A Laser Ranging-Based Positioning and Correction Method for Steel Pipe Welding

Abstract

This invention discloses a laser ranging-based method for positioning and correcting steel pipe welding, belonging to the field of steel pipe welding. The method includes: clamping and displacing the steel pipe to be welded to an automated welding unit using a robotic arm; acquiring an image of the steel pipe to be welded and locating the welding start point; activating a target laser rangefinder to spatially locate the welding start point and obtain welding position parameters; obtaining the welding torch position parameters of the welding robot; analyzing and generating the welding torch displacement trajectory; correcting the welding trajectory to obtain the welding movement path; and jointly controlling the welding robot to complete automated welding. This application solves the technical problems of insufficient positioning accuracy, difficulty in adapting to different entry angles, and weld seam offset caused by deformation during welding in existing technologies. It achieves the technical effects of improving the positioning accuracy of steel pipe welding, realizing automatic tracking welding of steel pipes entering at any angle, and real-time correction of the weld seam position during welding.

Description

Technical Field

[0001] This invention relates to the field of steel pipe welding, and specifically to a method for positioning and correcting steel pipe welding based on laser ranging. Background Technology

[0002] Steel pipe welding is a crucial step in industrial production, and its quality directly impacts the performance and safety of the final product. With the development of industrial automation, automated welding technology has been widely applied. However, in actual production, steel pipe welding still faces technical challenges such as insufficient positioning accuracy, difficulty adapting to different entry angles, and weld seam displacement due to deformation during the welding process. Existing steel pipe welding technologies typically employ fixed welding workstations, requiring the steel pipe to enter the workstation at a specific angle and position. This limits production flexibility and makes it difficult to adapt to steel pipes with different entry angles, often requiring additional equipment or manual intervention to adjust the pipe's position. Regarding welding positioning, traditional methods mainly rely on mechanical positioning devices or simple vision systems, which have limited accuracy and cannot meet the requirements of high-precision welding. Especially in complex industrial environments, relying solely on vision systems is easily affected by factors such as lighting and reflection, making it difficult to obtain reliable spatial position information. Furthermore, during the welding process, the steel pipe deforms due to heat, causing weld seam displacement. Summary of the Invention

[0003] This application provides a laser ranging-based method for positioning and correcting steel pipe welding, aiming to solve the technical problems of insufficient positioning accuracy, difficulty in adapting to different entry angles, and weld seam offset caused by deformation during welding in the prior art.

[0004] The laser ranging-based steel pipe welding positioning and correction method disclosed in this application includes: clamping and displacing the steel pipe to be welded to an automated welding unit using a robotic arm, wherein a laser ranging array and an industrial camera array are arranged around the welding operation space of the automated welding unit; using the industrial camera array to perform image acquisition on the steel pipe to be welded, and locating the welding start point based on the image acquisition results; activating a target laser rangefinder in the laser ranging array according to the welding start point, and scheduling the target laser rangefinder to run independently to spatially locate the welding start point and obtain welding position parameters; interacting with the automated welding unit to obtain the welding torch position parameters of the welding robot; pre-constructing a welding torch displacement analysis model, which generates a welding torch displacement trajectory based on the welding torch position parameters and the welding position parameters; interactively obtaining preset welding trajectory data, and correcting the welding trajectory based on the welding position parameters and the preset welding trajectory data to obtain a welding movement path; and using the welding torch displacement trajectory and the welding movement path to jointly control the welding robot to complete the automated welding of the steel pipe to be welded.

[0005] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0006] By employing a robotic arm to grip and move the steel pipe to be welded to an automated welding unit, the welding operation space of which is surrounded by a laser ranging array and an industrial camera array, automatic transport of the steel pipe is achieved, preparing for subsequent precise positioning and welding. The surrounding laser ranging array and industrial camera array provide the hardware foundation for comprehensive detection and positioning. The industrial camera array performs image acquisition on the steel pipe to be welded, and the welding start point is located based on the image acquisition results. Visual technology is used to initially determine the welding start point, laying the foundation for subsequent precise positioning. Based on the welding start point, the target laser rangefinder is activated in the laser ranging array and scheduled to operate independently to spatially locate the welding start point and obtain welding position parameters. Precise spatial positioning of the welding start point using laser ranging technology obtains coordinate information, improving positioning accuracy. The automated welding unit interacts with the welding robot to obtain the welding gun position parameters, providing a basis for subsequent path planning. The process involves: pre-constructing a welding torch displacement analysis model, which generates the torch displacement trajectory based on torch position parameters and welding position parameters, and planning the optimal path from the current position to the welding start point; interactively obtaining preset welding trajectory data, and correcting the welding trajectory based on welding position parameters and preset welding trajectory data to obtain the welding movement path; considering the actual welding position, correcting the preset welding trajectory to adapt to the current welding task; using the torch displacement trajectory and welding movement path to jointly control the welding robot, completing the automated welding of the steel pipe to be welded, realizing a technical solution for precise control and automated welding of the welding robot, solving the technical problems of insufficient positioning accuracy, difficulty in adapting to different entry angles, and weld seam offset caused by deformation during welding in existing technologies, achieving the technical effects of improving the positioning accuracy of steel pipe welding, realizing automatic tracking welding of steel pipes entering at any angle, and real-time correction of weld seam position during welding.

