A method and system for controlling the welding quality of a flanged circular pipe girth weld

Abstract

The application relates to the technical field of robot welding control, and discloses a flange and circular pipe girth weld welding quality control method and system. The method acquires pre-welding detection data of a flange and a circular pipe after clamping and establishes a girth weld workpiece coordinate system; a local joint section contour is obtained by intercepting the girth weld in the circumferential direction, the local joint section contour is matched with a preset qualified weld section, a to-be-filled section area is obtained, and a local section filling model is constructed; a circumferential section control curve is generated based on the local section filling model, and a welding control section is divided; a molten pool target corridor and an initial welding control quantity are generated according to the welding control section; during welding, the molten pool target corridor is mapped to a molten pool image coordinate, and an actual molten pool envelope is extracted; position deviation components and filling deviation components are obtained by decomposing the envelope deviation between the actual molten pool envelope and the molten pool target corridor, welding gun posture correction quantities and welding heat input correction quantities are respectively generated, and a post-section heat memory quantity is generated according to actual heat input and correction residual error, so as to update subsequent welding control sections.

Description

Technical Field

[0001] This application relates to the field of robotic welding control technology, and in particular to a method and system for controlling the welding quality of flange circular pipe circumferential welds. Background Technology

[0002] Circumferential welds between flanges and round pipes are widely used in the manufacturing of pipelines, pressure vessels, equipment supports, and other metal structural components. With the improvement of automated manufacturing, welding robots are increasingly being applied to circumferential welds on flange and round pipe workpieces. Existing robotic circumferential weld methods typically obtain the weld position before welding through visual inspection, 3D inspection, or weld tracking devices, and then generate a welding trajectory based on the detected weld centerline or bevel boundary. During the welding process, some methods also incorporate molten pool images, arc voltage, arc current, or wire feeding status to provide feedback corrections to the welding torch position and welding process parameters.

[0003] For circumferential welds with relatively stable assembly and minimal changes in weld morphology, the above method can meet certain automatic welding requirements. However, after the flange and the round pipe are clamped, the relative position between the end of the round pipe and the end of the flange, the bevel opening, local misalignment, and the transition position of the tack weld often change along the circumference of the circumferential weld. If control is still mainly based on the overall weld centerline or a uniform welding trajectory, the actual differences in the morphology to be welded at different circumferential positions are difficult to be reflected in the welding torch posture and welding heat input control in a timely manner, which can easily lead to problems such as insufficient deposition, weld pool offset, heat input mismatch, or unstable formation in local positions.

[0004] The molten pool images acquired during robotic welding are easily affected by arc light, flares, spatter highlights, and reflections from the metal surface. Existing correction methods based on molten pool images may misjudge high-brightness interference from non-real molten pool boundaries as changes in welding status if they directly rely on feedback control based on brightness abrupt changes, boundary changes, or molten pool width changes in a single frame image. This could lead to incorrect corrections of the welding torch posture or welding parameters.

[0005] Furthermore, circumferential welds have a continuous circumferential path characteristic, and the heat input and weld formation state of the current welding section will affect subsequent adjacent sections. Existing welding control methods focus more on real-time detection and correction of the current welding position, neglecting the influence between the welding result of the current section and the control basis of subsequent sections. This can easily lead to subsequent sections entering the welding process according to preset parameters, making it difficult to correct the local heat effects or cumulative forming deviations caused by previous welding.

[0006] Therefore, in the automatic welding process of flange circular pipe circumferential welds, there is still a need for a welding quality control method that can take into account the actual assembly morphology before welding, the identification of the molten pool state during welding, and the influence of circumferential continuous welding, so as to improve the matching between welding torch posture adjustment and welding heat input control and the actual welding state. Summary of the Invention

[0007] The purpose of this invention is to provide a welding quality control method and system for flange circumferential welds, in order to solve the problem that in the existing automatic welding process of flange circumferential welds, it is difficult to coordinate the welding torch posture and welding heat input according to the actual joint morphology and the state of the molten pool during welding.

[0008] To solve the above-mentioned technical problems, the present invention is implemented using the following technical solution.

[0009] On one hand, the present invention provides a method and system for controlling the welding quality of circumferential welds on flanged pipes, comprising: Obtain pre-welding inspection data after clamping the flange and round pipe fitting, and establish a workpiece coordinate system for the circumferential weld. Multiple local joint cross-sectional profiles are cut along the circumference of the circumferential weld. Each local joint cross-sectional profile is registered with a preset qualified weld cross-section to obtain the cross-sectional area to be filled, and a local cross-sectional filling model is constructed. Based on the local section filling model, a circumferential section control curve is generated, and welding control segments are divided accordingly. Generate the target corridor of the molten pool and the initial welding control quantity based on the local cross-section filling model corresponding to each welding control segment; When welding the current welding control segment, the target corridor of the molten pool is mapped to the coordinates of the molten pool image, and the actual molten pool envelope is extracted; Based on the envelope deviation between the actual molten pool envelope and the target corridor of the molten pool, the position deviation component and the filling deviation component are decomposed and the welding gun attitude correction amount and the welding heat input correction amount are generated respectively. The heat memory value after the current welding control segment is generated based on the actual heat input and the correction residual, and the subsequent welding control segment is updated based on the heat memory value after the segment.

[0010] Preferably, multiple local joint cross-sectional profiles are cut along the circumference of the circumferential weld, including: The circumferential welding path is determined using the axis of the circular tube in the workpiece coordinate system of the circumferential weld. Multiple cross-sectional sampling positions are set along the circumferential welding path; At each of the aforementioned cross-sectional sampling locations, a local joint cross-sectional profile perpendicular to the welding advance direction is extracted; The local joint cross-sectional profile is formed by the intersection profile of the pre-welding inspection data on the corresponding cross-sectional plane, and includes the flange end profile, the round pipe end profile, and the bevel area profile.

