A fine welding method for preparing a reinforcement cage based on a roll welding mechanism

Abstract

The present application relates to a kind of based on roll welding mechanism preparation steel reinforcement cage precision welding method, the method includes initial coiling data acquisition, welding execution, welding fusion determination, post-welding data acquisition, welding point judgment and repair welding step, by collecting the image of intersection point before and after welding, the relative position is extracted, combined with the thermal capacity judgment parameter obtained by thermal imaging, respectively generate offset determination score and welding determination score, fusion calculation virtual welding determination score to accurately identify virtual welding point.The present application constructs the whole process closed-loop control system of quality determination, geometric determination, dynamic repair welding, repair welding verification and to physical marker traceability, realizes accurate identification of virtual welding, efficient repair welding and quality whole link control, avoids traditional process defects, improves the overall welding quality and production stability of steel reinforcement cage, ensures that product meets the stringent acceptance standards of engineering, applicable to steel reinforcement cage roll welding large-scale production scene.

Description

Technical Field

[0001] This invention relates to the field of steel cage preparation and processing technology, and specifically to a precision welding method for preparing steel cages based on a roll welding machine. Background Technology

[0002] As a core load-bearing component in the construction engineering field, steel reinforcement cages are widely used in critical scenarios such as bridge pile foundations, subway pit support, and high-rise building foundation reinforcement. Their structural strength directly determines the bearing capacity and overall stability of the foundation engineering, and has a decisive impact on the quality and service life of the project. With the development of modern engineering construction towards large-scale and refined directions, the requirements for the precision and efficiency of steel reinforcement cage manufacturing are constantly increasing. Traditional manual binding or semi-automatic welding methods, due to their cumbersome procedures, large quality fluctuations, and long production cycles, can no longer meet the needs of large-scale construction. Automated manufacturing technology using roll welding machines has emerged and become the industry mainstream.

[0003] Currently, fully automated roll welding machines, driven by CNC systems, coordinate the feeding of longitudinal reinforcement bars, the winding of annular reinforcement bars, and the welding actuators to achieve standardized mass production of rebar cages, significantly improving manufacturing efficiency and dimensional accuracy. Its core working principle is as follows: multiple longitudinal reinforcement bars are evenly arranged and clamped along the circumference. A rotating mechanism drives the longitudinal reinforcement bars to rotate synchronously, while annular reinforcement bars are wound around the outside of the longitudinal reinforcement bars according to a preset pitch. When the annular reinforcement bars and longitudinal reinforcement bars form intersecting nodes, the welding torch of the welding machine continuously welds the intersecting nodes at fixed points, ultimately forming a cage-like structure. The application of this technology effectively reduces the intensity of manual operation, standardizes welding process parameters, and has become a key support for promoting the industrialization of rebar cage manufacturing.

[0004] However, in actual production, the welding quality of steel cages manufactured by roll welding machines still faces significant technical bottlenecks. Among these, the problem of incomplete welding caused by the misalignment of the intersection points of the annular and longitudinal reinforcement bars is particularly typical and has become a major hidden danger restricting the structural performance of the steel cage. Specifically, the welding actuators of existing roll welding machines typically adopt a fixed trajectory design, and the welding point of the welding torch is determined by the arrangement parameters of the longitudinal and annular reinforcement bars preset by the CNC system. Its positional accuracy depends on the precise alignment of the annular and longitudinal reinforcement bars. However, during processing and conveying, the ring reinforcement is prone to deviation due to the following multiple factors: First, ring reinforcement is mostly formed by cold bending using a rebar bending machine. During the bending process, the elastic-plastic rebound of the rebar material, the gap error of the bending die, and uneven stress can easily lead to a deviation between the actual radius of curvature of the ring reinforcement and the design value, resulting in morphological defects such as excessive ellipticity or local bulges. Second, if the ring reinforcement feeding channel of the rolling welding machine has problems such as excessive guide gap or uneven wear of the conveying rollers, it can cause the ring reinforcement to move axially or shift circumferentially during the winding process, making it unable to fit with the longitudinal reinforcement according to the preset trajectory. Third, if the longitudinal reinforcement has uneven circumferential distribution or axial positioning error when clamping and fixing, it will also indirectly cause the intersection node with the ring reinforcement to deviate from the theoretical position.

[0005] The aforementioned offset problem directly causes misalignment between the welding point and the actual intersection point: when the annular reinforcement shifts to one side, its actual contact point with the longitudinal reinforcement deviates from the preset welding trajectory of the welding torch, preventing the welding torch from effectively depositing fusion at the intersection, thus resulting in incomplete welding defects. These defects mainly manifest as insufficient weld penetration, incomplete weld filler, and weak bond between the weld and the reinforcing steel matrix. From a structural mechanics perspective, incomplete welding nodes cannot effectively transfer stress between the longitudinal and annular reinforcements, causing stress concentration at a few effective welding nodes when the reinforcing cage is subjected to axial pressure, bending moment, or shear force. This easily leads to node cracking, annular reinforcement loosening, and other failures. Engineering practice data shows that the overall load-bearing capacity of a reinforcing cage with incomplete welding defects can be reduced by more than 30%, failing to meet design requirements and potentially causing safety accidents such as pile fracture and foundation pit collapse during construction, or shortening the service life of the project by more than 50% due to fatigue damage during use, posing a serious threat to project safety.

[0006] To address the aforementioned issues, existing industry improvements primarily focus on enhancing the processing accuracy of ring reinforcements or optimizing the fixing structure of longitudinal reinforcements. Examples include using high-precision bending dies to control the forming error of ring reinforcements and adding positioning sensors to the clamping mechanism of longitudinal reinforcements. However, these methods only reduce the probability of misalignment at a single point and cannot fundamentally resolve the contradiction between fixed welding points and dynamic misalignment at intersections. When ring reinforcements or longitudinal reinforcements experience sudden misalignment due to uncontrollable factors, the welding mechanism at the fixed points still cannot respond and adjust in time, making incomplete welds difficult to avoid. Summary of the Invention

[0007] To address the shortcomings of existing technologies, the present invention aims to provide a precision welding method for preparing rebar cages based on a roll welding machine. This method addresses the issue of low rebar cage production quality due to the lack of welding quality inspection in the roll welding process. It constructs a closed-loop control system encompassing welding quality assessment, geometric position determination, dynamic and precise repair welding, repair welding verification, and physical marking traceability. This system enables accurate identification of incomplete welds at rebar cage weld points, efficient repair welding, and comprehensive quality control throughout the entire process, ultimately improving the overall welding quality of the rebar cages to meet stringent engineering acceptance standards.

[0008] This invention provides a precision welding method for preparing reinforcing cages based on a roll welding machine, the precision welding method comprising:

[0009] The initial coiling data acquisition step involves acquiring the first image to obtain the relative positions of the annular reinforcement and longitudinal reinforcement before welding, and then calling the corresponding welding parameters from a pre-established database based on the relative positions to generate welding instructions.

[0010] Visual image analysis improves the accuracy of welding parameter matching, and the first image provides basic data for post-weld positional deviation.

[0011] The welding execution steps involve the welding unit receiving welding instructions and executing the welding operation.

[0012] The welding fusion judgment step uses thermal imaging to obtain the heat capacity judgment parameters at the weld joint in order to obtain the welding judgment score;

[0013] This step assesses the fusion quality of the weld joint from the perspective of thermal properties, avoiding the problem that traditional visual inspection cannot identify internal non-fusion defects.

[0014] The post-weld data acquisition step involves acquiring a second image at the intersection of the annular reinforcement and the longitudinal reinforcement after welding, obtaining the relative position of the annular reinforcement and the longitudinal reinforcement based on the second image, and comparing it with the relative position of the first image to obtain an offset judgment score.

[0015] This step assesses the welding quality of the reinforcing bars from a geometric position perspective, overcoming the limitations of a single fusion judgment and achieving dual-dimensional detection from both welding thermal and geometric position perspectives.

[0016] The solder joint judgment step calculates a cold solder joint judgment score based on the welding judgment score and the offset judgment score. The cold solder joint judgment score represents the probability of a cold solder joint. If the cold solder joint judgment score is lower than a preset qualified threshold, the solder joint is a cold solder joint, and a re-soldering command is issued.

[0017] By using a dual-dimensional judgment score based on both the angle of welding thermal characteristics and the angle of the geometric position of the reinforcing bar, the accuracy of the judgment of incomplete welds is significantly improved.

[0018] The repair welding step calculates the repair welding trigger time based on the repair welding trigger time strategy, and controls the repair welding unit to perform the repair welding operation at the repair welding trigger time.