[0007] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0008] Figure 1 A schematic flowchart of a laser ranging-based steel pipe welding positioning and correction method is provided for embodiments of this application;

[0009] Figure 2 This application provides a schematic flowchart of a method for obtaining welding position parameters in a steel pipe welding positioning and correction method based on laser ranging, which is provided for an embodiment of the present application. Detailed Implementation

[0010] The overall concept of the technical solution provided in this application is as follows:

[0011] This application provides a laser ranging-based method for positioning and correcting steel pipe welding. First, a robotic arm transports the steel pipe to be welded to an automated welding unit equipped with a surrounding array of industrial cameras and a laser ranging array. The industrial camera array performs initial image acquisition and welding start-up location, laying the foundation for subsequent precise positioning. Then, based on the initial positioning results, the corresponding laser ranging instrument is activated to perform high-precision spatial positioning of the welding start-up, obtaining accurate coordinate parameters. Based on this, a pre-constructed welding torch displacement analysis model, combined with the current position of the welding torch and the welding target position, is used to plan the optimal welding torch movement trajectory. Simultaneously, considering actual welding requirements, preset welding trajectory data is introduced and dynamically corrected based on the actual positioning results to generate the final welding movement path. Finally, through joint control of the welding robot, precise positioning and movement of the welding torch are achieved, completing a high-quality automated welding task.

[0012] After introducing the basic principles of this application, the non-limiting embodiments of this application will be described in detail below with reference to the accompanying drawings.

[0013] like Figure 1 As shown in the embodiment of this application, a method for positioning and correcting steel pipe welding based on laser ranging is provided. The method includes:

[0014] S1: The steel pipe to be welded is gripped and moved to the automated welding unit by a robotic arm, wherein the welding operation space of the automated welding unit is surrounded by a laser rangefinder array and an industrial camera array.

[0015] Specifically, firstly, a robotic arm is used to grip the steel pipe to be welded. This robotic arm can be a multi-axis industrial robot or a dedicated gripping device, possessing sufficient load capacity and precise positioning capabilities to ensure stable gripping and movement of the steel pipe. Subsequently, the robotic arm moves the steel pipe to be welded into the operating space of the automated welding unit according to a preset trajectory. The automated welding unit is a comprehensive workstation integrating welding equipment, a positioning system, and a control system. Around the welding operating space of this automated welding unit, a laser ranging array and an industrial camera array are arranged in a ring to achieve omnidirectional coverage of the welding operating space. The laser ranging array consists of multiple laser rangefinders arranged in a horizontal cylindrical shape. Each laser rangefinder can accurately measure the distance between itself and the target object, thus providing accurate data support for subsequent spatial positioning. Similarly, the industrial camera array also adopts a similar ring arrangement, consisting of multiple high-resolution industrial cameras, capable of acquiring images of the steel pipe to be welded from different angles, providing comprehensive visual information for subsequent image processing and analysis.

[0016] By circling around the steel pipe, it can be effectively tracked and located regardless of the angle at which the pipe enters the welding operation space, laying the foundation for subsequent precise welding.

[0017] S2: The industrial camera array is used to perform image acquisition on the steel pipe to be welded, and the welding start point is located based on the image acquisition results.

[0018] Specifically, firstly, an industrial camera array is activated to acquire images of the steel pipe to be welded as it enters the welding operation space. Since the industrial camera array is arranged in a ring around the welding operation space, it can simultaneously capture image information of the steel pipe from multiple angles. This multi-angle acquisition method ensures comprehensive capture of the steel pipe's surface features, effectively avoiding information omissions that might occur with a single angle. During image acquisition, each industrial camera generates a local image of the steel pipe. Each image is assigned a unique image source identifier, recording which industrial camera captured the image. Next, the acquired multiple local steel pipe images are input into a pre-built welding start-point recognition model. Based on image processing and machine learning algorithms, the welding start-point recognition model can automatically analyze features in the images and identify suitable locations as welding start points. Then, the welding start-point recognition model outputs an image marked with the welding start point, known as the start-point image. In this image, the welding start point is identified, providing precise location information for subsequent welding operations.

[0019] S3: Activate the target laser rangefinder in the laser ranging array according to the welding starting point, and schedule the target laser rangefinder to run independently to spatially locate the welding starting point and obtain the welding position parameters.

[0020] Specifically, firstly, based on the determined welding starting point, the most suitable laser rangefinder is selected from the laser ranging array. The selection principle is: the laser rangefinder is closest to the welding starting point in space, providing the most accurate distance measurement results. The selected device is referred to as the target laser rangefinder. After activating the target laser rangefinder, it is scheduled to operate independently. This independent operation mode ensures that the ranging process is not interfered with by other devices, improving measurement accuracy.

[0021] After the target laser rangefinder is activated, it performs precise spatial positioning of the welding start point. Specifically, the laser rangefinder emits a laser beam to the welding start point and then receives the reflected laser signal. By calculating the time difference between laser emission and reception, the distance between the laser rangefinder and the welding start point is precisely measured. After the measurement is completed, welding position parameters are obtained, including the X, Y, and Z coordinates of the welding start point in three-dimensional space. These coordinate values ​​are relative to a predefined standard coordinate system and can accurately describe the precise position of the welding start point in the welding operation space.