[0011] Preferably, the profile of each local joint section is registered with a preset qualified weld section to obtain the section region to be filled, and a local section filling model is constructed, including: The preset qualified weld section is established based on the target weld forming size, weld reinforcement range, and the coverage boundaries on both sides of the weld. The coordinates of the preset qualified weld section and the local joint section profile are registered. Within the coverage area of ​​the preset qualified weld section, the area already occupied by the flange end, the round pipe end, and the bevel entity is deducted to obtain the section area to be filled; Based on the cross-sectional area to be filled, generate local cross-sectional filling units corresponding to the cross-sectional sampling positions; Multiple local cross-section filling units are arranged in circumferential order according to the sampling positions of the cross-section to form the local cross-section filling model.

[0012] Preferably, a circumferential section control curve is generated based on the local section filling model, and welding control segments are divided accordingly, including: Extract the center point of the weld coverage from the section area to be filled in each local section filling unit; Connect multiple weld coverage center points in circumferential order according to the sampling positions of the cross section to generate a circumferential cross section control curve; The circumferential section control curve is smoothed. The circumferential interval where the circumferential section control curve changes continuously is divided into welding control segments, and the circumferential interval where the circumferential section control curve undergoes a local abrupt change and its adjacent intervals are divided into transition welding control segments.

[0013] Preferably, the target corridor of the molten pool and the initial welding control quantity are generated based on the local cross-section filling model corresponding to each welding control segment, including: Read multiple local section filling elements within the current welding control segment; Generate the weld coverage centerline based on the weld coverage center point of multiple cross-sectional areas to be filled; Based on the distribution range of the cross-sectional area to be filled on both sides of the center line of the weld coverage, a permissible boundary zone for the weld pool is generated; The weld pool target corridor is defined by the weld coverage centerline and the weld pool boundary allowance band; The initial trajectory of the welding torch is generated based on the center line of the weld coverage, and the initial amount of welding heat input is generated based on the allowable zone of the weld pool boundary.

[0014] Preferably, when welding the current welding control segment, the target corridor of the molten pool is mapped to the coordinates of the molten pool image, and the actual molten pool envelope is extracted, including: Based on the current position of the welding torch, the welding torch posture, and the calibration relationship of the vision acquisition device, the target corridor of the molten pool is mapped to the coordinates of the molten pool image to obtain the image constraint region; Acquire continuous molten pool image frames within the image constraint region; Extract candidate molten pool boundary points from the continuous molten pool image frames; Based on the spatial adjacency relationship and motion trend continuity of candidate molten pool boundary points in adjacent molten pool image frames, the candidate molten pool boundary points are sequentially connected. The stable boundary line or boundary region of the melt pool formed by the candidate melt pool boundary points that have completed the timing connection is determined as the actual melt pool envelope. Highlighted areas located outside the image constraint region, or areas with abrupt brightness changes that do not form temporally continuous boundaries, are excluded from the actual melt pool envelope.

[0015] Preferably, based on the envelope deviation between the actual molten pool envelope and the target corridor of the molten pool, position deviation components and filling deviation components are decomposed, and welding torch attitude correction amount and welding heat input correction amount are generated respectively, including: Multiple envelope comparison sections are set along the target corridor of the molten pool; The target center point of the target corridor of the molten pool is determined in each envelope comparison section, and the actual envelope center point is determined according to the actual molten pool envelope; Based on the offset of the actual envelope center point relative to the target center point, a position deviation component is generated; A filling deviation component is generated based on the amount of insufficient or excessive coverage of the actual molten pool envelope relative to the allowable zone of the molten pool boundary. When the positional deviation components in multiple envelope comparison sections characterize the overall translation of the actual molten pool envelope, a correction amount for the center point of the welding torch tool is generated. When the positional deviation component in multiple envelope comparison sections represents the angular deviation of the actual molten pool envelope relative to the target corridor of the molten pool, a welding torch angle correction amount is generated. Based on the filling deviation component, a welding current correction, a welding speed correction, or a wire feed speed correction is generated.

[0016] Preferably, generating post-segment thermal memory based on the actual heat input and corrected residual of the current welding control segment, and updating subsequent welding control segments based on the post-segment thermal memory, includes: The actual welding heat input of the current welding control section is determined based on the welding current, welding voltage, and welding speed within the current welding control section. The residual amount of envelope deviation is determined based on the position deviation component and fill deviation component that still exist at the end of the current welding control segment; A heat-affected component is generated based on the actual welding heat input, a forming deviation component is generated based on the envelope deviation residual amount, and a post-segment thermal memory amount is formed by the heat-affected component and the forming deviation component. The heat memory value after the segment is superimposed onto the subsequent welding control segment according to the attenuation relationship between the welding advance direction and the circumferential distance; Based on the superimposed post-segment thermal memory, the target corridor of the molten pool or the initial welding control quantity of the subsequent welding control segment is pre-corrected.

[0017] Preferably, after dividing the circumferential section control curve where a local abrupt change occurs and its adjacent sections into transition welding control segments, the method further includes: Based on the target corridor of the molten pool corresponding to the welding control segments on both sides of the transition welding control segment, a target corridor of the transition molten pool is generated. A gradual initial welding control quantity is generated based on the target corridor of the transition molten pool; When welding the transition welding control section, the unit time variation range of the welding torch posture correction amount and the welding heat input correction amount is limited according to the gradual initial welding control amount.