[0019] Furthermore, the welding fusion determination step includes a welding determination strategy, which specifically includes:

[0020] The heat capacity determination parameters include temperature distribution parameters, heat distribution profile parameters, and heat conduction continuity parameters. The heat distribution profile parameters reflect the heated areas of the annular ribs, welding molten material, and longitudinal ribs. The temperature distribution parameters reflect the temperature distribution within the heated area at the weld point. The heat conduction continuity parameters reflect the continuous change in heat within the heated area.

[0021] The welding judgment score is calculated based on each hot melt judgment parameter.

[0022] The parameters for judging hot melt in the welding fusion determination step are clearly defined as temperature distribution parameters, heat distribution profile parameters, and heat conduction continuity parameters, and the physical meaning reflected by each parameter is defined.

[0023] Furthermore, the initial winding data acquisition step includes a relative position acquisition strategy, which specifically includes:

[0024] The camera captures the first image of the intersection of the annular rib and the longitudinal rib before welding. The longitudinal rib centerline is obtained from the image based on the longitudinal rib outline, and the annular rib centerline is obtained based on the annular rib outline. The angle between the intersection point is obtained by the longitudinal rib centerline and the annular rib centerline. The angle between the intersection point reflects the relative angle between the annular rib and the longitudinal rib.

[0025] From the overlapping area of ​​the longitudinal rib contour and the annular rib contour, multiple line segments perpendicular to the central axis of the longitudinal rib and the central axis of the annular rib are obtained in the overlapping area. The line segments are the pixel diameters of the longitudinal rib and the annular rib. The average pixel diameter of the longitudinal rib and the average pixel diameter of the annular rib are calculated using the multiple line segments.

[0026] The positional relationship of the longitudinal rib relative to the first camera is obtained based on the ratio of the actual diameter of the longitudinal rib to the average pixel diameter of the longitudinal rib; the positional relationship of the annular rib relative to the first camera is obtained based on the ratio of the actual diameter of the annular rib to the average pixel diameter of the annular rib.

[0027] The front-to-back positional relationship between the longitudinal reinforcement and the camera is obtained by using the positions of the annular reinforcement and the camera.

[0028] When obtaining the pixel diameters of longitudinal and annular ribs in an image, the extraction range of the diameter is limited to the overlapping area to avoid interference from non-welded areas. Furthermore, a mean-averaging method is used instead of single-point measurement to reduce errors caused by contour distortion and improve the reliability of relative position parameters.

[0029] Furthermore, the location acquisition strategy includes a pixel filtering strategy, specifically:

[0030] Compare the slope fluctuation value of each pixel in the longitudinal rib contour with the fluctuation threshold, and remove pixels that are less than the fluctuation threshold; compare the curvature value of each pixel in the annular rib contour with the curvature threshold, and remove pixels that are less than the curvature threshold.

[0031] Perform connected component analysis on the remaining pixels after removal, and remove pixels in isolated connected components;

[0032] The central axes of the longitudinal and annular ribs are obtained by fitting the pixels in the connected domain along the specified directions of the longitudinal and annular ribs.

[0033] By employing a pixel-based filtering strategy, noise points with abnormal slope fluctuations and isolated connected regions caused by corrosion or dust are identified in the contour region. This retains the effective contour pixels of the steel reinforcement body, thereby improving the accuracy of the centerline fitting and ensuring the calculation accuracy of core parameters such as intersection angle and pixel diameter.

[0034] Furthermore, the post-weld data acquisition step is configured with an offset determination strategy, which is as follows:

[0035] The relative angles of the ring reinforcement and longitudinal reinforcement are obtained from the second image by using a relative position acquisition strategy. The relative angle deviations are then obtained by comparing them with the relative angles of the ring reinforcement and longitudinal reinforcement in the first image.

[0036] The positional deviation is obtained by comparing the front and rear positions of the annular reinforcement and longitudinal reinforcement obtained in the second image with the front and rear positions obtained in the first image.

[0037] The offset judgment score is obtained by quantifying the relative angle deviation and position deviation.

[0038] By using an offset determination strategy to form weld points for longitudinal and annular bars, the changes in relative position before and after welding are quantified from two dimensions: the relative angle and the front and back positions of the two bars. This analysis helps to determine whether there are any issues with welding quality caused by the offset of their positions.

[0039] Furthermore, the solder joint determination step is also equipped with a simplified determination strategy. The simplified determination strategy is triggered after the execution conditions are met. The execution conditions are: when the score of the cold solder joint determination is not lower than the preset qualified threshold, the solder joint is a qualified solder joint, the number of consecutive times that the solder joint is determined to be a qualified solder joint is recorded, and the number of consecutive times exceeds the preset number threshold.

[0040] The simplified execution strategy is as follows: calculate the welding judgment score based on the hot melt judgment parameter; if the welding judgment score is lower than the judgment threshold, the weld point is a false weld point.

[0041] When the simplified judgment strategy determines that the solder joint is a dummy solder joint, the relative position on the first image obtained at this time is compared with the relative position in the previous first image. If the similarity is lower than the threshold, the dummy solder joint judgment score is recalculated to determine the solder joint.

[0042] By designing a simplified judgment strategy, simplified judgment logic is executed after consecutive weld points are judged as qualified, thereby reducing the amount of calculation for weld point judgment and improving the detection efficiency of the production line. Furthermore, a position similarity composite mechanism is set up in the simplified judgment process to avoid the simplified judgment of a false weld being caused by a sudden change in the position of the reinforcing bar, thus balancing efficiency and accuracy.

[0043] Furthermore, both the initial winding data acquisition step and the post-weld data acquisition step include an image preprocessing strategy, which specifically includes:

[0044] Multiple images with different exposures are continuously acquired for the same solder joint. Effective feature points are extracted from each image, and the fused image is obtained by feature point matching and fusion.

[0045] Identify bright spot areas in the fused image, replace the bright spot pixels with normal pixel values ​​around the bright spot, and complete the removal and repair of reflective bright spots;

[0046] The fused image with reflective bright spots removed is filtered to eliminate the image blurring area caused by thermal fog, and edge feature points are enhanced by edge sharpening to obtain the processed first image and second image.

[0047] Image preprocessing strategies are employed to preprocess the acquired first and second images. Multi-exposure fusion enhances the dynamic range of the images, corrects reflections and thermal fog defects, ensures the clarity of contour features, and edge sharpening enhances the contour feature points of the steel bars, providing a high-quality image foundation for subsequent pixel extraction and centerline fitting.

[0048] Furthermore, the welding repair step is further configured with a welding repair triggering timing strategy, specifically as follows:

[0049] Obtain the historical circumferential angle at the second image acquisition moment;

[0050] The real-time rotational angular velocity of the ring rib is obtained by the encoder, as well as the delay time from the image acquisition time to the command issuance time. The circumferential angle compensation amount is calculated based on the real-time rotational angular velocity and the delay time.

[0051] The real-time circumferential angle is obtained by summing the historical circumferential angle with the circumferential angle compensation.

[0052] Based on the target circumferential angle and real-time circumferential angle in the coordinates of the repair area, the estimated time for the dummy weld point to reach the repair area is calculated.

[0053] The welding trigger time is obtained by summing the time when the welding repair command is issued with the estimated time.

[0054] This strategy addresses the impact of equipment delay by calculating the angle compensation amount, enabling precise control of the welding timing. This adapts to the continuous production characteristics of the roll welding machine, eliminating the need for machine downtime for welding and improving production efficiency.

[0055] Furthermore, the welding repair step is also equipped with a welding repair verification strategy, which specifically includes:

[0056] By assembling an ultrasonic transmitter next to the welding unit, ultrasonic waves are emitted from the transmitting end of the ultrasonic transmitter toward the welding point of the annular rib and the longitudinal rib. If the reflected wave value received by the receiving end is lower than the preset threshold, the welding point is determined to be a qualified welding point.

[0057] If the received reflected wave value is higher than the preset threshold, the solder joint is determined to be an unqualified solder joint, and a marking strategy is executed.

[0058] After the repair welding, ultrasonic testing is used for re-inspection to identify whether the welding quality of the repaired weld points is up to standard. Unqualified repaired weld points are marked to facilitate subsequent manual review and processing, thereby improving the final product quality of the steel cage.

[0059] Furthermore, the labeling strategy specifically includes:

[0060] When a weld repair point is determined to be an unqualified weld repair point, a marking instruction is sent to the physical marking unit;

[0061] The marking trigger time is obtained by summing the time when the marking command is issued with the time when the unqualified repair weld point is moved to the physical marking unit. The physical marking unit is then controlled to perform the marking operation on the unqualified repair weld point at the trigger time.

