Multi-dimensional space fitting mapping manifold groove welding track planning method and system

Abstract

The invention relates to a multi-dimensional space fitting mapping manifold groove welding track planning method and system, and the method achieves the high-precision robust extraction of geometrical characteristics of a main pipe and branch pipes in complex point cloud data through a step-by-step random sampling consistency algorithm based on normal vector estimation. Further combining theoretical intersecting line guiding and self-adaptive distance statistical characteristic screening, and accurately separating the boundary of the inner side and the outer side of the groove; and finally, a high-precision and smooth three-dimensional welding track is automatically generated by utilizing a dimensionality reduction projection and least square arc fitting technology. According to the scheme, the problems that in the prior art, under industrial field interference of oil contamination, oxide skin and the like, the recognition precision is low, the anti-interference capacity is poor, and trajectory planning depends on manual teaching are effectively solved, and the automatic welding efficiency and quality of the collecting pipe are effectively improved.

Description

Technical Field

[0001] This application relates to the field of robot automation technology, and in particular to a method and system for planning welding trajectories for pipe bevels using multi-dimensional spatial fitting and mapping. Background Technology

[0002] Manifolds, as core pipeline connectors in key industrial sectors such as petrochemicals, power generation, and hydropower, typically consist of a main pipe connected to multiple branch pipes in an intersecting manner. The intersecting weld between the main and branch pipes forms a complex three-dimensional spatial curve, and its welding quality directly determines the operational safety and reliability of the entire pipeline system under harsh conditions such as high temperature and high pressure. Therefore, achieving high-quality and high-efficiency welding of manifold welds is a long-standing and crucial technical requirement in this field.

[0003] For a long time, the welding of manifolds has mainly relied on manual operation by highly skilled welders. However, manual welding is easily affected by subjective factors such as the welder's skill level, fatigue state, and on-site environment. In order to overcome the above drawbacks, industrial robot welding technology has emerged and has achieved great success in standardized welding scenarios such as straight seams and circumferential seams of conventional pipelines.

[0004] Currently, mainstream robotic welding systems, such as those from Yaskawa and KUKA, effectively improve welding stability and efficiency by equipping themselves with high-degree-of-freedom robot bodies, welding power sources, and tooling fixtures. However, when these advanced robotic systems are faced with non-standard workpieces such as manifolds with complex three-dimensional intersecting weld seams, existing technologies reveal serious technical bottlenecks and limitations, making it difficult to directly meet the actual needs of automated welding. Summary of the Invention

[0005] To address the aforementioned issues, this application provides a multi-dimensional space fitting mapping method and system for planning welding trajectories of manifold bevels, which enables accurate identification of manifold bevels and automatic generation of high-precision welding trajectories under complex interference.

[0006] To achieve the above objectives, in a first aspect, embodiments of this application provide a method for planning the welding trajectory of a manifold bevel through multi-dimensional spatial fitting mapping, comprising the following steps: S100. Obtain point cloud data of the area where the main pipe and branch pipe of the manifold intersect, and identify the cylindrical geometric parameters of the main pipe and the conical and cylindrical geometric parameters of the branch pipe bevel based on the point cloud data. S200. Extract the bevel boundary points of the intersecting region from the point cloud data, calculate the characteristic distance of each bevel boundary point relative to the surface of the main cylinder, and based on the statistical characteristics of the characteristic distance, separate the bevel boundary points into an inner boundary point sequence and an outer boundary point sequence. S300. Based on the inner boundary point sequence and the outer boundary point sequence, perform circular arc fitting in a two-dimensional plane perpendicular to the main axis, and then map it back to three-dimensional space to generate a welding trajectory for robot welding operations.

[0007] Preferably, step S100 specifically includes: S101. Filter and denoise the point cloud data and estimate the normal vector; S102. Based on the normal vector information, use the random sampling consensus algorithm to perform a first fitting on the point cloud data, extract the cylindrical model parameters of the main pipe, the cylindrical model parameters include the axial direction, center position and radius, and separate the point cloud data into the main pipe point cloud and the remaining point cloud; S103. For the remaining point cloud, use the random sampling consensus algorithm to perform a second fitting, extract the conical model parameters and cylindrical model parameters of the branch pipe welding bevel, the conical model parameters include the vertex position, axial direction and half vertex angle.

[0008] Preferably, before extracting the bevel boundary points in step S200, the following processing steps are also included: S201a. Based on the main pipe cylindrical parameters and the branch pipe conical and cylindrical parameters identified in step S100, the discrete point sequence of the theoretical intersection line is obtained analytically by solving the simultaneous equations of the cylinder equation and the cone equation. S201b. In the point cloud data, perform boundary detection on the point cloud of the branch pipe bevel and extract an initial candidate boundary point set; S201c. Using the theoretical intersection line as a reference, the initial candidate boundary point set is spatially filtered, and points located within a preset distance range around the theoretical intersection line are retained as candidate bevel boundary points for subsequent separation.