[0018] On the other hand, the present invention provides a welding quality control system for flange circular pipe circumferential welds, comprising: The cross-section model construction module is used to obtain pre-welding inspection data after the flange and round pipe fitting are clamped, establish the workpiece coordinate system of the circumferential weld, and construct a local cross-section filling model along the circumference of the circumferential weld. The control segment generation module is used to generate a circumferential section control curve based on the local section filling model, and to divide the welding control segment according to the circumferential section control curve; The target corridor generation module is used to generate the target corridor of the molten pool and the initial welding control quantity based on the local cross-section filling model corresponding to each welding control segment. The molten pool envelope extraction module is used to map the molten pool target corridor to the molten pool image coordinates, and extract the actual molten pool envelope based on the molten pool target corridor; The decoupling correction module is used to generate position deviation components and filling deviation components based on the envelope deviation of the actual molten pool envelope relative to the target corridor of the molten pool, and to generate welding torch attitude correction amount and welding heat input correction amount respectively. The thermal memory update module is used to generate the post-segment thermal memory amount based on the actual heat input and correction residual of the current welding control segment, and update the subsequent welding control segment based on the post-segment thermal memory amount.

[0019] Compared with the prior art, the beneficial effects achieved by the present invention are as follows: 1. Convert the pre-weld inspection data of the flange and round pipe fitting after clamping into a local section filling model distributed along the circumference of the circumferential weld. By registering the local joint section profile with the preset qualified weld section, the section area to be filled is obtained. This transforms the welding control basis from the traditional weld centerline or single assembly deviation into a section filling object that can directly reflect the local deposition compensation requirements. This allows gaps, misalignments, bevel openings or local morphological changes at different circumferential positions to enter the subsequent control process through the same section area to be filled, avoiding the control fragmentation caused by establishing control rules for multiple isolated deviations. 2. Generate a circumferential section control curve based on the local section filling model, and divide the welding control segment according to the continuous change range and local abrupt change range of the curve. Make the welding control segment correspond to the circumferential change trend of the deposition requirements. Compared with the method of segmenting by fixed angle or continuous welding by a single weld centerline, it can maintain continuous control at the local joint section where the change is stable and form transition control at the local morphology abrupt change, thereby reducing the risk of unreasonable jumps in welding torch posture or heat input at the abrupt change position. 3. Generate a target corridor for the molten pool based on the section to be filled within the welding control section, and map the target corridor to the coordinates of the molten pool image as an image constraint region. This makes the extraction of the actual molten pool envelope during welding no longer dependent on brightness abrupt changes or single-frame boundary changes in the entire image. The actual molten pool envelope is formed by the temporal boundaries in continuous molten pool image frames under the constraint of the target corridor. This can reduce the interference of arc light, flare and spatter highlights on the judgment of the molten pool state, and ensure that the image recognition results during welding correspond to the local filling requirements determined before welding. 4. Compare the actual molten pool envelope with the target molten pool corridor, and decompose the envelope deviation into position deviation component and filling deviation component. The position deviation component enters the welding torch posture correction chain, and the filling deviation component enters the welding heat input correction chain. This gives the welding torch pointing correction and molten deposit filling correction clear data sources and control division of labor, thereby reducing the risk of incorrect correction caused by simply adjusting the welding torch posture or process parameters uniformly based on molten pool anomalies.

[0020] 5. Based on the actual heat input of the current welding control segment and the correction residual, generate the heat memory value after the segment, and superimpose the heat memory value after the segment onto the subsequent welding control segment. This is used to pre-correct the target corridor of the molten pool or the initial welding control value of the subsequent welding control segment, so that the heat influence and forming deviation of the current segment no longer remain in the instantaneous feedback control, but can enter the control basis of the subsequent segment along the welding advance direction, thereby forming a welding quality control chain that is continuously updated along the circumference of the circumferential weld. Attached Figure Description

[0021] Figure 1 The diagram shows a flowchart of a robotic welding quality control method for circumferential welds of flanges and round pipe fittings. Figure 2 The diagram shows a flowchart of obtaining the local joint cross-sectional profile using the workpiece coordinate system of a circumferential weld. Figure 3 The diagram shows the process of generating a local section filling model from a preset qualified weld section. Figure 4 The diagram shows the process of generating circumferential section control curves and dividing welding control segments using local section filling elements. Figure 5 The diagram shows the process of generating the target corridor of the molten pool and the initial welding control quantities for the local section filling unit; Figure 6 The diagram shows the process of generating welding correction amounts from the actual molten pool envelope and performing subsequent control corrections. Detailed Implementation

[0022] This invention can be modified in many ways and has many embodiments, with specific embodiments shown in the accompanying drawings for detailed description. However, this does not mean that the invention is limited to a specific implementation; it should be understood that all modifications, equivalents, and even substitutions falling within the concept and technical scope of this invention are included in this invention. Similar reference numerals are used for similar constituent elements in the description of the drawings.

[0023] The terms “first,” “second,” “A,” “B,” etc., are used to describe a wide variety of constituent elements, but these constituent elements are not limited by these terms. These terms are used to distinguish one constituent element from others. For example, without departing from the scope of the invention, a first constituent element may be named a second constituent element, and similarly, a second constituent element may be named a first constituent element. The term “and / or” includes a combination of multiple associated descriptions, or one of multiple associated descriptions.

[0024] When it is mentioned that a certain constituent element is "connected" or "linked" to other constituent elements, it can mean not only that the element is directly connected or linked to the other constituent element, but also that there are other constituent elements in between. Conversely, when it is mentioned that a certain constituent element is "directly connected" or "directly linked" to other constituent elements, there are no other constituent elements in between.

[0025] The terminology used in this application is for illustrative purposes only and is not intended to limit the scope of the invention. Singular expressions include plural expressions unless the context clearly indicates otherwise. Terms such as "comprising" or "having" as used herein do not preclude the possibility of the presence or addition of features, numbers, stages, actions, constituent elements, components, or combinations thereof described in the specification.

[0026] Unless otherwise defined, all terms used herein, including technical or scientific terms, have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0027] Terms defined in common dictionaries should be interpreted as having the same meaning as in the context of the relevant technology, and should not be interpreted as having an ideal or overly formal meaning unless explicitly defined in this application.