[0062] By using a marking strategy to accurately calculate the marking trigger time, the marking position is ensured to correspond perfectly with the defective solder joint. Furthermore, the material marking unit enables the visual marking of defective solder joints, facilitating rapid identification and processing in subsequent processes.

[0063] The beneficial effects of this invention are:

[0064] The welding method proposed in this invention integrates a two-dimensional judgment logic of thermal imaging fusion characteristic detection and geometric position deviation detection before and after welding. Combined with pixel point screening strategy and image preprocessing strategy optimization, it not only solves the problem that traditional appearance inspection cannot identify internal false welds, but also avoids the one-sidedness of single convenient judgment, greatly improves the accuracy of false weld identification, and reduces the structural safety hazards of steel cages caused by weld defects.

[0065] This invention automatically matches suitable welding parameters based on image analysis before welding, standardizes welding execution during welding, and determines and performs repair welding based on post-weld images and precise calculation of the repair welding trigger time. The entire process requires no manual intervention. At the same time, a simplified judgment strategy is set up to reduce the amount of calculation for determining the weak weld while ensuring judgment accuracy. It is adapted to the production characteristics of continuous rotation operation of the roll welding machine, realizes closed-loop automation from welding to detection of repair welding, and significantly improves the production efficiency of rebar cages.

[0066] This invention adds a secondary verification step after welding repair. Unqualified weld points are precisely located using physical markings, and the target weld points can be quickly identified during manual re-inspection, improving positioning efficiency. At the same time, it constructs a closed-loop process from detection, welding repair, verification to marking, ensuring traceability of welding repair quality and rapid identification and handling of unqualified weld points, thereby improving the standardization level and product quality stability of the entire steel cage welding process. Attached Figure Description

[0067] Figure 1 This is a front view schematic diagram of the structure of the rebar cage welding machine for the present invention, showing the positional distribution of each working unit in the welding machine.

[0068] Figure 2 This is a side view of the structure of the roll welding machine for preparing the reinforcing cage according to the present invention, showing the process of welding the ring bars and longitudinal bars to form the reinforcing cage.

[0069] Figure 3 This is a flowchart illustrating the precision welding method for preparing reinforcing cages based on a roll welding machine, as presented in this invention. It shows the relationship between the determination of incomplete welds and the repair welding during the entire welding process.

[0070] The attached figures are labeled as follows:

[0071] 1. Roll welding machine; 101. Ring reinforcement; 102. Longitudinal reinforcement; 103. Reinforcing cage; 2. Welding unit; 3. First camera; 4. Thermal imager; 5. Second camera; 6. Repair welding unit; 7. Ultrasonic transmitter; 8. Physical marking unit. Detailed Implementation

[0072] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0073] It should be noted that when a component is described as "fixed to" another component, it can be directly on the other component or may have a component in between. When a component is considered "connected to" another component, it can be directly connected to the other component or may have a component in between. When a component is considered "set on" another component, it can be directly set on the other component or may have a component in between. The terms "vertical," "horizontal," "left," "right," and similar expressions used in this document are for illustrative purposes only.

[0074] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0075] This invention provides a precision welding method for preparing reinforcing cages using a roll welding machine. This method is primarily aimed at the roll welding equipment used in preparing reinforcing cages, such as… Figure 1 and Figure 2As shown, the forming roller of the rolling welding machine 1 has positioning slots for longitudinal ribs 102 distributed circumferentially on the outer ring of the forming roller. Multiple longitudinal ribs 102 of the same length are placed into the positioning slots of the rolling welding machine 1 one by one, and the longitudinal ribs 102 are fixed by clamps set at both ends of the positioning slot opening. The clamps can be pneumatic clamps composed of clamping cylinders and arc-shaped clamps to ensure that all longitudinal ribs 102 remain parallel and perpendicular to the main shaft of the equipment. An axial feed drive assembly is set on the rear side of the forming roller to drive all longitudinal ribs 102 to move forward synchronously. As the forming roller continues to rotate and the longitudinal ribs 102 move forward at a uniform speed, the annular ribs 101 are continuously wrapped around the outside of the longitudinal ribs 102, completing the initial forming of the reinforcing cage 103.

[0076] After one end of the annular rib 101 passes through the guide wheel for guiding traction, it is welded and fixed to the longitudinal rib 102 through the welding unit 2 to form a traction anchor point. Therefore, during the continuous rotation of the forming roller, the rotation of the forming roller will generate a continuous radial traction force on the annular rib 101, realizing the passive feeding of the annular rib 101. The annular rib 101 is gradually bent along the inner arc surface of the bending die, so that the steel bar gradually forms a circle. The intersection of the annular rib 101 and the longitudinal rib 102 is welded and fixed by the welding unit 2.

[0077] Furthermore, the forming roller of the roll welding machine 1 is equipped with a repair welding unit 6 and a physical marking unit 8. The repair welding unit 6 performs repair welding on the points determined to be poor welds, and the physical marking unit 8 is used to perform inkjet operation to mark the unqualified repair welds. In addition, two sets of image acquisition units are set on the circumference of the forming roller. The image acquisition units complete the image acquisition at the welding points of the longitudinal rib 102 and the annular rib 101.

[0078] The two sets of image acquisition units mentioned above have the same structure, including a first camera 3, a second camera 5, a high-frequency pulse ring light source, and an anti-arc light isolation cover. It should be noted that the first camera 3 and the second camera 5 should be industrial cameras with anti-ambient light interference function, high-precision imaging, high resolution, and adaptability to harsh environments. For example, RVC-M series industrial cameras or EpicEye series industrial cameras can be selected.

[0079] The first camera 3 is mounted and fixed on the side of the welding unit 2 in the roll welding machine 1 via a frame. It is necessary to ensure that the camera lens is facing the intersection area of ​​the annular rib and the longitudinal rib. The first camera 3 is used to acquire the first image at the intersection point, which is the image before welding. The second camera 5 is also fixed on the frame, and the camera lens is also ensured to be facing the intersection area of ​​the annular rib and the longitudinal rib. The second camera 5 is used to acquire the second image at the intersection point, which is the image after welding. In addition, a high-frequency pulsed ring light source is installed around the two cameras for image acquisition supplementary lighting, and an anti-arc light isolation cover is set, which is an arc-shaped metal cover to isolate welding arc light interference.

[0080] A thermal imager 4 is mounted on the periphery of the forming roller and located behind the welding unit 2 via a fixing frame. The thermal imager 4 is used to obtain thermal imaging images of the intersection of the longitudinal rib 102 and the annular rib 101 after welding. It should be noted that the thermal imager 4 should be an infrared thermal imager designed specifically for high temperature and dynamic welding environments to cope with strong arc light interference, wide temperature range and rapid temperature changes. For example, a FLIR Ex-XT series infrared thermal imager or a Gewu Youxin X38H series infrared thermal imager can be used.

[0081] An encoder group is configured in the roll welding machine 1 and installed at the drive shaft end of the forming roller of the roll welding machine 1. The drive shaft end drives the imaging roller of the roll welding machine 1 to rotate. The encoder group is used to collect the real-time rotational angular velocity and rotation angle of the rebar cage.

[0082] Specifically, welding unit 2 is a resistance welding unit, fixed on the frame and corresponding to the overlap position of the longitudinal rib 102 and annular rib 101 of the active forming roller. It includes a spot welding electrode head, an electrode drive cylinder, a welding power supply, and a control module. The spot welding electrode head is made of chromium zirconium copper and has an electrode wheel structure. The electrode drive cylinder is used to drive the spot welding electrode head to press the overlap of the annular rib 101 and the longitudinal rib 102. The control module is linked with the encoder group, and the electrode wheel makes rolling contact with the steel bar to form a continuous or intermittent welding circuit, which perfectly matches the dynamic welding requirements of the annular rib and longitudinal rib when the steel cage is rotated and formed. It should be noted that the repair welding unit 6 has the same structure as welding unit 2. Its installation position is also along the rotation direction of the forming roller, arranged behind welding unit 2, and connected to the image acquisition unit to receive operation commands.

[0083] The physical marking unit 8 can be an industrial inkjet marking unit, fixed on the frame and located behind the welding unit 6. It needs to be adapted to the production characteristics of continuous rotation of the steel cage. This unit includes a high-temperature resistant inkjet printhead, an ink storage tank, a pneumatic pressurization component and a trigger controller. The nozzle diameter of the inkjet printhead is 0.1~0.3mm, and the distance between the printhead and the surface of the steel cage is controlled at 5~10mm. The pneumatic pressurization component is used to provide stable inkjet pressure. The trigger controller is connected to the image acquisition unit to receive operation instructions and is used to apply visual markings to unqualified welding points.