[0022] Once the welding position parameters are obtained, the welding start point can be precisely located, providing accurate spatial information for subsequent welding torch movement and welding operations. Compared to purely vision-based methods, laser-based positioning offers higher accuracy and reliability, especially when dealing with steel pipes with complex curved surfaces or special materials. It provides more accurate spatial positioning information, laying the foundation for subsequent automated welding processes and effectively improving welding quality and efficiency.

[0023] S4: Interact with the automated welding unit to obtain the welding gun position parameters of the welding robot.

[0024] Specifically, firstly, a communication connection is established with the automated welding unit to ensure the real-time and reliable transmission of data. After establishing the connection, a data request command is sent to the automated welding unit, requesting the acquisition of the welding robot's current welding torch position parameters. Upon receiving the request, the automated welding unit immediately reads the welding robot's current spatial coordinates from its control system as the welding torch position parameters, providing necessary starting information for subsequent welding torch displacement analysis and trajectory planning. By comparing the welding torch position parameters with the welding position parameters, the distance and path the welding torch needs to move can be accurately calculated, thereby achieving efficient and precise welding operations.

[0025] S5: Pre-build a welding torch displacement analysis model, which generates a welding torch displacement trajectory based on the welding torch position parameters and welding position parameters.

[0026] Specifically, a welding torch displacement analysis model is pre-constructed. This model analyzes and generates the optimal welding torch displacement trajectory based on the input welding torch position parameters and welding position parameters. First, the obtained welding torch position parameters and welding position parameters are input into the welding torch displacement analysis model. The model processes the spatial coordinate data of the welding torch position and welding position to generate the optimal welding torch displacement trajectory. The generated trajectory describes the path the welding torch takes from its current position (specified by the welding torch position parameters) to the welding start point (specified by the welding position parameters), ensuring smooth and safe movement of the welding torch and avoiding collisions with the workpiece or other equipment.

[0027] By using a welding torch displacement analysis model, intelligent and optimized welding torch displacement can be achieved. Compared to traditional fixed-program methods, dynamic programming methods based on spatial coordinates offer greater flexibility and adaptability, automatically adjusting displacement strategies according to different welding scenarios. This effectively improves welding efficiency and accuracy while reducing unnecessary welding torch movements.

[0028] S6: Interact to obtain preset welding trajectory data, and perform welding trajectory correction based on the welding position parameters and preset welding trajectory data to obtain the welding movement path.

[0029] Specifically, firstly, preset welding trajectory data is obtained interactively. This data is pre-defined by welding process experts based on factors such as the specifications, material, and welding requirements of the steel pipe. The preset trajectory data includes information about the ideal path the welding torch should follow, such as the weld geometry, welding speed, and welding angle. Next, the previously obtained welding position parameters are compared and analyzed with the preset trajectory data, and a welding trajectory correction process is performed. This process adjusts and optimizes the preset trajectory data according to the actual welding position parameters, including but not limited to the actual spatial position of the welding start point, the actual placement angle and direction of the steel pipe, and any potential workpiece deformation or positional deviations. Through correction, a welding movement path that better matches the actual situation is generated, considering not only ideal welding requirements but also various variables under actual working conditions, thereby ensuring the accuracy and reliability of the welding process. The resulting welding movement path is a series of corrected and optimized spatial coordinate points that collectively describe the actual movement trajectory the welding torch should follow during the welding process, guiding the welding robot to precisely perform welding operations along the surface of the steel pipe.

[0030] By intelligently adjusting the welding trajectory, the ideal welding requirements are effectively combined with the actual working conditions, greatly improving the adaptability and accuracy of the welding process. It can cope with common problems in actual production, such as workpiece placement errors and material deformation, thereby ensuring the consistency and reliability of welding quality.

[0031] S7: The welding robot is jointly controlled by the welding torch displacement trajectory and the welding movement path to complete the automated welding of the steel pipe to be welded.

[0032] Specifically, the obtained welding torch displacement trajectory and welding movement path are used as input parameters for joint control of the welding robot. This joint control strategy fully utilizes the obtained precise spatial information and optimized motion path to achieve highly accurate and efficient welding operations. First, the welding robot moves the welding torch according to the welding torch displacement trajectory. During this process, the welding torch moves from its initial position to the welding start point along a pre-calculated optimal path. When the welding torch reaches and is tangent to the welding start point, the welding power of the welding robot is activated, and the formal welding process begins. The tangent state between the welding torch and the welding start point ensures the precise positioning of the welding start point and the ideal welding angle, laying the foundation for high-quality welding. Subsequently, according to the obtained welding movement path, the welding robot's welding torch is precisely guided to move on the surface of the steel pipe to be welded. The welding torch moves strictly according to the calibrated welding movement path, ensuring the continuity and uniformity of the weld. After the welding torch completes the entire welding movement path, the welding power is automatically turned off, and the welding torch is controlled to return to a safe position, completing the entire automated welding process.