[0028] Example 1: refer to Figures 1 to 6 As shown, this embodiment relates to a welding quality control method for circumferential welds of flanges and pipes, which is applicable to scenarios where welding robots perform circumferential welds on clamped flanges and pipes. After clamping, the flange and pipe form an annular area to be welded, and the welding robot performs welding along the circumferential path of the annular area to be welded.

[0029] The method in this embodiment is executed by a welding quality control system for flange circular pipe circumferential welds. The system includes a cross-section model construction module, a control segment generation module, a target corridor generation module, a molten pool envelope extraction module, a decoupling correction module, and a thermal memory update module. These modules can be integrated into a welding robot controller.

[0030] The method includes: S101. Obtain the pre-welding inspection data after the flange and round pipe fitting are clamped, and establish the workpiece coordinate system for the circumferential weld. S102. Cut out multiple local joint cross-sectional profiles along the circumferential direction of the circumferential weld, register each local joint cross-sectional profile with the preset qualified weld cross-section to obtain the cross-sectional area to be filled, and construct a local cross-sectional filling model. S103. Generate a circumferential section control curve based on the local section filling model, and divide the welding control segment accordingly; S104. Generate the target corridor of the molten pool and the initial welding control quantity based on the local section filling model corresponding to each welding control segment; S105. When welding the current welding control segment, map the target corridor of the molten pool to the coordinates of the molten pool image and extract the actual molten pool envelope; S106. Based on the envelope deviation between the actual molten pool envelope and the target corridor of the molten pool, the position deviation component and the filling deviation component are decomposed and the welding gun posture correction amount and the welding heat input correction amount are generated respectively. S107. Generate post-segment thermal memory based on the actual heat input and corrected residual of the current welding control segment, and update the subsequent welding control segment based on the post-segment thermal memory.

[0031] Specifically, the cross-section model construction module acquires the pre-welding inspection data after the flange and the round pipe are clamped. The pre-welding inspection data is collected by a three-dimensional vision inspection device, a line laser inspection device, or a structured light inspection device, and includes the three-dimensional contours of the flange end, the round pipe end, and the bevel area in the clamped state.

[0032] The cross-section model construction module fits the axis of the circular tube according to the outer circle contour of the circular tube, and determines the flange reference plane according to the flange end face contour; using the axis of the circular tube as the circumferential positioning reference and the flange reference plane as the axial positioning reference, the coordinate system of the circumferential weld workpiece is established.

[0033] In the workpiece coordinate system of the circumferential weld, the cross-section model construction module sets multiple cross-section sampling positions along the circumference of the circumferential weld, and extracts a local joint cross-section profile perpendicular to the welding advance direction at each cross-section sampling position. The local joint cross-section profile is formed by the intersection profile of the pre-weld inspection data on the corresponding cross-section plane, including the flange end profile, the circular pipe end profile, and the bevel area profile within the cross-section.

[0034] When using the local joint cross-sectional profile as the treatment object, the actual area that the weld metal needs to compensate for is the gap area in the cross-sectional space, rather than the geometric deviation in the welding process alone. Furthermore, axial clearance, local misalignment, bevel opening, and end face inclination are ultimately reflected as changes in the position, boundary, or area of ​​the area to be welded in the cross-sectional profile.

[0035] Based on the target weld forming size, weld reinforcement range, and weld coverage boundaries on both sides, a preset qualified weld section is established. When qualified sample data is available, the section model construction module also extracts the weld outer contour boundary from the cross-sectional contour of the qualified sample and uses this weld outer contour boundary as the boundary template of the preset qualified weld section.

[0036] For each cross-sectional sampling location, the preset qualified weld cross-section and the local joint cross-section contour are coordinate registered. During the registration process, the contour of the circular pipe end and the contour of the flange end in the current cross-section are used as the joint reference, so that the coverage boundary of the preset qualified weld cross-section falls into the corresponding position of the current local joint cross-section.

[0037] After registration, the area occupied by the flange end, the circular pipe end, and the bevel entity is deducted from the coverage area of ​​the preset qualified weld section. The remaining area is designated as the section area to be filled, which is the section space that needs to be compensated by the deposited metal at the current circumferential sampling position. Local section filling units are generated based on the section area to be filled. Each local section filling unit records the boundary of the section area to be filled and its position in the coordinate system of the circumferential weld workpiece. Multiple local section filling units are arranged in the order of the circumferential sampling positions to form a local section filling model. This converges the local geometric changes caused by clamping deviations into the section area to be filled, so that subsequent control segment division, target corridor generation, and in-weld correction all revolve around the same section filling object.

[0038] The local section filling model is read and the local section filling units are processed sequentially. For each local section filling unit, the control segment generation module extracts the fusion coverage center point from the section area to be filled. The fusion coverage center point is determined by the regional centroid, boundary centerline point or target fusion center point of the section area to be filled.

[0039] By sequentially connecting multiple weld coverage center points through circumferential sampling positions, a circumferential section control curve is generated. This circumferential section control curve characterizes the change in weld coverage position along the circumference of the circumferential weld. The circumferential section control curve is smoothed, and circumferential paths are divided according to the curve's change state. Circumferential intervals with continuous curve changes are defined as welding control segments; circumferential intervals with local abrupt changes and their adjacent intervals are defined as transition welding control segments.

[0040] Multiple local section filling units within the current welding control segment are read, and a weld coverage centerline is generated based on the weld coverage center point of each section area to be filled. The weld coverage centerline corresponds to the path position that the molten pool should mainly cover within the current welding control segment. Based on the distribution range of the section areas to be filled on both sides of the weld coverage centerline, a weld pool boundary allowable zone is generated. The weld pool boundary allowable zone varies with the actual width of the section area to be filled; where the section area to be filled is wider, the boundary allowable zone is correspondingly wider, and where the section area to be filled is narrower, the boundary allowable zone is correspondingly narrower.