[0084] An ultrasonic transmitter 7 is installed behind the welding repair unit 6. It is used to verify whether the welding repair point is qualified. The ultrasonic transmitter 7 is mounted on the side of the welding repair unit 6 by a fixed bracket. After the ultrasonic transmitter 7 is installed, the corresponding position should be adjusted to ensure that the center line connecting the transmitting end and the receiving end is perpendicular to the welding repair point.

[0085] like Figure 3As shown, in one embodiment of the present invention, the present invention also proposes a precision welding method for preparing reinforcing cages based on a roll welding machine 1, the specific contents of which are as follows:

[0086] The initial coiling data acquisition step involves acquiring a first image to obtain the relative positions of the annular rib 101 and the longitudinal rib 102 before welding, and then calling the corresponding welding parameters from a pre-established database based on the relative positions to generate welding instructions.

[0087] The initial coiling data acquisition step is a preparatory step for the entire precision welding method. Its core purpose is to provide a reliable basis for matching subsequent welding parameters by accurately collecting the relative position information of the annular rib 101 and the longitudinal rib 102 before welding, and at the same time establish a benchmark for comparing the positions before and after welding.

[0088] Among them, the annular rib 101 refers to the steel reinforcement member distributed in a ring in the steel cage 103, used to wrap around the outside of the longitudinal rib 102 to form a cage-like structure with the longitudinal rib 102; the longitudinal rib 102 refers to the steel reinforcement member distributed along the length direction of the steel cage 103; the relative position refers to the relative angular relationship between the annular rib 101 and the longitudinal rib 102 in space and their front-to-back position relationship; the first image refers to the two-dimensional image of the intersection of the annular rib 101 and the longitudinal rib 102 before welding, which is captured by the first camera 3 installed at a designated position on the roll welding machine 1. In actual application, the first camera 3 is fixed in front of the welding unit 2 in the roll welding machine 1 to ensure that the camera lens is facing the intersection area of ​​the annular rib 101 and the longitudinal rib 102. In addition, a high-frequency pulse ring light source is configured to provide supplementary lighting to avoid the influence of ambient light on image quality before welding.

[0089] The pre-establishment of the database in this step refers to the construction of a correlation database containing different relative position parameters and optimal welding parameters through process experiments and data accumulation before the mass production of the steel cage 103. The welding parameters include process parameters that directly affect the welding quality, such as welding current, welding voltage, welding time, and welding torch pressure. The welding command refers to the standardized operation signal generated by the control system based on the database matching results, which can be recognized and executed by the welding unit 2.

[0090] After acquiring the first image, an image preprocessing strategy needs to be executed to preprocess the first image. The specific steps are as follows:

[0091] Multiple images with different exposures are continuously acquired at the same intersection of the annular rib 101 and the longitudinal rib 102. For example, three images are acquired, and the exposures are set to 100ms, 150ms and 200ms respectively. The contour feature points of the annular rib 101 and the longitudinal rib 102 in each image are then extracted. The three images are then fused into a single fused image using a feature point matching algorithm.

[0092] Then, identify the bright spot areas on the fused image caused by the reflection of the steel bar surface, and replace the bright spot pixels with the average value of normal pixels within a 5×5 pixel range around the bright spot to complete the repair of the reflective bright spot;

[0093] The repaired fused image is then filtered using a Gaussian filtering algorithm. The filter kernel size can be set to 3×3 to eliminate the blurred areas caused by thermal fog. Finally, the edge sharpening is performed using the Laplacian operator to enhance the contour feature points of the annular rib 101 and the longitudinal rib 102, resulting in the first image after processing.

[0094] Based on the processed first image, the relative positions of the annular rib 101 and the longitudinal rib 102 are extracted. First, a pixel selection strategy is executed, the specific content of which is as follows:

[0095] Each pixel in the contour of the longitudinal rib 102 is taken as the analysis object. The slope of the fitted line in its 3×3 neighborhood is calculated for each pixel. Then, the difference between this slope and the average slope of the fitted lines of all pixels in the neighborhood is calculated. This difference is the slope fluctuation value. The fluctuation threshold can be set to 3°. This fluctuation threshold can be determined through process experiments. Pixels with slope fluctuation values ​​less than 3° are judged as noise points and removed. The slope fluctuation value refers to the difference between the local fitted line slope of a pixel in the contour of the longitudinal rib 102 and the average fitted line slope of the local area where the pixel is located. It is used to measure the slope stability of the pixel.

[0096] Since the annular rib 101 is an arc-shaped component, the curvature of its contour pixels should be kept within a specific range. Pixels with excessively small curvature values ​​indicate that the position is close to a straight line, which does not conform to the arc characteristics of the annular rib 101 and is considered noise points that need to be removed. Therefore, each pixel in the contour of the annular rib 101 is taken as the analysis object, and the curvature value of the fitted curve of each pixel in its 3×3 neighborhood is calculated and compared with the curvature threshold. Pixels with curvature values ​​less than the curvature threshold are defined as noise points and removed. The curvature threshold is also set by standard specimen calibration.

[0097] Subsequently, connected component analysis was performed on the pixels after initial removal. A connected component is a region in an image that consists of adjacent pixels with the same gray value. For example, isolated connected components with an area of ​​less than 50 pixels were identified as impurities and interference. Isolated connected components refer to areas such as rust or dust on the surface of steel bars. The two connected components with the largest areas were retained, corresponding to the contours of longitudinal reinforcement 102 and annular reinforcement 101, respectively.

[0098] Based on the pixels in the retained connected components, the central axis of the longitudinal rib 102 is obtained by fitting along the length direction of the longitudinal rib 102, and the central axis of the annular rib 101 is obtained by fitting along the tangent direction of the circular arc of the annular rib 101. The central axis of the longitudinal rib 102 refers to a straight line that can reflect the center position and direction of the longitudinal rib 102, and the central axis of the annular rib 101 refers to a straight line that can reflect the center position and tangent direction of the local circular arc of the annular rib 101.

[0099] After filtering based on the first image pixels, the central axes of the longitudinal rib 102 and the annular rib 101 are obtained. The relative angles and positional relationships between the longitudinal rib 102 and the annular rib 101 are then acquired, specifically:

[0100] First, the angle between the longitudinal reinforcement 102 and the annular reinforcement 101 is calculated to obtain the angle at the intersection point. This angle at the intersection point is the relative angle between the annular reinforcement 101 and the longitudinal reinforcement 102.

[0101] Subsequently, the pixel diameter is extracted in the overlapping area of ​​the contours of the longitudinal rib 102 and the annular rib 101. The overlapping area refers to the area in the first image where the annular rib 101 and the longitudinal rib 102 intersect and their contours overlap.

[0102] First, five sampling points are uniformly selected along the length of the central axis of the longitudinal rib 102 within the overlapping area, with a distance of 20 pixels between adjacent sampling points. With each sampling point as the center, a line segment perpendicular to the central axis of the longitudinal rib 102 is drawn. The pixel distance between the two intersection points of this line segment and the contour of the longitudinal rib 102 is the local pixel diameter of the longitudinal rib 102 at that sampling point. Similarly, four sampling points are uniformly selected along the tangent direction of the central axis of the annular rib 101 within the overlapping area. With each sampling point as the center, a line segment perpendicular to the central axis of the annular rib 101 is drawn. The pixel distance between the two intersection points of this line segment and the contour of the annular rib 101 is the local pixel diameter of the annular rib 101 at that sampling point. The average value of the local pixel diameters of the five longitudinal ribs 102 and the average value of the local pixel diameters of the four annular ribs 101 are calculated respectively to obtain the average pixel diameter of the longitudinal rib 102 and the average pixel diameter of the annular rib 101.

[0103] By calculating the average pixel diameter of the longitudinal rib 102 and comparing it with the known actual diameter of the longitudinal rib 102, the ratio of the actual diameter of the longitudinal rib 102 to the average pixel diameter of the longitudinal rib 102 can be obtained. This ratio reflects the actual length corresponding to each pixel in the image. Based on this ratio, the distance relationship between the longitudinal rib 102 and the first camera 3 can be obtained.

[0104] Similarly, by obtaining the pixel diameter of the annular rib 101 and the actual diameter of the annular rib 101, the ratio of the actual diameter of the annular rib 101 to the average pixel diameter of the annular rib 101 is obtained, and the distance relationship of the annular rib 101 relative to the first camera 3 is obtained based on this ratio.

[0105] By comparing the ratio of the longitudinal reinforcement 102 with the ratio of the annular reinforcement 101, it can be determined that the annular reinforcement 101 is closer to the camera and located in front of the longitudinal reinforcement 102. Furthermore, by comparing the two ratios, the difference reflects the relative distance between the annular reinforcement 101 and the longitudinal reinforcement 102. For example, if the ratio of the longitudinal reinforcement 102 is 0.268 and the ratio of the annular reinforcement 101 is 0.248, the difference between the two reinforcement ratios is 0.02.