[0033] By employing high-precision and high-efficiency automated welding of steel pipes, the positioning accuracy of steel pipe welding has been improved, automatic tracking welding of steel pipes entering the site at any angle has been achieved, and the weld position has been corrected in real time during the welding process. This has improved welding quality and production efficiency, and reduced reliance on operator skills.

[0034] Furthermore, embodiments of this application also include:

[0035] A pre-constructed image acquisition window is used; when the robotic arm's stop time reaches the image acquisition window, the industrial camera array is activated to acquire images of the steel pipe to be welded, obtaining multiple local steel pipe images, wherein the multiple local steel pipe images have multiple image source identifiers; a pre-constructed welding start identification model is used, and the welding start identification is performed by synchronizing the multiple local steel pipe images to the welding start identification model, and a start image is output, wherein the welding start is identified in the start image.

[0036] In one feasible implementation, an image acquisition window is first predefined. The image acquisition window is a time parameter used to determine the optimal time to perform image acquisition. The purpose of pre-defining the image acquisition window is to ensure that image acquisition is performed when the steel pipe is in a stable state, thereby obtaining clear and accurate image data. Timing begins after the robotic arm completes the steel pipe transport and stops operating. When the stop time reaches the preset image acquisition window, the industrial camera array is automatically activated to perform image acquisition on the steel pipe to be welded. Multiple cameras in the industrial camera array work simultaneously, capturing images of the steel pipe from different angles, obtaining multiple partial images of the steel pipe. Each image is assigned a unique image source identifier to record which industrial camera captured the image.

[0037] Subsequently, a pre-constructed welding start-point recognition model is built. Specifically, firstly, a large number of steel pipe welding image samples are collected, including steel pipes of various types, sizes, and shapes, as well as images under different welding conditions. Each sample is marked with the ideal welding start-point position by professional welders. Secondly, the collected steel pipe welding image samples are preprocessed, including image normalization, denoising, and enhancement, to improve the subsequent feature extraction effect. Then, edge detection and corner detection algorithms are applied to extract features from the preprocessed images. Edge detection can use the Canny operator or the Sobel operator, while corner detection can use the Harris corner detection algorithm or the FAST corner detection algorithm. Next, a convolutional algorithm is selected... Using a network as its basic architecture, the input layer receives extracted features, and the output layer provides the predicted welding start point. Subsequently, the model is trained using prepared sample data. Annotated images are input into the network, and the network parameters are continuously adjusted using the backpropagation algorithm. During training, cross-validation is used to evaluate model performance, and hyperparameters are adjusted to optimize the model. Afterward, the trained model is tested and validated using a set of unused image samples to evaluate its generalization ability. If the model performs poorly, the feature extraction method or network structure is adjusted, and then retraining is performed. Through this iterative construction process, a welding start point recognition model capable of accurately identifying welding start points is obtained. Then, multiple local steel pipe images are simultaneously input into the welding start point recognition model. The model analyzes the input images, identifies the most suitable location as the welding start point, and outputs a start point image. The identified welding start point is clearly marked in the start point image, serving as a reference point for subsequent welding processes.

[0038] By accurately locating the welding start point, the accuracy and reliability of the welding start point positioning are improved, laying the foundation for subsequent welding operations and effectively meeting the welding needs of steel pipes with various complex shapes.

[0039] Furthermore, embodiments of this application also include:

[0040] The process involves: interactively obtaining spatial parameter information of the welding operation space; modeling and reconstructing the welding operation space based on the spatial parameter information to obtain an operation space model; interactively obtaining standard workpiece parameters of the steel pipe to be welded, and modeling and reconstructing the steel pipe to be welded based on the standard workpiece parameters to obtain a standard workpiece model; interactively obtaining the initial clamping trajectory of the robotic arm; using the initial clamping trajectory to guide the standard workpiece model to fit to the operation space model to obtain an updated space model; performing equipment layout analysis based on the updated space model to obtain equipment model information and equipment arrangement information, wherein the equipment arrangement information includes laser ranging arrangement parameters and industrial camera arrangement parameters; and arranging the equipment around the welding operation space according to the equipment arrangement information to obtain the laser ranging array and industrial camera array.

[0041] In a preferred embodiment, firstly, spatial parameter information of the welding operation space is obtained through a human-machine interface, including the dimensions, internal structural features, and potential obstacles of the operation space. The operator can provide this information by inputting specific values ​​or uploading CAD drawings, and can also specify specific work areas or restricted areas, providing guidance for subsequent equipment layout optimization. Secondly, using 3D modeling technology, a high-precision virtual operation space model is constructed based on the obtained spatial parameter information, including not only the external contour of the space but also accurately reproducing its internal structure and features. Then, again through the human-machine interface, detailed parameters of the steel pipe to be welded are obtained, including dimensions, material, and surface treatment. Based on these parameters, a precise standard workpiece model is constructed using parametric modeling technology, reflecting not only the geometric features of the steel pipe but also material property information, providing a foundation for subsequent welding simulation. Simultaneously, the initial gripping trajectory of the robotic arm is obtained, describing how the robotic arm grasps the steel pipe and moves it to the welding operation space, including the spatial coordinates of a series of key points and the angular changes of each joint of the robotic arm.