[0041] The center line of the weld deposition coverage and the boundary allowable band of the weld pool jointly define the target corridor of the weld pool. The target corridor of the weld pool forms the target weld pool coverage range of the current welding control segment in the workpiece coordinate system of the circumferential weld seam, and is mapped to the image constraint area extracted from the weld pool boundary during welding image processing.

[0042] Initial welding control quantities are generated based on the target corridor of the molten pool. Specifically, the weld deposition centerline is included in the welding torch trajectory generation process, and the molten pool boundary allowable zone is included in the initial welding heat input generation process. Therefore, both the welding torch pointing and the molten pool coverage capability are determined by the same cross-sectional filling foundation.

[0043] When welding the current welding control segment, the molten pool envelope extraction module maps the target corridor of the molten pool from the workpiece coordinate system of the circumferential weld to the molten pool image coordinate system based on the current position of the welding torch, the welding torch posture, and the calibration relationship between the vision acquisition device and the welding torch, thus obtaining the image constraint region.

[0044] Within the image constraint region, continuous molten pool image frames are acquired, and candidate molten pool boundary points are extracted in each molten pool image frame. For candidate molten pool boundary points in adjacent image frames, the molten pool envelope extraction module connects them according to spatial adjacency and motion trend continuity to form a stable molten pool boundary line.

[0045] The boundary region enclosed or defined by the stable boundary line of the molten pool is defined as the actual molten pool envelope. The actual molten pool envelope includes a closed boundary region and a set of non-closed boundary lines formed when locally blocked by arc light. When the set of non-closed boundary lines maintains a temporal connection relationship in consecutive molten pool image frames, it enters the actual molten pool envelope.

[0046] Highlighted areas outside the image constraint region are not included in the candidate boundary point extraction process. Brightness abrupt changes that occur only in a single frame and cannot form a temporally continuous boundary within the image constraint region are removed from the candidate boundary points. Therefore, arcs, flares, and spatters do not directly constitute the actual melt pool envelope.

[0047] The actual molten pool envelope is overlapped and compared with the target corridor of the molten pool. During the comparison, the center line of the weld coverage in the target corridor of the molten pool is used as the position reference, and the allowable zone of the molten pool boundary is used as the coverage range reference. To obtain the position deviation component, the decoupling correction module sets multiple envelope comparison sections in the target corridor of the molten pool along the welding advance direction. Each envelope comparison section intersects with the weld coverage center line to form the target center point, and intersects with the actual molten pool envelope to form the actual envelope cross section. The actual envelope center point is determined according to the midpoint of the actual envelope cross section or the center line point of the stable boundary. The offset of the actual envelope center point relative to the target center point is calculated. The offsets of multiple envelope comparison sections are converted to the coordinate system of the circumferential weld workpiece through the calibration relationship of the vision acquisition device to form the position deviation component.

[0048] The position deviation component includes the offset direction and the offset distance. The offset direction is the offset direction of the actual envelope center point relative to the target center point in the coordinate system of the circumferential weld workpiece, and the offset distance is the spatial distance between the two. To reduce the impact of single-frame image disturbance on the correction amount, the position deviation component in multiple consecutive acquisition cycles is smoothed to obtain the position deviation input amount of the current control cycle.

[0049] The welding gun attitude correction method is determined based on the spatial distribution of the position deviation input. When the actual molten pool envelope deviates from the weld coverage center line in the same direction in the envelope comparison sections corresponding to the front, middle and tail ends, and the difference in offset distance of each envelope comparison section is within the preset translation judgment range, it is determined that the actual molten pool envelope has undergone overall translation relative to the molten pool target corridor, and the welding gun tool center point correction amount is generated. The welding gun tool center point correction amount is a displacement vector in the workpiece coordinate system of the circumferential weld or the robot base coordinate system, which is used to compensate for the deviation of the welding gun tool center point relative to the weld coverage center line.

[0050] When the offset distance of the actual molten pool envelope in the envelope comparison section corresponding to the front, middle and tail ends increases or decreases along the welding advance direction, or when the center line of the actual molten pool envelope deviates from the center line of the weld deposition coverage, it is determined that the actual molten pool envelope has deviated from the attitude angle relative to the target corridor of the molten pool, and a welding torch angle correction amount is generated. The welding torch angle correction amount is the angle increment of the welding torch working angle or the welding torch travel angle, which is used to correct the incident direction of the welding torch relative to the current welding control section.

[0051] When the actual molten pool envelope exhibits both overall translation and attitude angle deviation, the decoupling correction module generates a tool center point correction and a tool angle correction, respectively, and combines them into a tool attitude correction. During the combination, the tool center point correction applies to the tool's pointing position, and the angle correction applies to the tool's incident direction, to prevent simply rotating the tool from causing further deviation of the molten pool center from the weld deposition coverage centerline.

[0052] To obtain the filling deviation component, the relationship between the actual envelope cutoff line and the allowable zone of the molten pool boundary is compared on the same set of envelope comparison sections. If the actual envelope cutoff line does not reach the inner side requirement of the allowable zone of the molten pool boundary, an insufficient coverage amount is generated; if the actual envelope cutoff line exceeds the outer boundary of the allowable zone of the molten pool boundary, an excessive coverage amount is generated. The insufficient coverage amount and the excessive coverage amount together form the filling deviation component.

[0053] A welding heat input correction is generated based on the filling deviation component. This correction is at least one of a welding current correction, a welding speed correction, or a wire feed speed correction. When the under-coverage increases, a heat input correction is generated to improve the weld filling capability; when the over-coverage increases, a heat input correction is generated to limit the expansion of the molten pool.