[0106] It should be noted that the spatial order is derived by utilizing the imaging characteristic that pixel diameter is inversely proportional to actual distance. When the camera is fixed, the closer a steel bar of the same specification is to the camera, the larger the pixel diameter appears in the image; the farther away it is, the smaller the pixel diameter. Given that the actual diameter is known, the ratio of the actual diameter to the average pixel diameter (i.e., the actual length corresponding to a unit pixel) can indirectly reflect the object distance: the larger the ratio, the longer the actual length corresponding to a unit pixel, and the farther the steel bar is from the camera; the smaller the ratio, the shorter the actual length corresponding to a unit pixel, and the closer the steel bar is to the camera.

[0107] Using the relative angle and front-to-back position relationship of the longitudinal rib 102 and the annular rib 101 obtained from the first image analysis as retrieval parameters, matching welding parameters are searched from a pre-established database. For example, the database stores a relative angle of 45° and a difference of 0.02 between the ratio of the longitudinal rib 102 and the ratio of the annular rib 101. The corresponding welding parameters are welding current 180A, welding voltage 28V, welding time 0.8s, and welding torch pressure 0.3MPa. The control system then generates welding instructions containing these parameters, completing the initial coiling data acquisition step. This step matches the corresponding welding parameters through visual image analysis, without relying on human experience judgment, which greatly improves the accuracy of welding parameter matching. At the same time, the relative position data extracted from the first image serves as benchmark data, providing a direct comparative basis for subsequent post-weld position deviation analysis.

[0108] The welding execution step is the process of converting welding instructions into actual welding operations. In this step, welding unit 2 receives the welding instructions and executes the welding operations. Welding unit 2 is a resistance welding unit. When welding unit 2 receives the welding instructions sent by the control system, the electrode wheel outputs a welding current of 180A and a welding voltage of 28V according to the parameters set in the welding instructions, and continues to supply power for 0.8s. At the same time, the spot welding electrode head presses the intersection with a pressure of 0.3MPa, so that the annular rib 101 and the longitudinal rib 102 achieve fusion welding under high temperature. During the welding process, the action of welding unit 2 is synchronized with the rotation and feeding action of the roller welding machine 1 to ensure that each intersection can be welded according to the preset parameters.

[0109] The welding fusion judgment step involves acquiring heat capacity judgment parameters at the weld joint through thermal imaging to derive a welding judgment score. This step evaluates the fusion quality of the weld joint from a thermal characteristic perspective. Thermal imaging refers to the technology of detecting infrared radiation at the weld joint using a thermal imager and converting the temperature distribution into a visual image. In practical applications, the thermal imager is installed beside welding unit 2, with its detection range covering the weld joint and a surrounding 20mm area. The detection frequency is set to 10 frames per second. During the welding process, the thermal imager acquires thermal imaging data of the weld joint in real time.

[0110] The heat capacity judgment parameter refers to a quantitative indicator that reflects the heat distribution and conduction characteristics at the weld joint, including temperature distribution parameters, heat distribution profile parameters, and heat conduction continuity parameters. Specifically, the temperature distribution parameter reflects the temperature distribution within the heated area at the weld joint, specifically the ratio of the highest temperature to the average temperature within the heated area. The heat distribution profile parameter reflects the heated areas of the annular reinforcement 101, the welding molten material, and the longitudinal reinforcement 102, specifically the ratio of the area of ​​the heated area to the preset standard heated area. Here, the welding molten material refers to the molten steel reinforcement material and any welding materials added during the welding process. The heat conduction continuity parameter reflects the continuous change of heat within the heated area, specifically the average of the absolute values ​​of the rate of change of the central temperature of the heated area at two adjacent moments.

[0111] The welding judgment score is a numerical value calculated based on the heat capacity judgment parameter to quantify the fusion quality of the weld joint. The value ranges from 0 to 100 points. The higher the score, the better the fusion quality. It can effectively identify internal non-fusion defects that cannot be detected by traditional visual inspection. For example, if there is a non-fusion area inside the weld joint, it will cause the heat distribution profile parameter to be small and the heat conduction continuity parameter to be abnormal, which will reduce the welding judgment score.

[0112] For the aforementioned welding judgment score, a weighted summation method can be used for calculation. The influence weights of the three types of heat capacity judgment parameters on fusion quality are determined through process experiments, and the welding judgment score is calculated using a linear weighted summation formula. The calculation formula is as follows:

[0113] ,

[0114] in, The score for welding assessment is limited to 0-100 points. If the calculated result is >100, then 100 is taken; if the calculated result is <0, then 0 is taken. These are temperature distribution parameters; These are the thermal distribution profile parameters; For continuous parameters of heat conduction; Weights for temperature distribution parameters; Weights for the heat distribution profile parameters; Let the weights be the continuous parameters of heat conduction, and the weights satisfy... + + =1.

[0115] Specifically, temperature distribution parameters The calculation formula is:

[0116] ,

[0117] in, The highest temperature in the heated area of ​​the solder joint is obtained by searching the pixel temperature of the thermal imager. The average temperature within the heated area of ​​the solder joint is obtained by statistically averaging the temperatures of all pixels in the heated area using a thermal imager.

[0118] Thermal distribution profile parameters The calculation formula is:

[0119] ,

[0120] in, The actual heated area is defined as the area above a set temperature threshold in the thermal imager, which is calculated using the pixel area. The standard heated area is determined based on the specifications of the reinforcing bars. For example, the standard heated area of ​​the weld joint between the 20mm diameter longitudinal bar 102 and the 16mm diameter ring bar 101 is set to 120mm².

[0121] Thermal conductivity continuity parameter The calculation formula is:

[0122] =1- ,

[0123] in, The average absolute value of the rate of change of center temperature between adjacent time points is used to calculate the rate of change between adjacent frames. The sampling frequency of the thermal imager is set to 10 frames / second. Ten consecutive thermal images are selected during the welding process, and the center temperature of the weld point in each frame is extracted to calculate the rate of change between adjacent frames. The preset standard temperature change rate is determined by process experiments, and this preset value is the maximum allowable change rate for stable heat conduction.

[0124] After analyzing the fusion characteristics of the weld joint through thermal imaging, the post-weld data acquisition step is initiated. This step evaluates the welding quality from a geometric position perspective. The second camera 5 is used to acquire a second image of the intersection of the annular rib 101 and the longitudinal rib 102 after welding. Based on the second image, the relative positions of the annular rib 101 and the longitudinal rib 102 are obtained and compared with the relative positions of the longitudinal rib 102 and the annular rib 101 obtained in the first image to obtain an offset judgment score. The offset judgment score is a numerical value of the degree of positional deviation obtained by comparing the relative positions of the annular rib 101 and the longitudinal rib 102 in the first and second images. The value range is 0-100 points. The higher the score, the smaller the positional deviation and the better the welding quality.

[0125] After acquiring the second image, it is necessary to preprocess the image, filter the pixels on the image, and obtain the relative positions of the longitudinal rib 102 and the annular rib 101. The entire execution process is the same as the image acquisition and preprocessing process in the initial coiling data acquisition step. The second camera 5 takes a second image of the intersection after welding. It also goes through the same image preprocessing steps, including multi-exposure fusion, reflection spot repair, filtering and deblurring, and edge sharpening, to obtain a clear second image. Then, using the same pixel filtering strategy and centerline fitting method as in the initial coiling data acquisition step, the relative angle and front-back position relationship of the annular rib 101 and the longitudinal rib 102 are extracted from the second image.

[0126] Based on the offset determination strategy configured in the post-weld data acquisition step, the relative angle obtained in the second image is compared with the relative angle of the annular rib 101 and the longitudinal rib 102 in the first image to obtain the relative angle deviation. For example, if the relative angle obtained in the second image is 46° and the relative angle obtained from the first image is 45°, then the relative angle deviation can be calculated to be 1°.

[0127] By comparing the front and rear positions of the annular reinforcement 101 and the longitudinal reinforcement 102 obtained in the second image with the front and rear positions obtained in the first image, the positional deviation is obtained. For example, if the difference between the ratio of the longitudinal reinforcement 102 and the ratio of the annular reinforcement 101 obtained in the second image is 0.032, and the difference between the ratio of the longitudinal reinforcement 102 and the ratio of the annular reinforcement 101 obtained in the first image is 0.02, then the deviation in the degree of change in the distance between the front and rear positions of the two reinforcement bars is 0.012.