[0042] Next, in the virtual environment, the fitting process of the robotic arm moving the standard workpiece model into the operating space model according to the initial clamping trajectory is simulated. After fitting, an updated space model is generated, which contains the precise spatial information of the steel pipe in its final position. Subsequently, based on the updated space model, a comprehensive equipment layout analysis is performed. This analysis considers multiple factors such as spatial constraints, measurement coverage, measurement accuracy, inter-device interference, and cost-effectiveness. Heuristic or optimization algorithms are used to search for the optimal equipment layout scheme. The analysis results include the specific model, installation coordinates, and orientation angle of each device, obtaining equipment model information and equipment layout information. The equipment layout information includes laser ranging array layout parameters and industrial camera layout parameters. Then, based on the equipment layout information, detailed equipment installation instructions are generated, including the precise installation coordinates, installation angles, calibration parameters, etc., for each device, ensuring accurate equipment placement. After installation, automatic calibration and testing are performed to ensure that the laser ranging array and industrial camera array can operate normally according to design requirements, obtaining the laser ranging array and industrial camera array data.

[0043] Furthermore, embodiments of this application also include:

[0044] A preset slicing direction is used as a constraint to geometrically segment the updated space model, resulting in multiple slice space models. Multiple cross-sectional views are extracted from these slice space models. Based on these cross-sectional views, the straight-line distance between the standard workpiece model and the operating space model is collected, yielding multiple sets of spatial distance parameters. These spatial distance parameters are serialized and their maximum values ​​are extracted to obtain spatial distance extreme values. The spatial distance extreme values ​​are used to traverse a pre-constructed equipment configuration table to obtain equipment model information, including a first equipment model and a second equipment model. Based on the first equipment model and the spatial distance extreme values, the equipment acquisition range is analyzed to obtain the laser ranging interval distance. This laser ranging interval distance is then fitted to the updated space model to obtain the laser ranging arrangement parameters. This process is repeated to analyze and obtain the industrial camera arrangement parameters.

[0045] In a preferred embodiment, firstly, a slicing direction is preset, which is selected to best reflect the axis of the operating space and workpiece features. Then, constrained by the preset slicing direction, the updated space model is geometrically segmented, transforming the 3D model into a series of 2D slices, obtaining multiple slice space models. Each slice space model represents a cross-section of the updated space model at a specific location. Next, multiple cross-sectional views are extracted from these slice space models. Each cross-sectional view clearly shows the relative positional relationship between the standard workpiece model and the operating space model at that location, providing an intuitive 2D representation for subsequent distance analysis. Then, based on these cross-sectional views, the straight-line distance between the standard workpiece model and the operating space model is acquired. Specifically, on each cross-sectional view, the shortest distance from a point on the workpiece contour to the boundary of the operating space is calculated, obtaining multiple sets of spatial distance parameters, each set of parameters corresponding to the distance information of a slice location.

[0046] Subsequently, multiple sets of spatial distance parameters were serialized, and the maximum values ​​were extracted to represent the maximum distance between the workpiece and the operating space, known as spatial distance extrema. Simultaneously, a pre-constructed equipment configuration table was built. Specifically, detailed technical parameters of commercially available laser rangefinders and industrial cameras were collected, including but not limited to measurement range, accuracy, resolution, response time, operating temperature range, protection level, and interface type. These parameters were categorized and organized according to equipment type (laser rangefinders and industrial cameras). Each equipment type formed a sub-table, with each row representing a specific equipment model and columns corresponding to various technical parameters. A unique identifier, i.e., the equipment model code, was assigned to each device. An applicable range field was added to each device, calculated based on parameters such as measurement range and accuracy, representing the optimal working distance range for that device in the current application scenario. This data was then organized into a structured data table, the equipment configuration table. Finally, the obtained spatial distance extrema were used to iterate through the pre-constructed equipment configuration table. By comparing the extreme spatial distances with the measurement ranges of each device, the system selects the device models that meet the requirements, obtaining model information for two types of devices: a first type (laser rangefinder) and a second type (industrial camera). For the laser rangefinder, based on the selected first device model and the extreme spatial distances, the system analyzes the device acquisition range, considering factors such as the measurement range and accuracy requirements of the devices, to determine the optimal spacing between the laser rangefinders, thus obtaining the laser ranging interval. This laser ranging interval is then applied to update the spatial model, yielding the laser ranging layout parameters, including the specific installation position and orientation of each rangefinder. Subsequently, the system uses a similar method to analyze and determine the industrial camera layout parameters based on the determined second device model, ensuring that the camera array can fully cover the area to be welded and provide sufficient image clarity.

[0047] Furthermore, such as Figure 2 As shown, embodiments of this application also include:

[0048] Based on the image source identifier of the starting point image, locate the target camera and the target image acquisition position in the industrial camera array; locate the target laser rangefinder in the laser ranging array, wherein the laser rangefinder is closest to the target image acquisition position in space; schedule the target laser rangefinder to operate independently to spatially locate the welding starting point and obtain the welding position parameters.