[0054] In one implementation, the decoupling correction module pre-stores attitude correction coefficients and thermal input correction coefficients. The attitude correction coefficients convert the position deviation input into the displacement of the welding torch tool center point or the welding torch angle increment; the thermal input correction coefficients convert the filling deviation component into corrections for welding current, welding speed, or wire feed speed. These attitude correction coefficients and thermal input correction coefficients are obtained through welding process experiments or robot calibration processes and are stored corresponding to the current workpiece specifications, welding materials, and welding process parameters.

[0055] The welding torch posture correction and welding heat input correction are output to the robot controller and welding power control unit. The robot controller updates the welding torch posture based on the welding torch tool center point correction and welding torch angle correction. The welding power control unit or wire feed control unit updates the welding current, welding speed or wire feed speed based on the welding heat input correction.

[0056] The position deviation component is fed into the welding torch posture correction chain, and the filling deviation component is fed into the welding heat input correction chain. Both the position deviation component and the filling deviation component are derived from the same comparison between the actual molten pool envelope and the target corridor of the molten pool, so that the welding torch pointing correction and the deposition filling correction have a clear data source.

[0057] After the current welding control segment is completed, the thermal memory update module reads the welding current, welding voltage, welding speed and welding duration of that segment, and calculates the actual welding heat input of the current welding control segment based on the above welding process parameters. The actual welding heat input is determined according to the energy input per unit welding length, or according to the cumulative value of heat input in multiple control cycles within the current welding control segment.

[0058] The position deviation component and fill deviation component at the end of the current welding control segment are read, and the residual envelope deviation is generated based on the two. The residual value of the position deviation component reflects the offset of the actual molten pool envelope center relative to the weld coverage center line that has not yet been eliminated, and the residual value of the fill deviation component reflects the actual molten pool envelope relative to the allowable zone of the molten pool boundary that still has insufficient or excessive coverage.

[0059] A heat-affected zone (HAZ) component is generated based on the actual welding heat input, and a forming deviation component is generated based on the residual envelope deviation. The HAZ and forming deviation components together form the post-segment thermal memory. The post-segment thermal memory is the state quantity retained after the current welding control segment ends, characterizing the residual heat effect and forming deviation impact of the current segment on subsequent welding control segments. The post-segment thermal memory is superimposed onto subsequent welding control segments along the welding advance direction. During superposition, an attenuation coefficient is determined based on the circumferential distance between the current and subsequent welding control segments, and this attenuation coefficient is used to reduce the impact of the post-segment thermal memory on distant control segments.

[0060] For subsequent welding control segments based on the post-segment thermal memory, the target corridor generation module pre-corrects the target corridor of the molten pool or the initial welding control quantity based on the superimposed post-segment thermal memory. When the heat effect component in the post-segment thermal memory indicates that the actual welding heat input of the current segment is too high, the initial welding heat input of the subsequent welding control segment is reduced; when the forming deviation component in the post-segment thermal memory indicates that the current segment still has insufficient coverage, the initial amount of the molten pool boundary allowable band or wire feed of the subsequent welding control segment is compensated; when the forming deviation component in the post-segment thermal memory indicates that the current segment still has excessive coverage, the initial amount of the molten pool boundary allowable band or heat input of the subsequent welding control segment is converged, so that the welding result of the current welding control segment enters the pre-correction process of the subsequent control segment through the post-segment thermal memory, instead of only remaining in the instantaneous feedback control of the current control segment.

[0061] For the local abrupt change interval in the circumferential section control curve, the control segment generation module divides it into the transition welding control segment. The local abrupt change interval corresponds to the position where the area to be filled changes rapidly within a short circumferential range, including local bulges in tack welds, local abrupt changes in the bevel, or local misalignment positions.

[0062] Within the transition welding control section, the target corridor generation module generates a transition weld pool target corridor based on the weld pool target corridors corresponding to the welding control sections on both sides of the abrupt change interval. This transition weld pool target corridor allows the weld deposition centerline and weld pool boundary to change continuously within the abrupt change interval. Based on the transition weld pool target corridor, a gradual initial welding control quantity is generated. When the welding robot executes the transition welding control section, the welding torch posture correction and welding heat input correction quantities change in a limited manner according to the gradual initial welding control quantity, thereby avoiding sudden jumps in welding torch posture or heat input at locations of local morphological abrupt changes.

[0063] Example 2: The difference from Embodiment 1 is that this embodiment discloses a flange circular pipe circumferential weld quality control system for executing a flange circular pipe circumferential weld quality control method as described in Embodiment 1. The system includes: The cross-section model construction module is used to obtain pre-welding inspection data after the flange and round pipe fitting are clamped, establish the workpiece coordinate system of the circumferential weld, and construct a local cross-section filling model along the circumference of the circumferential weld. The control segment generation module is used to generate a circumferential section control curve based on the local section filling model, and to divide the welding control segment according to the circumferential section control curve; The target corridor generation module is used to generate the target corridor of the molten pool and the initial welding control quantity based on the local cross-section filling model corresponding to each welding control segment. The molten pool envelope extraction module is used to map the molten pool target corridor to the molten pool image coordinates, and extract the actual molten pool envelope based on the molten pool target corridor; The decoupling correction module is used to generate position deviation components and filling deviation components based on the envelope deviation of the actual molten pool envelope relative to the target corridor of the molten pool, and to generate welding torch attitude correction amount and welding heat input correction amount respectively. The thermal memory update module is used to generate the post-segment thermal memory amount based on the actual heat input and correction residual of the current welding control segment, and update the subsequent welding control segment based on the post-segment thermal memory amount.