[0128] The offset judgment score is then obtained by quantifying the relative angle deviation and position deviation. The offset judgment score can also be calculated using a linear weighted scoring method.

[0129] The formula for calculating the offset determination score is:

[0130] ,

[0131] in, The offset determination score is constrained to be between 0 and 100. The relative angle deviation is obtained from the difference between the relative angle in the second image and the relative angle in the first image; This is a preset value to represent the maximum allowable relative angular deviation. The difference between the ratio difference of the second image and the ratio difference of the first image is obtained; The maximum allowable positional deviation is a preset value; The relative angle deviation weight, and Let be the position deviation weight, and satisfy . + =1.

[0132] This step, by comparing the relative positions of the longitudinal rib 102 and the annular rib 101 before and after welding, geometrically assesses whether the positional shift of the annular rib 101 and the longitudinal rib 102 is caused by factors such as thermal deformation and welding stress during the welding process. It overcomes the limitation of a single fusion judgment that can only assess the internal fusion state, and realizes dual-dimensional detection of thermal properties and geometric position.

[0133] By analyzing the fusion state of the weld joint and determining the geometric positional relationship between the two reinforcing bars, the weld joint is analyzed to determine whether it is a false weld. The weld joint judgment step is then initiated. This step is the core link for comprehensively evaluating the false weld situation and issuing a repair welding instruction. Specifically, the false weld judgment score is calculated based on the welding judgment score and the offset judgment score. The false weld judgment score represents the probability of a false weld at the weld joint. If the false weld judgment score is lower than the preset qualified threshold, the weld joint is a false weld joint, and a repair welding instruction is issued.

[0134] Among them, the cold weld judgment score refers to the fusion weld judgment score and the offset judgment score, which quantifies the probability of cold welds. The value range is 0-100 points, and the higher the score, the lower the probability of cold welds. The preset qualified threshold is the critical value set according to the engineering quality requirements to judge whether a weld is a cold weld, which is usually set to 80 points. A cold weld is a weld where the ring rib 101 and the longitudinal rib 102 have not achieved effective fusion and the bonding force is insufficient after welding. The repair welding command is the operation signal sent by the control system to the repair welding unit 6 for secondary welding of the cold weld.

[0135] The formula for calculating the score for determining poor solder joints is: ,in, The score for determining a cold solder joint is limited to 0-100 points. The welding score is used to determine the score. The score is used to determine the offset. Weighting for weld fusion determination, and The weights for determining the position offset satisfy the following conditions: + =1.

[0136] For example, It is 0.6. The weight is 0.4. This weight allocation is determined based on the conclusion that the influence of fusion quality on weld strength is greater than that of positional deviation in engineering practice. Taking the above-calculated welding judgment score of 93.62 and offset judgment score of 63.33 as an example, substituting them into the formula, we get the false weld judgment score = 93.62×0.6 + 63.33×0.4 = 56.17 + 25.33 = 81.5 points. This score is higher than the preset qualified threshold of 80 points, so the weld is judged as a qualified weld and no repair welding instruction is issued. If the welding judgment score of a weld is 70 points and the offset judgment score is 75 points, then the false weld judgment score = 70×0.6 + 75×0.4 = 42 + 30 = 72 points, which is lower than the preset qualified threshold of 80 points. The weld is judged as a false weld, and the control system immediately generates a repair welding instruction and sends it to the repair welding unit 6.

[0137] The repair welding step involves calculating the repair welding trigger time through position compensation and controlling the repair welding unit 6 to perform the repair welding operation at the trigger time. This step involves the repair welding unit 6 to perform precise repair welding on the sparse weld points. Position compensation refers to the position correction calculated by taking into account the dynamic changes in the position of the sparse weld points caused by the continuous rotation of the roller welding machine 1 and the position deviation caused by the system command transmission delay. The repair welding trigger time is the precise time point at which the repair welding unit 6 begins to perform the repair welding operation. The repair welding unit 6 is a mechanism assembled around the roller welding machine 1 for secondary welding of sparse weld points. Its structure is consistent with that of the welding unit 2, but the welding parameters of the repair welding unit 6 can be adjusted according to the repair welding requirements.

[0138] First, based on the additional welding triggering timing strategy configured in the welding repair step, the welding repair triggering timing is obtained, specifically:

[0139] The encoder acquires the second image at the moment of acquisition of the dummy solder joint, and the circumferential angle of the roller of the roll welding machine 1 corresponding to the dummy solder joint;

[0140] The real-time rotational angular velocity of the annular rib 101 is then obtained through an encoder. The annular rib 101 rotates synchronously with the roller of the welding machine 1, so the rotational angular velocity of the roller can be obtained in real time through the encoder. The encoder is installed at the drive shaft end of the forming roller of the welding machine 1 to collect the rotational angle and angular velocity of the roller in real time; and to obtain the delay time from the image acquisition time to the command issuance time. The delay time represents the time interval from the completion of the second image acquisition to the system issuing the welding command. Then, the circumferential angle compensation amount is calculated based on the real-time rotational angular velocity and the delay time. The circumferential angle compensation amount represents the angle by which the roller drives the rebar cage 103 to rotate within the delay time.

[0141] The real-time circumferential angle is obtained by summing the historical circumferential angle with the circumferential angle compensation. The calculation formula is as follows: + ,in, This is the actual circumferential angle; ω represents the historical circumferential angle corresponding to the second image acquisition moment; ω represents the rotational angular velocity of the welding machine 1. For the duration of the delay;

[0142] Based on the target circumferential angle and the real-time circumferential angle in the coordinates of the repair area, the estimated time for the dummy weld point to reach the repair area is calculated. The estimated time represents the time required for the dummy weld point to rotate from the real-time circumferential angle to the target circumferential angle.

[0143] The formula for calculating the estimated time is:

[0144] t= ,

[0145] in, t For estimated time; The target circumferential angle is the circumferential angle of the roller corresponding to the installation position of the welding unit 6, which is a fixed value; For real-time circumferential angle;

[0146] The welding trigger time is obtained by summing the time when the welding repair command is issued with the estimated time.

[0147] For example, the second image acquisition time is 10.2s. At this time, the encoder feedback shows that the circumferential angle of the dummy solder joint is 120°. According to the real-time data acquired by the encoder, the roller rotates by 5° within 100ms. Therefore, ω is 50° / s.

[0148] The delay time from the image acquisition time to the system issuing the welding repair command is 0.3s. The welding repair command issuance time can be calculated to be 10.5s. The circumferential angle compensation amount is calculated to be 15°. The real-time circumferential angle is then calculated to be 135°. Using the difference of 45° between the target circumferential angle of 180° and the real-time circumferential angle of 135°, the estimated time is calculated to be 0.9s. The welding repair command issuance time and the estimated time are then summed to obtain the welding repair trigger time of 11.4s. The welding repair unit 6 accurately triggers the welding repair according to the welding repair trigger time, completely offsetting the position deviation caused by equipment delay and rotation.

[0149] The repair welding unit 6 performs repair welding operations on the malfunctioning weld points according to the pre-set welding parameters, ensuring accurate repair positions and effectively repairing the malfunctioning weld points.

[0150] After the repair welding is completed, in order to verify the quality of the repaired weld points, a repair welding verification step is added. Ultrasonic testing technology is used to determine whether there are internal defects such as incomplete fusion or slag inclusions in the repair weld points, providing a basis for the subsequent marking strategy. The repair welding verification strategy specifically includes: by assembling an ultrasonic transmitter 7 next to the repair welding unit 6, ultrasonic waves are emitted from the transmitting end of the ultrasonic transmitter 7 toward the repair weld points of the annular rib 101 and the longitudinal rib 102. If the reflected wave value received by the receiving end is lower than a preset threshold, the repair weld point is determined to be a qualified repair weld point; if the reflected wave value received by the receiving end is higher than the preset threshold, the repair weld point is determined to be a unqualified repair weld point, and the marking strategy is executed.

[0151] The ultrasonic transmitter 7 in the above-mentioned weld repair verification strategy is a testing device with the function of transmitting and receiving ultrasonic waves. It consists of a transmitter, a receiver, and a signal processing module. It is suitable for detecting internal defects in metal components. The transmitter is the component in the ultrasonic transmitter 7 that generates and emits ultrasonic waves. It is usually a piezoelectric ceramic probe that can convert electrical signals into ultrasonic signals. Ultrasonic waves are mechanical waves with a frequency higher than 20,000 Hz. They have strong penetrating power and can detect internal defects, making them suitable for quality inspection of metal welds. The receiver is the component in the ultrasonic transmitter 7 that receives reflected ultrasonic signals. It is also a piezoelectric ceramic probe that can convert ultrasonic signals into electrical signals and transmit them to the signal processing module. The reflected wave value is the peak voltage value of the electrical signal converted by the receiver from the reflected ultrasonic signal. The unit is volts. Its magnitude is related to the dielectric properties inside the weld repair point. Defective areas will cause the reflected wave value to rise abnormally. The preset threshold is the critical value of the reflected wave value for judging whether the weld repair point is qualified or not, determined through process experiments. The unit is volts. If it is lower than this value, it means that there are no obvious defects inside the weld repair point. If it is higher than this value, it means that there are defects such as lack of fusion or slag inclusion.