[0049] In one feasible implementation, firstly, the target camera and target image acquisition position are located within an industrial camera array based on the image source identifier of the starting point image. Each acquired image carries a unique identifier corresponding to a specific camera and its installation position. By parsing the image source identifier, it is possible to accurately determine which camera captured the welding starting point and its specific spatial location, laying the foundation for subsequent laser rangefinder selection. Next, the most suitable laser rangefinder is selected from the laser rangefinder array. The selection principle is to find the laser rangefinder with the closest spatial distance to the target image acquisition position. The rangefinder closest to the image acquisition position can usually measure the welding starting point with the best angle and accuracy. By calculating the spatial distance between each laser rangefinder and the target image acquisition position, the one with the smallest distance is selected as the target laser rangefinder. After determining the target laser rangefinder, it is scheduled to operate independently to accurately locate the welding starting point in space. The target laser rangefinder first aligns with the estimated welding starting point and then emits a laser beam to measure the precise distance between itself and the welding starting point. After acquiring the measured distance data, it is combined with pre-calibrated standard coordinate system information. By using a spatial geometric transformation algorithm, the ranging results are converted into three-dimensional coordinates in a unified coordinate system of the welding operation space, thereby obtaining the precise spatial position parameters of the welding start point and the welding position parameters.

[0050] By combining image positioning from an industrial camera with precise measurement from a laser rangefinder, high-precision spatial positioning of the welding start point was achieved. Using the nearest laser rangefinder not only improved measurement accuracy but also reduced potential obstructions or interference. The ranging results were converted into a unified coordinate system, providing accurate spatial references for subsequent welding path planning and robot control, thus laying a solid foundation for the accuracy and reliability of the entire welding process.

[0051] Furthermore, embodiments of this application also include:

[0052] The target laser rangefinder is scheduled to operate independently, and the spatial distance between the welding starting point and the target laser rangefinder is collected to obtain the weld point-equipment spatial distance; the automated welding unit is interacted with to obtain the standard coordinate system of the welding operation space and the equipment coordinate parameters of the target laser rangefinder; the weld point-equipment spatial distance is transformed according to the equipment coordinate parameters and the standard coordinate system to obtain the welding position parameters.

[0053] In a preferred embodiment, firstly, the target laser rangefinder is scheduled to operate independently. During this process, the target laser rangefinder is activated and aligned with the previously identified welding start point. The target laser rangefinder emits a laser beam, and the straight-line distance between the welding start point and the rangefinder is precisely calculated by measuring the time it takes for the laser to reflect back to the receiver, thus obtaining the weld point-equipment spatial distance value. Next, it interacts with the automated welding unit to obtain the standard coordinate system of the welding operation space and the equipment coordinate parameters of the target laser rangefinder. The standard coordinate system defines the three-dimensional reference frame of the entire welding operation space, with a fixed point on the welding worktable as the origin; the equipment coordinate parameters of the target laser rangefinder describe the precise position of the rangefinder in this standard coordinate system. Subsequently, coordinate transformation calculations are performed. Using the previously obtained weld point-equipment spatial distance, combined with the equipment coordinate parameters of the target laser rangefinder and the standard coordinate system of the welding operation space, a series of spatial geometric transformations are executed, including matrix operations such as translation and rotation. By calculation, the distance information of the target laser rangefinder is converted into three-dimensional coordinates in the standard coordinate system, and the precise position parameters of the welding starting point in the standard coordinate system of the welding operation space are obtained, which are expressed as (X, Y, Z) coordinate values, and the welding position parameters are obtained.

[0054] By combining high-precision laser ranging and accurate coordinate system transformation, the welding start point can be accurately located, thus providing consistent and accurate position data throughout the entire welding operation space, which supports subsequent welding path planning.

[0055] Furthermore, embodiments of this application also include:

[0056] Historical displacement data is collected from the automated welding unit to obtain multiple sample displacement trajectories, multiple sample welding torch positions, and multiple sample welding positions. These sample displacement trajectories, welding torch positions, and welding positions are then divided to obtain K sample displacement analysis data. A standard displacement analysis model is constructed based on a backpropagation neural network, and the model parameters of the standard displacement analysis model are optimized using the K sample displacement analysis data to obtain K welding torch displacement analysis branches. These K welding torch displacement analysis branches are then connected in parallel to complete the construction of the welding torch displacement analysis model. The welding torch position parameters and welding position parameters are synchronized to the welding torch displacement analysis model to obtain K backup displacement trajectories. The K backup displacement trajectories are serialized, and the shortest trajectory is extracted based on the sorting results to obtain the welding torch displacement trajectory.

[0057] In a preferred embodiment, firstly, a large-scale historical displacement data collection is performed on the automated welding unit, recording numerous sample displacement trajectories, corresponding sample welding torch positions, and sample welding positions. This historical data covers various welding scenarios and conditions, providing rich learning materials for subsequent model training. Next, the multiple sample displacement trajectories, multiple sample welding torch positions, and multiple sample welding positions are divided into K subsets. Each subset contains complete displacement trajectory, welding torch position, and welding position information, forming K independent sample displacement analysis data. Subsequently, a standard displacement analysis model is constructed based on a backpropagation neural network to learn the complex nonlinear relationship between the welding torch position, welding position, and optimal displacement trajectory. Then, the standard displacement analysis model is trained and its parameters optimized using the K sample displacement analysis data. Training each sample displacement analysis data generates an optimized model, resulting in a welding torch displacement analysis branch. Thus, K trained welding torch displacement analysis branches are ultimately obtained. After training each branch, the K welding torch displacement analysis branches are combined in parallel to form a welding torch displacement analysis model.