[0064] In summary, this invention converts the pre-weld inspection data of the flange and round pipe fitting after clamping into a local cross-section filling model distributed along the circumference of the circumferential weld. By registering the local joint cross-section profile with the preset qualified weld cross-section, the cross-section area to be filled is obtained. This transforms the welding control basis from the traditional weld centerline or single assembly deviation into a cross-section filling object that can directly reflect the local deposition compensation requirements. This allows gaps, misalignments, bevel openings, or local morphological changes at different circumferential positions to enter the subsequent control process through the same cross-section area to be filled, avoiding the control fragmentation caused by establishing control rules separately for multiple isolated deviations. A circumferential cross-section control curve is generated based on the local cross-section filling model, and welding control segments are divided according to the continuous change range and local abrupt change range of the curve. This makes the welding control segments correspond to the circumferential change trend of deposition requirements. Compared with the method of segmenting by fixed angle or continuous welding by a single weld centerline, it can maintain continuous control at the local joint cross-section where the change is stable and form transition control at the local morphological abrupt change, thereby reducing the risk of unreasonable jumps in welding torch posture or heat input at abrupt change positions.

[0065] The target corridor of the molten pool is generated based on the section to be filled within the welding control segment, and mapped to the molten pool image coordinates as an image constraint region. This eliminates the reliance on brightness abrupt changes or single-frame boundary variations in the entire image for extracting the actual molten pool envelope during welding. Under the constraint of the target corridor, the actual molten pool envelope is formed by the temporal boundaries in continuous molten pool image frames, reducing interference from arc light, flares, and spatter highlights on the molten pool state judgment. This ensures that the image recognition results during welding correspond to the local filling requirements determined before welding. The actual molten pool envelope is compared with the target corridor, and the envelope deviation is decomposed into position deviation and filling deviation components. The position deviation component enters the welding torch posture correction chain, and the filling deviation component enters the welding heat input correction chain. This provides clear data sources and control division for welding torch pointing correction and molten deposit filling correction, thereby reducing the risk of erroneous corrections caused by simply adjusting the welding torch posture or process parameters uniformly based on molten pool anomalies.

[0066] Based on the actual heat input of the current welding control segment and the heat memory amount generated after the correction residual, the heat memory amount after this segment is superimposed on the subsequent welding control segment. This is used to pre-correct the target corridor of the molten pool or the initial welding control amount of the subsequent welding control segment. This allows the heat influence and forming deviation of the current segment to no longer remain in the instantaneous feedback control, but to enter the control basis of the subsequent segment along the welding advance direction, thereby forming a welding quality control chain that is continuously updated along the circumference of the circumferential weld.

[0067] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0068] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0069] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0070] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0071] The embodiments of the present invention have been described above with reference to the accompanying drawings. However, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of the present invention without departing from the spirit and scope of the claims. All of these forms are within the protection scope of the present invention.

Claims

1. A method for controlling the welding quality of circumferential welds on flanged pipes, characterized in that, include: Obtain pre-welding inspection data after clamping the flange and round pipe fitting, and establish a workpiece coordinate system for the circumferential weld. Multiple local joint cross-sectional profiles are cut along the circumference of the circumferential weld. Each local joint cross-sectional profile is registered with a preset qualified weld cross-section to obtain the cross-sectional area to be filled, and a local cross-sectional filling model is constructed. Based on the local section filling model, a circumferential section control curve is generated, and welding control segments are divided accordingly. Generate the target corridor of the molten pool and the initial welding control quantity based on the local cross-section filling model corresponding to each welding control segment; When welding the current welding control segment, the target corridor of the molten pool is mapped to the coordinates of the molten pool image, and the actual molten pool envelope is extracted; Based on the envelope deviation between the actual molten pool envelope and the target corridor of the molten pool, the position deviation component and the filling deviation component are decomposed and the welding gun attitude correction amount and the welding heat input correction amount are generated respectively. The heat memory value after the current welding control segment is generated based on the actual heat input and the correction residual, and the subsequent welding control segment is updated based on the heat memory value after the segment.

2. The welding quality control method for flange circular pipe circumferential welds according to claim 1, characterized in that, Multiple local joint cross-sectional profiles were cut along the circumference of the circumferential weld, including: The circumferential welding path is determined using the axis of the circular tube in the workpiece coordinate system of the circumferential weld. Multiple cross-sectional sampling positions are set along the circumferential welding path; At each of the aforementioned cross-sectional sampling locations, a local joint cross-sectional profile perpendicular to the welding advance direction is extracted; The local joint cross-sectional profile is formed by the intersection profile of the pre-welding inspection data on the corresponding cross-sectional plane, and includes the flange end profile, the round pipe end profile, and the bevel area profile.

3. The welding quality control method for flange circular pipe circumferential welds according to claim 2, characterized in that, The profiles of each local joint section are registered with the preset qualified weld section to obtain the section area to be filled, and a local section filling model is constructed, including: The preset qualified weld section is established based on the target weld forming size, weld reinforcement range, and the coverage boundaries on both sides of the weld. The coordinates of the preset qualified weld section and the local joint section profile are registered. Within the coverage area of ​​the preset qualified weld section, the area already occupied by the flange end, the round pipe end, and the bevel entity is deducted to obtain the section area to be filled; Based on the cross-sectional area to be filled, generate local cross-sectional filling units corresponding to the cross-sectional sampling positions; Multiple local cross-section filling units are arranged in circumferential order according to the sampling positions of the cross-section to form the local cross-section filling model.

4. The welding quality control method for flange circular pipe circumferential welds according to claim 3, characterized in that, Based on the local section filling model, a circumferential section control curve is generated, and welding control segments are divided accordingly, including: Extract the center point of the weld coverage from the section area to be filled in each local section filling unit; Connect multiple weld coverage center points in circumferential order according to the sampling positions of the cross section to generate a circumferential cross section control curve; The circumferential section control curve is smoothed. The circumferential interval where the circumferential section control curve changes continuously is divided into welding control segments, and the circumferential interval where the circumferential section control curve undergoes a local abrupt change and its adjacent intervals are divided into transition welding control segments.