[0009] Preferably, step S200 specifically includes: S201. For each candidate bevel boundary point P i Calculate the candidate bevel boundary point P i Calculate the characteristic distance d relative to the surface of the main cylindrical tube. i S202. Calculate the characteristic distance d of all candidate bevel boundary points. i arithmetic mean : Where N is the total number of candidate bevel boundary points; S203. Based on the arithmetic mean A separation threshold is set to filter the candidate bevel boundary points and separate them into the inner boundary point sequence and the outer boundary point sequence.

[0010] Preferably, the feature distance d iThe calculation method includes the following steps: a) Establish a spatial coordinate system such that the Z-axis of the coordinate system coincides with the central axis of the main pipe; b) For each candidate bevel boundary point P i Establish from the candidate slope boundary point P i A ray pointing along the axis of the branch pipe toward the surface of the main cylindrical pipe; c) Calculate the intersection point P between the ray and the surface of the main cylindrical tube. i ′; d) Calculate the candidate bevel boundary point P i Intersection point P i The absolute value of the coordinate difference between ′ and ′ in the Z-axis direction is taken as the feature distance d. i d i =∣P i,z -P i,z ′∣.

[0011] Preferably, the specific method for screening the candidate bevel boundary points in step S203 is as follows: for the outer boundary point sequence, those satisfying condition d are selected. i >d avg Candidate bevel boundary points of -T1 are assigned to the outer boundary point sequence, where T1 is the first preset offset; for the inner boundary point sequence, those satisfying condition d are... avg -T2 <d i <d avg The candidate bevel boundary point of +T2 is included in the inner boundary point sequence, where T2 is the second preset offset and T2>T1.

[0012] Preferably, the first preset offset T1 is 0.3mm~1.0mm, and the second preset offset T2 is 4mm~8mm.

[0013] Preferably, step S300 specifically includes: S301. Establish a two-dimensional projection plane perpendicular to the branch pipe axis, and project the inner boundary point sequence and the outer boundary point sequence onto the two-dimensional projection plane to obtain the two-dimensional projection point set {Q}. k S302. Establish a welding coordinate system, for each point Q in the two-dimensional projection point set. k Determine its two-dimensional local coordinates (x) in the welding coordinate system. k ,y k ) and the corresponding angle parameter θ k ; S303. The least squares method is used to fit the circular arc trajectory by minimizing the objective function. The circular arc parameters (A, ϕ, C, D) are solved using this method. Where A is the radius of the fitted circular arc. Let C be the phase angle, and D be the offset of the center of the circle in the x and y directions, respectively. S304. Based on the arc parameters, generate a first trajectory representing the outer side of the weld bead, a second trajectory representing the inner side of the weld bead, and a robot path trajectory representing the center of the weld bead in the two-dimensional projection plane. S305. The first trajectory, the second trajectory, and the robot path trajectory are reverse-mapped to three-dimensional space to form the welding trajectory.

[0014] Preferably, after step S305, the method further includes a step of precisely correcting the welding trajectory: S306. For each path point R(θ) on the robot's path trajectory, establish a ray pointing from the path point R(θ) into the interior of the main body, the parametric equation of the ray being: P intersection =R(θ)+t·d c , where d c Let t be a unit vector pointing to the interior of the cylinder, and t be a parameter. S307. Calculate the intersection point P between the ray and the surface of the main cylindrical tube. intersection The specific method is as follows: S307a. Define an auxiliary vector: w = R(θ) - C c C c The central point of the supervisor; S307b. Establish a quadratic equation in terms of parameter t: at 2 +bt+c=0, where: a=d c ·d c −(d c ·u c ) 2 b=2[w·d c −(w·u c )(d c ·u c )] c = w·w−(w·u) c ) 2 - Among them, u c r is the unit vector of the main axis. c The radius is the main radius, and · represents the vector dot product operation; S307c. Solve the quadratic equation and select the smallest non-negative solution t; S307d. Substituting t into the ray parameter equation, we obtain the intersection point coordinates P. intersection =R(θ)+t·d c ; S308. Using the intersection point P intersection Replace the trajectory point corresponding to angle θ in the second trajectory to correct the inner trajectory of the weld bead so that it fits the surface of the main cylindrical tube.

[0015] Secondly, embodiments of this application provide a welding trajectory generation system based on the geometric feature extraction of point cloud of manifold bevel, comprising: The point cloud acquisition module is used to acquire point cloud data of the area where the main pipe and branch pipes of the collection pipe intersect; The point cloud processing module is used for preprocessing the point cloud data, extracting geometric features, and separating bevel boundary points; The trajectory planning module is used to generate a welding trajectory based on the bevel boundary points; A control module includes a processor and a memory communicatively connected to the processor. The memory stores computer program instructions that, when executed by the processor, cause the system to perform the method described in any embodiment of the first aspect.