[0058] In practical applications, the current welding torch position parameters and welding position parameters are simultaneously input into the constructed welding torch displacement analysis model. The model's K welding torch displacement analysis branches are calculated concurrently, with each branch generating a backup displacement trajectory, resulting in K backup displacement trajectories. These K backup displacement trajectories are then serialized and sorted based on their length, prioritizing the shortest trajectory. The optimal welding torch displacement trajectory—the shortest path and most efficient movement scheme—is extracted from these K backup trajectories.

[0059] Furthermore, embodiments of this application also include:

[0060] The welding robot moves its welding torch tangent to the welding starting point using the welding torch displacement trajectory control, and then starts the welding power supply of the welding robot; the welding movement path guides the welding torch of the welding robot to move on the surface of the steel pipe to be welded, thereby completing the automated welding of the steel pipe to be welded.

[0061] In one feasible implementation, firstly, the movement of the welding robot is controlled by a generated welding torch displacement trajectory. Following this trajectory, the welding robot precisely moves the welding torch from its initial position to near the welding start point. This not only controls the torch to reach the correct spatial position but also ensures that the torch remains tangential to the welding start point, guaranteeing the optimal angle and distance between the torch and the welding surface. When the torch accurately reaches the predetermined position and is tangential to the welding start point, the system immediately activates the welding robot's power supply, ensuring the welding operation begins in the optimal position and state, thereby effectively improving the initial welding quality. Next, the system switches to using a welding movement path to guide the subsequent movements of the welding robot. While the torch moves, the welding process continues, achieving automated welding of the steel pipe to be welded, improving welding efficiency and consistency.

[0062] In summary, the laser ranging-based steel pipe welding positioning and correction method provided in this application has the following technical effects:

[0063] A robotic arm grips and moves the steel pipe to be welded to an automated welding unit. The automated welding unit's welding operation space is surrounded by a laser rangefinder array and an industrial camera array, preparing for subsequent precise positioning and welding. The industrial camera array acquires images of the steel pipe to be welded, and based on the image acquisition results, the welding start point is located, providing basic data for subsequent precise positioning. Based on the welding start point, the target laser rangefinder is activated in the laser rangefinder array and scheduled to operate independently to spatially locate the welding start point, obtaining welding position parameters and precisely positioning the welding start point. The automated welding unit interacts to obtain the welding torch position parameters of the welding robot and acquire the current status of the welding equipment, providing necessary information for subsequent path planning. A pre-built welding torch displacement analysis model is constructed, which generates the welding torch displacement trajectory based on the welding torch position parameters and the welding position parameters, planning the welding torch movement trajectory. Pre-set welding trajectory data is obtained interactively, and the welding trajectory is corrected based on the welding position parameters and the pre-set welding trajectory data to obtain the welding movement path and correct the welding trajectory to adapt to actual welding requirements. The welding robot is controlled by combining the welding torch displacement trajectory and the welding movement path to complete the automated welding of the steel pipe to be welded. This completes the entire automated welding process, improves the positioning accuracy of the steel pipe welding, enables automatic tracking welding of steel pipes entering the site at any angle, and allows for real-time correction of the weld position during the welding process.

[0064] In summary, any step of the method described above can be stored as a computer instruction or program in an unrestricted computer memory, and can be called and identified by an unrestricted computer processor to implement any method in the embodiments of this application, without any additional restrictions.

[0065] Furthermore, the "first" or "second" mentioned above may not only represent a sequential relationship, but may also represent a specific concept, and / or refer to the individual or collective selection of multiple elements. Clearly, those skilled in the art can make various modifications and variations to this application without departing from its scope. Therefore, if such modifications and variations fall within the scope of this application and its equivalents, this application intends to include such modifications and variations.

