clock source time reference axis. This forms a seamless equipment collaborative execution logic that meets the preset timing synchronization tolerance requirements on the six degrees of freedom in space, external auxiliary collaborative axis, and droplet transfer mechanism, generating an intelligent welding control result sequence.

[0019] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. An intelligent welding control method for an automatic pipeline welding workstation, characterized in that, Includes the following steps: S1: Obtain the straight pipe length, end clearance distance and roller frame X-axis movable position record, subtract the end clearance distance to obtain the effective support range, filter the position item by item, check the joint clamping spacing, match and filter the Z-axis target height, combine the support coordinate points, and generate the support layout coordinate set. S2: Based on the set of coordinates of the support arrangement, select the combination that meets the quantity in the X direction, write the X-direction position value and the Z-direction target height, read the contact point and the fitting state, compare the adjacent contact positions, mark the fitting change section, replace the adjacent support coordinate points one by one, sort by quantity, and generate the support execution position sequence; S3: Based on the support execution position sequence, drive the roller support assembly to the position. After the pipe support is positioned, the laser sensor continuously scans the annular weld, records the spatial position of the weld, generates direction vectors according to adjacent points and compares the included angles, retains the continuous connection relationship and splices the local trajectory segments to generate the continuous running trajectory of the weld. S4: Based on the continuous operation trajectory of the weld, read the local trajectory segment and splicing sequence, obtain the welding gun movement record, wire feeding action record and welding gun posture record corresponding to the weld spatial point along the splicing sequence, mark the posture change segment, limit the matching range and continue the corresponding action to generate welding operation process path 1.

2. The intelligent welding control method for the automatic pipeline welding workstation according to claim 1, characterized in that: The set of support arrangement coordinates includes support coordinate points, support position distribution, and support height configuration. The support execution position sequence includes support execution coordinates, support execution order, and support combination scheme. The continuous operation trajectory of the weld includes a trajectory space point set, trajectory connection path, and trajectory continuous segment. The welding operation process path includes welding torch movement path, wire feeding action path, and welding torch posture path.

3. The intelligent welding control method for the automatic pipeline welding workstation according to claim 1, characterized in that: The steps for obtaining S1 are as follows: S11: Obtain the straight pipe length, end clearance distance, and roller frame X-direction movable position record. Determine the effective support interval by subtracting the end clearance distance from the straight pipe length. Substitute the roller frame X-direction movable position record into the effective support interval filter in sequence, retain the position arrangement within the interval, and generate the position corresponding sequence. S12: Call the position corresponding sequence to obtain the joint clamping spacing record, compare adjacent positions in turn, remove positions that do not meet the joint clamping spacing, form a spacing qualified position sequence, obtain the straight pipe outer diameter and Z-direction height corresponding record, match the Z-direction target height item by item according to the spacing qualified position sequence, and generate a height corresponding sequence. S13: Based on the height correspondence sequence, call the height adjustment range, substitute the Z-direction target height into the height adjustment range filter in sequence, remove positions that exceed the height adjustment range, form a height qualified position sequence, combine the X-direction position value in the height qualified position sequence with the corresponding Z-direction target height to form a support coordinate point, and generate a support layout coordinate set.

4. The intelligent welding control method for the automatic pipeline welding workstation according to claim 1, characterized in that: The steps for obtaining S2 are as follows: S21: Based on the support coordinate points of the support arrangement coordinate set, read the support coordinate points in ascending order of X-direction position value, compare the reading order with the support quantity requirement item by item, select the position that meets the support quantity requirement, retain the corresponding coordinate point arrangement, and record the support coordinate point sequence number. Write the X-direction position value and Z-direction target height into the roller support assembly. After the roller support assembly is in place, generate the current support position combination. S22: Call the current support position combination, read the contact point distribution state between the roller support assembly and the outer surface of the pipe, determine the contact point as the contact position, obtain the corresponding bonding state of the contact position, compare adjacent bonding states according to the contact position arrangement order, mark the bonding change section where the bonding state is inconsistent or the bonding difference exceeds the bonding difference threshold, retain the section order, count the number of marked sections, and obtain the number of bonding change sections; S23: Based on the number of fitting change segments, call the current support position combination, extract the adjacent support coordinate points before and after each support coordinate point in the current combination according to the X-direction position value from small to large, and replace them one by one. After each replacement, rewrite the X-direction position value and Z-direction target height, recount the number of fitting change segments, sort them according to the number of fitting change segments from small to large, select the support position combination corresponding to the first sorted segment, and generate the support execution position sequence.