[0016] The multi-dimensional space fitting mapping method and system for manifold bevel welding trajectory planning designed in this application achieves high-precision and robust extraction of the geometric features of main and branch pipes from complex point cloud data through a step-by-step random sampling consensus algorithm based on normal vector estimation. Furthermore, it combines theoretical intersection line guidance and adaptive distance statistical feature filtering to accurately separate the inner and outer boundaries of the bevel. Finally, it utilizes dimensionality reduction projection and least squares circular arc fitting techniques to automatically generate a high-precision, smooth three-dimensional welding trajectory. This scheme effectively solves the problems of low recognition accuracy, poor anti-interference ability, and reliance on manual teaching in existing technologies under industrial site interference such as oil stains and oxide scale, effectively improving the efficiency and quality of automated welding of manifolds. Attached Figure Description

[0017] Figure 1 This is a flowchart of the multi-dimensional space fitting mapping manifold weld trajectory planning method provided in the embodiments of this application.

[0018] Figure 2 This is a schematic diagram of the point cloud identification and trajectory planning results of a manifold bevel according to an embodiment of the present invention. Detailed Implementation

[0019] The preferred embodiments of this application are described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit this application.

[0020] This application provides a method and system for planning welding trajectories for manifold bevels using multidimensional spatial fitting mapping. In a specific application scenario, the method can be executed by a welding trajectory generation system based on the geometric features extracted from the point cloud of the manifold bevel. This system may include, for example, a six-degree-of-freedom industrial robot, a three-dimensional vision sensor such as a structured light camera, and a welding device such as a MAG or TIG welding torch, as well as a computing unit for data processing and control, such as an industrial computer or embedded controller.

[0021] The following reference Figure 1 The steps of the method provided in this embodiment are described in detail.

[0022] In a first aspect, embodiments of this application provide a method for planning the welding trajectory of a manifold bevel by fitting a multi-dimensional space, comprising the following steps: S100. Obtain point cloud data of the intersection area between the main pipe and the branch pipe of the manifold, and identify the cylindrical geometric parameters of the main pipe and the conical and cylindrical geometric parameters of the branch pipe bevel based on the point cloud data. In this step, the three-dimensional point cloud data of the intersection area between the main pipe and the branch pipe is obtained through a three-dimensional vision sensor, such as a structured light camera. The overall point cloud data of the cylindrical surface of the main pipe and the conical surface of the branch pipe obtained and identified are used in... Figure 2 The image is visually represented as a blue point cloud, containing the cylindrical surface features of the main pipe and the conical surface features of the branch pipe bevel. In practical applications, a structured light camera can be mounted on a robot end effector or a stand-alone support, aligned with the intersecting welding area of ​​the manifold pipes, to calculate the three-dimensional coordinates of each pixel by projecting an encoded structured light pattern and capturing reflected images, thereby generating a point cloud dataset containing the three-dimensional coordinates.

[0023] The specific identification process is as follows: S101. Filter and denoise the point cloud data and estimate the normal vector. Specifically, the acquired raw point cloud data can be filtered and denoised to remove stray points caused by sensor noise, environmental reflection, or surface contamination; simultaneously, the normal vector of the point cloud is estimated to provide a basis for subsequent geometric feature extraction. For example, a filter based on statistical principles can be used to remove outliers, and a local plane fitting method can be used to estimate the normal vector of each point.

[0024] S102. Based on the normal vector information obtained in step S101, the point cloud data is fitted for the first time using the Random Sample Consensus (RANSAC) algorithm. The RANSAC algorithm can accurately extract the cylindrical model parameters of the main tube, including the axial direction (e.g., unit vector u). c ), center position (e.g., center point C) c ) and radius (e.g., r) cAfter extracting the cylindrical model of the supervisor, the point cloud data is separated into the supervisor point cloud and the remaining point cloud, that is, the part belonging to the supervisor is removed from the original point cloud.

[0025] S103. For the remaining point cloud separated in step S102, a second fitting is performed using the random sampling consensus algorithm. The goal of this fitting is to extract the conical and cylindrical model parameters of the branch pipe weld bevel: for the conical part of the branch pipe, the extracted conical model parameters include its vertex position (e.g., A). c ), axis direction (e.g., unit vector u) t The parameters of the main pipe and branch pipes are extracted, including the axial direction, center position, and radius. Through these two RANSAC algorithm fittings, accurate geometric models of the main pipe and branch pipes of the manifold can be obtained.