5. The method for controlling the welding quality of the circumferential weld of a flanged circular pipe according to claim 1, characterized in that, Based on the local cross-section filling model corresponding to each welding control segment, the target corridor of the molten pool and the initial welding control quantities are generated, including: Read multiple local section filling elements within the current welding control segment; Generate the weld coverage centerline based on the weld coverage center point of multiple cross-sectional areas to be filled; Based on the distribution range of the cross-sectional area to be filled on both sides of the center line of the weld coverage, a permissible boundary zone for the weld pool is generated; The weld pool target corridor is defined by the weld coverage centerline and the weld pool boundary allowance band; The initial trajectory of the welding torch is generated based on the center line of the weld coverage, and the initial amount of welding heat input is generated based on the allowable zone of the weld pool boundary.

6. The method for controlling the welding quality of flange circular pipe circumferential welds according to claim 5, characterized in that, When welding the current welding control segment, the target corridor of the molten pool is mapped to the coordinates of the molten pool image, and the actual molten pool envelope is extracted, including: Based on the current position of the welding torch, the welding torch posture, and the calibration relationship of the vision acquisition device, the target corridor of the molten pool is mapped to the coordinates of the molten pool image to obtain the image constraint region; Acquire continuous molten pool image frames within the image constraint region; Extract candidate molten pool boundary points from the continuous molten pool image frames; Based on the spatial adjacency relationship and motion trend continuity of candidate molten pool boundary points in adjacent molten pool image frames, the candidate molten pool boundary points are sequentially connected. The stable boundary line or boundary region of the melt pool formed by the candidate melt pool boundary points that have completed the timing connection is determined as the actual melt pool envelope. Highlighted areas located outside the image constraint region, or areas with abrupt brightness changes that do not form temporally continuous boundaries, are excluded from the actual melt pool envelope.

7. The method for controlling the welding quality of the circumferential weld of a flanged circular pipe according to claim 1, characterized in that, Based on the envelope deviation between the actual molten pool envelope and the target corridor of the molten pool, position deviation components and filling deviation components are decomposed, and welding torch attitude correction amount and welding heat input correction amount are generated respectively, including: Multiple envelope comparison sections are set along the target corridor of the molten pool; The target center point of the target corridor of the molten pool is determined in each envelope comparison section, and the actual envelope center point is determined according to the actual molten pool envelope; Based on the offset of the actual envelope center point relative to the target center point, a position deviation component is generated; A filling deviation component is generated based on the amount of insufficient or excessive coverage of the actual molten pool envelope relative to the allowable zone of the molten pool boundary. When the positional deviation components in multiple envelope comparison sections characterize the overall translation of the actual molten pool envelope, a correction amount for the center point of the welding torch tool is generated. When the positional deviation component in multiple envelope comparison sections represents the angular deviation of the actual molten pool envelope relative to the target corridor of the molten pool, a welding torch angle correction amount is generated. Based on the filling deviation component, a welding current correction, a welding speed correction, or a wire feed speed correction is generated.

8. The method for controlling the welding quality of the circumferential weld of a flanged pipe according to claim 1, characterized in that, Based on the actual heat input of the current welding control segment and the corrected residual, a post-segment heat memory value is generated, and subsequent welding control segments are updated based on the post-segment heat memory value, including: The actual welding heat input of the current welding control section is determined based on the welding current, welding voltage, and welding speed within the current welding control section. The residual amount of envelope deviation is determined based on the position deviation component and fill deviation component that still exist at the end of the current welding control segment; A heat-affected component is generated based on the actual welding heat input, a forming deviation component is generated based on the envelope deviation residual amount, and a post-segment thermal memory amount is formed by the heat-affected component and the forming deviation component. The heat memory value after the segment is superimposed onto the subsequent welding control segment according to the attenuation relationship between the welding advance direction and the circumferential distance; Based on the superimposed post-segment thermal memory, the target corridor of the molten pool or the initial welding control quantity of the subsequent welding control segment is pre-corrected.

9. The method for controlling the welding quality of the circumferential weld of a flanged circular pipe according to claim 1, characterized in that, After dividing the circumferential section control curve where a local abrupt change occurs and its adjacent sections into transition welding control segments, the process further includes: Based on the target corridor of the molten pool corresponding to the welding control segments on both sides of the transition welding control segment, a target corridor of the transition molten pool is generated. A gradual initial welding control quantity is generated based on the target corridor of the transition molten pool; When welding the transition welding control section, the unit time variation range of the welding torch posture correction amount and the welding heat input correction amount is limited according to the gradual initial welding control amount.

10. A welding quality control system for circumferential welds of flanged pipes, characterized in that, include: The cross-section model construction module is used to obtain pre-welding inspection data after the flange and round pipe fitting are clamped, establish the workpiece coordinate system of the circumferential weld, and construct a local cross-section filling model along the circumference of the circumferential weld. The control segment generation module is used to generate a circumferential section control curve based on the local section filling model, and to divide the welding control segment according to the circumferential section control curve; The target corridor generation module is used to generate the target corridor of the molten pool and the initial welding control quantity based on the local cross-section filling model corresponding to each welding control segment. The molten pool envelope extraction module is used to map the molten pool target corridor to the molten pool image coordinates, and extract the actual molten pool envelope based on the molten pool target corridor; The decoupling correction module is used to generate position deviation components and filling deviation components based on the envelope deviation of the actual molten pool envelope relative to the target corridor of the molten pool, and to generate welding torch attitude correction amount and welding heat input correction amount respectively. The thermal memory update module is used to generate the post-segment thermal memory amount based on the actual heat input and correction residual of the current welding control segment, and update the subsequent welding control segment based on the post-segment thermal memory amount.

Images