Claims

1. A method for positioning and correcting steel pipe welding based on laser ranging, characterized in that, The method includes: The steel pipe to be welded is gripped and moved by a robotic arm to an automated welding unit, wherein the welding operation space of the automated welding unit is surrounded by a laser rangefinder array and an industrial camera array. The industrial camera array is used to acquire images of the steel pipe to be welded, and the welding start point is located based on the image acquisition results; Based on the welding starting point, the target laser rangefinder is activated in the laser ranging array, and the target laser rangefinder is scheduled to operate independently to spatially locate the welding starting point and obtain welding position parameters. Interact with the automated welding unit to obtain the welding gun position parameters of the welding robot; A pre-constructed welding torch displacement analysis model is used to generate a welding torch displacement trajectory based on the welding torch position parameters and welding position parameters. Interactively obtain preset welding trajectory data, and perform welding trajectory correction based on the welding position parameters and preset welding trajectory data to obtain the welding movement path; The welding robot is jointly controlled by the welding torch displacement trajectory and the welding movement path to complete the automated welding of the steel pipe to be welded; The industrial camera array is used to acquire images of the steel pipe to be welded, and the welding start point is located based on the image acquisition results, including: Pre-constructed image acquisition window; When the robotic arm's stop running time reaches the image acquisition window, the industrial camera array is activated to perform image acquisition on the steel pipe to be welded, obtaining multiple partial steel pipe images, wherein the multiple partial steel pipe images have multiple image source identifiers; A pre-constructed welding start point recognition model is used, and welding start recognition is performed by synchronizing the multiple local steel pipe images to the welding start point recognition model, and a start point image is output, wherein the welding start point is identified in the start point image; Based on the welding start point, the target laser rangefinder is activated in the laser ranging array, and the target laser rangefinder is scheduled to operate independently to spatially locate the welding start point and obtain welding position parameters, including: Based on the image source identifier of the starting image, locate the target camera and the target image acquisition position in the industrial camera array; The target laser rangefinder is positioned in the laser ranging array, wherein the target laser rangefinder is closest to the target image acquisition location in space; The target laser rangefinder is scheduled to operate independently to spatially locate the welding starting point and obtain the welding position parameters. A pre-constructed welding torch displacement analysis model is used to generate a welding torch displacement trajectory based on the welding torch position parameters and welding position parameters, including: Historical displacement data of the automated welding unit is collected to obtain multiple sample displacement trajectories, multiple sample welding torch positions, and multiple sample welding positions. The displacement trajectories of the multiple samples, the welding gun positions of the multiple samples, and the welding positions of the multiple samples are divided to obtain K sample displacement analysis data; A standard displacement analysis model is constructed based on a backpropagation neural network, and the model parameters of the standard displacement analysis model are optimized using the K sample displacement analysis data respectively to obtain K welding gun displacement analysis branches. The K welding torch displacement analysis branches are connected in parallel to complete the construction of the welding torch displacement analysis model; Synchronize the welding torch position parameters and welding position parameters to the welding torch displacement analysis model to obtain K backup displacement trajectories; The K backup displacement trajectories are serialized, and the shortest trajectory is extracted based on the sorting result to obtain the welding torch displacement trajectory; The method further includes: Interactively obtain spatial parameter information of the welding operation space; The welding operation space is modeled and reconstructed based on the spatial parameter information to obtain the operation space model; The standard workpiece parameters of the steel pipe to be welded are obtained interactively, and the steel pipe to be welded is modeled and reconstructed based on the standard workpiece parameters to obtain the standard workpiece model. The initial gripping trajectory of the robotic arm is obtained interactively; The initial clamping trajectory is used to guide the standard workpiece model to fit the operation space model, thereby obtaining an updated space model; Based on the updated spatial model, equipment layout analysis is performed to obtain equipment model information and equipment arrangement information, wherein the equipment arrangement information includes laser ranging arrangement parameters and industrial camera arrangement parameters; Based on the equipment layout information, the equipment is arranged around the welding operation space to obtain the laser ranging array and the industrial camera array.

2. The method as described in claim 1, characterized in that, Based on the updated spatial model, equipment layout analysis is performed to obtain equipment model information and equipment arrangement information, including: A preset slicing direction is used as a constraint to geometrically divide the updated space model, resulting in multiple slice space models. Multiple cross-sectional views are extracted from the multiple slice space models; Based on the multiple cross-sectional views, the straight-line distance between the standard workpiece model and the operating space model is collected to obtain multiple sets of spatial distance parameters; The multiple sets of spatial distance parameters are serialized and their maximum values ​​are extracted to obtain the spatial distance extreme values. The device model information is obtained by traversing the pre-constructed device configuration table using the spatial distance extreme value, wherein the device model information includes a first device model and a second device model; Based on the first device model and the extreme value of spatial distance, the device acquisition range is analyzed to obtain the laser ranging interval distance, and the laser ranging arrangement parameters are obtained by fitting the laser ranging interval distance to the updated spatial model. Similarly, the industrial camera layout parameters are obtained by analyzing the second equipment model.

3. The method as described in claim 1, characterized in that, The target laser rangefinder is scheduled to operate independently to spatially locate the welding starting point and obtain the welding position parameters, including: The target laser rangefinder is scheduled to operate independently to collect the spatial distance between the welding starting point and the target laser rangefinder, thereby obtaining the spatial distance between the welding point and the equipment. The automated welding unit interacts with the target laser rangefinder to obtain the standard coordinate system of the welding operation space and the device coordinate parameters of the target laser rangefinder. The welding position parameters are obtained by transforming the spatial distance between the weld point and the equipment according to the equipment coordinate parameters and the standard coordinate system.

4. The method as described in claim 1, characterized in that, The welding robot is jointly controlled using the welding torch displacement trajectory and welding movement path to complete the automated welding of the steel pipe to be welded, including: The welding robot moves its welding torch tangent to the welding starting point by controlling the welding torch displacement trajectory, and then the welding power supply of the welding robot is activated. The welding robot's welding torch is guided by the welding movement path to move on the surface of the steel pipe to be welded, thereby completing the automated welding of the steel pipe.

Images