[0152] It should be noted that the ultrasonic transmitter 7 needs to be calibrated after installation. The ultrasonic transmitter 7 is mounted next to the welding unit 6 using a fixed bracket, ensuring that the center line connecting the transmitter and receiver is perpendicular to the weld center of the welding point. The distance between the transmitter and the surface of the welding point is set to 9 mm. This distance ensures effective ultrasonic penetration while preventing direct contact between the probe and the high-temperature welding point, which could cause damage. After installation, the equipment is calibrated. Known qualified and unqualified welding points are selected as standard test pieces. Qualified welding points are those confirmed by destructive testing to be completely fused internally. Unqualified welding points are artificially created welding points with 1-2 mm of unfusion defects. The standard test pieces are tested using the ultrasonic transmitter 7, and the reflected wave value ranges of qualified and unqualified welding points are recorded. The maximum value of the reflected wave value of the qualified welding point is taken as the preset threshold.

[0153] After the welding repair operation is completed, the weld point moves to the detection area of ​​the ultrasonic transmitter 7 as the steel cage 103 rotates. At this time, the transmitting end of the ultrasonic transmitter 7 emits ultrasonic waves according to the preset parameters. After the ultrasonic waves are emitted from the transmitting end, they are incident vertically on the surface of the weld point. Part of the ultrasonic waves will be reflected on the upper surface of the weld point, and the other part will penetrate the steel bar and enter the weld. If the weld point is completely fused, the ultrasonic waves will be transmitted and refracted at the joint surface of the annular reinforcement 101 and the longitudinal reinforcement 102. The ultrasonic energy reflected back to the receiving end is relatively weak, and the corresponding reflected wave value is low. If there are defects such as incomplete fusion or slag inclusion inside the weld point, the ultrasonic waves will be strongly reflected at the defect interface. The ultrasonic energy reflected back to the receiving end is relatively strong, and the corresponding reflected wave value is high.

[0154] After receiving the reflected ultrasonic signal, the receiving end converts it into an electrical signal and transmits it to the signal processing module. The signal processing module filters, amplifies, and extracts the peak value of the electrical signal to obtain the reflected wave value. This reflected wave value is compared with a preset threshold. If the reflected wave value is lower than the preset threshold, the weld point is determined to be a qualified weld point, and no marking strategy is required. The weld point continues to rotate with the rebar cage 103 to enter the next process. If the reflected wave value is higher than the preset threshold, the weld point is determined to be a unqualified weld point. The signal processing module sends an unqualified signal to the control system, and the control system triggers the marking strategy.

[0155] To ensure the accuracy of the weld repair verification, the detection frequency of the ultrasonic transmitter 7 must be matched with the rotation speed of the steel cage 103.

[0156] If the weld point is found to be unqualified after the weld repair verification, a marking strategy is triggered. The strategy is as follows: when the weld point is determined to be unqualified, a marking command is sent to the physical marking unit 8; the time when the marking command is sent is summed with the time when the unqualified weld point moves to the physical marking unit 8 to obtain the marking trigger time, and the physical marking unit 8 is controlled to perform the marking operation on the unqualified weld point at the trigger time.

[0157] The above marking strategy is an operational procedure for accurately visually marking unqualified weld points after weld repair verification. The marking operation is performed by the physical marking unit 8 when the unqualified weld point moves to the corresponding position, which facilitates subsequent manual quick positioning and verification. The physical marking unit 8 is an industrial marking device installed behind the weld repair unit 6 of the roll welding machine 1. It is used to apply visual markings to the unqualified weld points. The physical marking unit 8 can be a high-temperature resistant inkjet marking unit or a laser marking unit, and it needs to be adapted to the production characteristics of the continuous rotation of the steel cage 103.

[0158] The marking trigger time of physical marking unit 8 is obtained. Using the same calculation method as the compensation triggering strategy used by the welding repair unit 6 to obtain the welding repair trigger time, the time difference between the unqualified welding repair point moving from the verification area to the marking area is compensated to ensure that the marking action and the welding point position are accurately synchronized.

[0159] The physical marking unit 8 immediately starts the marking operation at the marking trigger moment, and applies a mark to the unqualified weld repair point. The mark shape is a circular spot or a short line segment, which ensures that it can be quickly identified on the surface of the steel cage 103 during manual review.

[0160] In practical applications, the solder joint judgment step is also equipped with a simplified judgment strategy. The simplified judgment strategy is an optimized process set up in the solder joint judgment step to improve production efficiency. It is triggered after meeting specific execution conditions. By reducing the amount of calculation of judgment parameters, the detection time is shortened, while retaining key quality verification links to avoid misjudgment.

[0161] The simplified judgment strategy requires that the number of times a solder joint is consecutively judged as a qualified solder joint by the system exceeds a preset threshold. A qualified solder joint is one whose score for a faulty solder joint meets the preset qualification threshold. The consecutive qualification count is recorded sequentially by the system according to the welding order. Each time a faulty solder joint is detected, the count is reset to 0. The preset threshold is the critical number of consecutive qualified solder joints required to trigger the simplified judgment strategy. It is determined by production efficiency requirements and quality control standards, and is typically set to 50. Exceeding this value satisfies the execution condition.

[0162] It should be noted that when the number of consecutive qualified welds exceeds the threshold, a simplified judgment strategy is implemented to determine the weld as faulty. The principle is that in the production of steel cage roll welding, the positional deviation mainly comes from random factors such as the forming error of the annular rib 101 or the gap of the feeding channel. When the number of consecutive qualified welds exceeds the preset threshold, it means that the forming accuracy of the annular rib 101 and the feeding trajectory have become stable, the rotation and feeding linkage mechanism of the roll welding machine 1 are synchronized well, and the positional relationship between the longitudinal rib 102 and the annular rib 101 will most likely continue the previous stable state. The probability of the positional deviation exceeding the allowable range is extremely low, and the probability of sending positional deviation is greatly reduced. Therefore, only the welding judgment score needs to be calculated.

[0163] When the execution conditions are met, a simplified judgment strategy is executed. It only needs to determine whether the weld joint is a dummy weld joint by the welding judgment score, omitting the calculation of the offset judgment score. The welding judgment score is based on the hot melt judgment parameters, which are thermal quantitative indicators that reflect the fusion quality of the weld joint, including temperature distribution parameters, heat distribution profile parameters, and heat conduction continuity parameters. The welding judgment score is a numerical value that is calculated and quantified based on the hot melt judgment parameters to reflect the fusion quality of the weld joint. The value range is 0-100 points. The higher the score, the better the fusion quality. The judgment threshold is the critical value for judging whether the weld joint is a dummy weld joint in the simplified execution strategy. It is determined by process experiments and is usually set to 75 points.

[0164] In the simplified judgment process, the system only collects the hot melt judgment parameters to calculate the welding judgment score, and no longer calculates the offset judgment score. If the calculated welding judgment score is ≥75 points, the weld point is directly judged as a qualified weld point, and the number of qualified weld points continues to accumulate.

[0165] If the welding judgment score is below 75 points, the weld point is judged as a false weld point. In this case, a similarity verification is required to determine whether the relative positions of the longitudinal rib 102 and the annular rib 101 have changed. Specifically, the relative positions of the longitudinal rib 102 and the annular rib 101 are extracted from the first image and compared with the relative positions of the longitudinal rib 102 and the annular rib 101 extracted from the first image of the previous weld point located at the current weld point. The similarity is an index that measures the degree of difference in the relative position parameters between the current first image and the previous first image, with a value range of 0-1. The closer the value is to 1, the smaller the difference. The closer the similarity is to 0, the greater the difference. The calculated similarity is compared with the similarity threshold. The similarity threshold is the critical value for judging whether the relative position of the two steel bars at the current weld point has changed significantly compared with the relative position of the previous weld point. It is determined by process test. If the similarity is lower than this value, it means that the relative position of the longitudinal bar 102 and the ring bar 101 has changed significantly. From the perspective of geometric position, there is a positional deviation between the two steel bars. Then the relationship between the relative position of the longitudinal bar 102 and the ring bar needs to be considered. Then the conventional judgment strategy in the system is activated, that is, the welding judgment score and the offset judgment score are used together to calculate the false weld judgment score to judge the method of false weld.