5. The intelligent welding control method for the automatic pipeline welding workstation according to claim 1, characterized in that: The steps for obtaining S3 are as follows: S31: Based on the support position combination in the support execution position sequence, extract the X-direction position value and Z-direction target height of each support coordinate point, write them into the roller support assembly in the order of the support coordinate points, drive the roller support assembly to the corresponding support position, and generate the support position coordinate sequence after the pipe has completed the support positioning. S32: Call the support positioning coordinate sequence, call the laser sensor to perform continuous scanning along the annular weld, record the weld space points according to the acquisition sequence, match the weld space points with the acquisition sequence number one by one, arrange the weld space points according to the acquisition sequence number, compare the acquisition order of adjacent points, and obtain the weld space point sequence. S33: According to the weld space point sequence, the adjacent weld space points are continuously collected and connected sequentially to obtain the point direction vector. The angle between the previous direction vector and the next direction vector is compared, and the connection relationship within the preset continuous range is retained. The weld space points are connected to obtain local trajectory segments. The local trajectory segments whose endpoint distance reaches the preset splicing distance requirement are spliced ​​in the order of collection to generate the continuous running trajectory of the weld.

6. The intelligent welding control method for the automatic pipeline welding workstation according to claim 1, characterized in that: The steps for obtaining S4 are as follows: S41: Based on the local trajectory segments and their splicing order in the continuous operation trajectory of the weld, read the weld space points in the local trajectory segments one by one along the splicing order, and read the corresponding welding gun movement record, wire feeding action record and welding gun posture record in the order of the weld space points. Bind the three types of records to the weld space points one by one to generate a point action record sequence. S42: Call the point action record sequence, compare the welding gun posture records corresponding to adjacent weld space points in sequence, read the posture angle difference and posture offset, compare the posture angle difference and posture offset with the preset posture angle threshold and preset posture offset threshold respectively, mark the corresponding segment of the position exceeding the threshold, and obtain the posture change segment. S43: Based on the posture change segment, read the starting and ending weld space point limit matching range, match the welding gun movement and wire feeding action according to the sequence of weld space points within the limit range, and after completing the action matching of each posture change segment, continue the corresponding actions of each posture change segment along the local trajectory segment splicing sequence to generate the welding operation process path.

7. The intelligent welding control method for the automatic pipeline welding workstation according to claim 1, characterized in that: The method further includes: S5: Based on the welding operation process path, retrieve the status of the welding robot, C-type positioner and wire feeding mechanism, match the welding torch movement position and the positioner rotation position according to the same posture change segment, compare the wire feeding action status and the sequence of movement, eliminate inconsistent execution segments, and generate an intelligent welding control result sequence. The intelligent welding control result sequence includes welding robot control commands, C-type positioner control commands, and wire feeding mechanism control commands.

8. The intelligent welding control method for the automatic pipeline welding workstation according to claim 7, characterized in that: The steps for obtaining S5 are as follows: S51: Based on the welding torch movement, wire feeding action and action continuation relationship corresponding to each posture change segment in the welding operation process path, retrieve the operating status records of the welding robot, C-type positioner and wire feeding mechanism, read the welding torch movement position and the positioner rotation position, perform position matching according to the same posture change segment, and generate equipment position matching sequence. S52: Call the device position matching sequence, extract the wire feeding action status from the wire feeding mechanism operation status record, read the sequential relationship of the welding gun movement position according to the same posture change segment, compare the start, stop and continuous intervals of the wire feeding action with the sequential relationship of the welding gun movement position in turn, mark the positions where the action relationship is inconsistent, and obtain the execution inconsistency segment. S53: Based on the inconsistency section, call the corresponding equipment actions of each posture change section in the welding operation process path, remove the welding torch movement action, wire feeding action and positioner rotation action corresponding to the inconsistency section, and continue the corresponding equipment actions in the welding operation process path according to the remaining posture change sections to generate an intelligent welding control result sequence.