[0166] In one embodiment of the present invention, the similarity calculation formula in the above determination process is as follows:

[0167] T= ×0.5+ ×0.5,

[0168] in, The formula for calculating relative angular similarity is:

[0169] =1- ,

[0170] in, The relative angle of the current first image; The relative angle between the previous solder joint and the first image; To allow the maximum relative angular deviation;

[0171] The similarity of the ratio difference between the front and rear positions of longitudinal reinforcement 102 and ring reinforcement 101 is calculated using the following formula:

[0172] =1- ,

[0173] in, The difference in the ratio between the longitudinal rib 102 and the annular rib 101 in the current first image; The difference between the ratio of longitudinal rib 102 and annular rib 101 on the first image corresponding to the previous weld point; To allow the maximum ratio deviation.

[0174] The above are merely preferred embodiments of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principle of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A precision welding method for preparing reinforcing cages using a roll welding machine, characterized in that, The method includes: The initial coiling data acquisition step involves acquiring the first image to obtain the relative positions of the annular reinforcement and longitudinal reinforcement before welding, and then calling the corresponding welding parameters from a pre-established database based on the relative positions to generate welding instructions. The welding execution steps involve the welding unit receiving welding instructions and executing the welding operation. The welding fusion judgment step uses thermal imaging to obtain the thermal fusion judgment parameters at the weld joint to obtain the welding judgment score; The welding fusion determination step includes a welding determination strategy, which specifically includes: The hot melt judgment parameters include temperature distribution parameters, heat distribution profile parameters, and heat conduction continuity parameters. The heat distribution profile parameters reflect the heated areas of the annular ribs, welding molten material, and longitudinal ribs. The temperature distribution parameters reflect the temperature distribution within the heated area at the weld point. The heat conduction continuity parameters reflect the continuous change in heat within the heated area. The welding judgment score is calculated based on each hot melt judgment parameter; The post-weld data acquisition step involves acquiring a second image at the intersection of the annular reinforcement and the longitudinal reinforcement after welding, obtaining the relative position of the annular reinforcement and the longitudinal reinforcement based on the second image, and comparing it with the relative position of the first image to obtain an offset judgment score. The solder joint judgment step calculates a cold solder joint judgment score based on the welding judgment score and the offset judgment score. The cold solder joint judgment score represents the probability of a cold solder joint. If the cold solder joint judgment score is lower than a preset qualified threshold, the solder joint is a cold solder joint, and a re-soldering command is issued. The repair welding step calculates the repair welding trigger time based on the repair welding trigger time strategy, and controls the repair welding unit to perform the repair welding operation at the repair welding trigger time. The welding repair step also includes a welding repair triggering timing strategy, specifically: Obtain the historical circumferential angle at the second image acquisition moment; The real-time rotational angular velocity of the ring rib is obtained by the encoder, as well as the delay time from the image acquisition time to the command issuance time. The circumferential angle compensation amount is calculated based on the real-time rotational angular velocity and the delay time. The real-time circumferential angle is obtained by summing the historical circumferential angle with the circumferential angle compensation. Based on the target circumferential angle and real-time circumferential angle in the coordinates of the repair area, the estimated time for the dummy weld point to reach the repair area is calculated. The welding trigger time is obtained by summing the time when the welding repair command is issued with the estimated time.

2. The precision welding method for preparing reinforcing cages based on a roll welding machine according to claim 1, characterized in that, The initial winding data acquisition step includes a relative position acquisition strategy, which specifically includes: The camera captures the first image of the intersection of the annular rib and the longitudinal rib before welding. The longitudinal rib centerline is obtained from the image based on the longitudinal rib outline, and the annular rib centerline is obtained based on the annular rib outline. The angle between the intersection point is obtained by the longitudinal rib centerline and the annular rib centerline. The angle between the intersection point reflects the relative angle between the annular rib and the longitudinal rib. From the overlapping area of ​​the longitudinal rib contour and the annular rib contour, multiple line segments perpendicular to the central axis of the longitudinal rib and the central axis of the annular rib are obtained in the overlapping area. The line segments are the pixel diameters of the longitudinal rib and the annular rib. The average pixel diameter of the longitudinal rib and the average pixel diameter of the annular rib are calculated using the multiple line segments. The positional relationship of the longitudinal rib relative to the first camera is obtained based on the ratio of the actual diameter of the longitudinal rib to the average pixel diameter of the longitudinal rib; the positional relationship of the annular rib relative to the first camera is obtained based on the ratio of the actual diameter of the annular rib to the average pixel diameter of the annular rib. The front-to-back positional relationship between the longitudinal reinforcement and the camera is obtained by using the positions of the annular reinforcement and the camera.

3. The precision welding method for preparing reinforcing cages based on a roll welding machine according to claim 2, characterized in that, The location acquisition strategy includes a pixel filtering strategy, specifically: Compare the slope fluctuation value of each pixel in the longitudinal rib contour with the fluctuation threshold, and remove pixels that are less than the fluctuation threshold; compare the curvature value of each pixel in the annular rib contour with the curvature threshold, and remove pixels that are less than the curvature threshold. Perform connected component analysis on the remaining pixels after removal, and remove pixels in isolated connected components; The central axes of the longitudinal and annular ribs are obtained by fitting the pixels in the connected domain along the specified directions of the longitudinal and annular ribs.

4. The precision welding method for preparing reinforcing cages based on a roll welding machine according to claim 3, characterized in that, The post-weld data acquisition step includes an offset determination strategy, which is as follows: The relative angles of the ring reinforcement and longitudinal reinforcement are obtained from the second image by using a relative position acquisition strategy. The relative angle deviations are then obtained by comparing them with the relative angles of the ring reinforcement and longitudinal reinforcement in the first image. The positional deviation is obtained by comparing the front and rear positions of the annular reinforcement and longitudinal reinforcement obtained in the second image with the front and rear positions obtained in the first image. The offset judgment score is obtained by quantifying the relative angle deviation and position deviation.

5. The precision welding method for preparing reinforcing cages based on a roll welding machine according to claim 1, characterized in that, The solder joint determination step also includes a simplified determination strategy, which is triggered after the execution conditions are met. The execution conditions are: when the score for determining a poor solder joint is not lower than a preset qualified threshold, the solder joint is a qualified solder joint, the number of consecutive times it is determined to be a qualified solder joint is recorded, and the number of consecutive times exceeds a preset threshold. The simplified judgment strategy is as follows: calculate the welding judgment score based on the hot melt judgment parameters; if the welding judgment score is lower than the judgment threshold, the weld point is a false weld point. When the simplified judgment strategy determines that the solder joint is a dummy solder joint, the relative position on the first image obtained at this time is compared with the relative position in the previous first image. If the similarity is lower than the threshold, the dummy solder joint judgment score is recalculated to determine the solder joint.

6. The precision welding method for preparing reinforcing cages based on a roll welding machine according to claim 1, characterized in that, Both the initial coiling data acquisition step and the post-weld data acquisition step include an image preprocessing strategy, which specifically includes: Multiple images with different exposures are continuously acquired for the same solder joint. Effective feature points are extracted from each image, and the fused image is obtained by feature point matching and fusion. Identify bright spot areas in the fused image, replace the bright spot pixels with normal pixel values ​​around the bright spot, and complete the removal and repair of reflective bright spots; The fused image with reflective bright spots removed is filtered to eliminate the image blurring area caused by thermal fog, and edge feature points are enhanced by edge sharpening to obtain the processed first image and second image.

7. The precision welding method for preparing reinforcing cages based on a roll welding machine according to claim 1, characterized in that, The welding repair step also includes a welding repair verification strategy, which specifically includes: By assembling an ultrasonic transmitter next to the welding unit, ultrasonic waves are emitted from the transmitting end of the ultrasonic transmitter toward the welding point of the annular rib and the longitudinal rib. If the reflected wave value received by the receiving end is lower than the preset threshold, the welding point is determined to be a qualified welding point. If the received reflected wave value is higher than the preset threshold, the solder joint is determined to be an unqualified solder joint, and a marking strategy is executed.

8. The precision welding method for preparing reinforcing cages based on a roll welding machine according to claim 7, characterized in that, The labeling strategy specifically includes: When a weld repair point is determined to be an unqualified weld repair point, a marking instruction is sent to the physical marking unit; The marking trigger time is obtained by summing the time when the marking command is issued with the time when the unqualified repair weld point is moved to the physical marking unit. The physical marking unit is then controlled to perform the marking operation on the unqualified repair weld point at the trigger time.

Images