[0026] S200. Extract the bevel boundary points of the intersecting region from the point cloud data, calculate the characteristic distance of each bevel boundary point relative to the surface of the main cylinder, and based on the statistical characteristics of the characteristic distance, separate the bevel boundary points into an inner boundary point sequence and an outer boundary point sequence. This step aims to extract the bevel boundary points of the intersecting region from the point cloud data and, based on the statistical characteristics of their characteristic distances relative to the surface of the main cylinder, separate these boundary points into an inner boundary point sequence and an outer boundary point sequence.

[0027] Before extracting the bevel boundary points, this embodiment also includes the following preprocessing steps: S201a. Calculate the theoretical intersection line: based on the main cylinder parameters (C) identified in step S100. c ,u c ,r c ), and the parameters of the branch pipe cone and cylinder (A) c ,u t By solving the simultaneous equations of the cylinder and cone, a discrete sequence of theoretical intersection points is analytically obtained. This calculated theoretical intersection line provides an accurate reference benchmark for subsequent boundary point spatial range filtering.

[0028] For example: Half-vertex angle calculation: , where angle is the cone apex angle input in degrees.

[0029] Calculation of base radius: Where h is the height parameter of the cone, and in this embodiment, h=1.

[0030] Coordinates of the center of the base: Where V is the vertex of the cone and D is the normalized direction vector.

[0031] The coordinates of discrete points on the circumference of the base are: Where i = 0, 1, 2, ..., n-1, and n is the number of sampling points for the circular discretization.

[0032] Generic vector: Busbar length: Specifically, for the parametric equations of the conical generatrix: r(t) = V + td, t ≥ 0 Where d is the direction vector of the generatrix.

[0033] The equation that the distance from a point on the surface of a cylinder to the axis is equal to the radius is: ||(r(t)-C)-((r(t)-C) T u)u||=r Where C is the center of the cylinder, u is the unit vector of the cylinder axis, and r is the radius of the cylinder.

[0034] Substituting r(t) into the cylinder equation, we can simplify it into a quadratic equation using the projection formula: at 2 +bt+c=0 The formula for calculating the coefficient is: a = d·d - (d·u) 2 b = 2·[w•d-(w·u)·(d·u)] c = w·w - (w·u) 2 -r 2 Where w = VC, and · represents the vector dot product operation.

[0035] By calculating the discriminant: D=b 2 -4ac, if D≥0, solve the quadratic equation to obtain two parametric solutions t1, t2: Substituting t1 and t2 into the equation of the busbar parameters, the coordinates of the intersection point can be calculated: Thus, a discrete point sequence of the theoretical intersection line is obtained.

[0036] S201b. Extracting the initial candidate boundary point set: In the point cloud data, perform boundary detection on the branch pipe bevel point cloud and extract the initial candidate boundary point set. For example, a boundary detection algorithm based on the variation characteristics of the normal vector angle can be used.

[0037] S201c. Spatial Range Filtering: Using the theoretical intersection line as a reference, the initial candidate boundary point set is spatially filtered to retain points located within a preset distance range around the theoretical intersection line, for example, within ±0.5mm around the theoretical intersection line, as candidate bevel boundary points for subsequent separation.

[0038] In specific implementation, step S200 includes: S201. For each candidate bevel boundary point P obtained in step S201c... i Calculate the candidate bevel boundary point P i Calculate the characteristic distance d relative to the surface of the main cylindrical tube. i The specific calculation method is as follows: Establish a spatial coordinate system such that the Z-axis of the coordinate system coincides with the central axis of the main pipe.

[0039] b) For each candidate bevel boundary point P i Establish from the candidate slope boundary point P i A ray pointing along the axis of the branch pipe toward the surface of the main cylindrical pipe.

[0040] c) Calculate the intersection point P between the ray and the surface of the main cylindrical tube. i ′.

[0041] d) Calculate the candidate bevel boundary point P i Intersection point P i The absolute value of the coordinate difference between ′ and ′ in the Z-axis direction is taken as the feature distance d. i , that is, d i =∣P i,z -P i,z ′∣。 This distance reflects the spatial positional relationship of the boundary point relative to the cylindrical surface.

[0042] S202. Traverse each candidate bevel boundary point and calculate the feature distance d of all candidate bevel boundary points. i And obtain their arithmetic mean. The calculation formula is as follows: Where N is the total number of candidate bevel boundary points.

[0043] S203. Based on the arithmetic mean A separation threshold is set to filter the candidate bevel boundary points, separating them into the inner boundary point sequence and the outer boundary point sequence. The weld inner boundary point sequence and the outer boundary point sequence obtained through this precise separation step are... Figure 2 The CCP is represented by a solid green line.

[0044] The specific screening method is as follows: Outer boundary point sequence filtering: For the outer boundary point sequence, those that satisfy condition d will be selected. i >d avg The candidate bevel boundary points of −T1 are included in the outer boundary point sequence, where T1 is the first preset offset, which is preferably 0.3mm~1.0mm, for example 0.5mm.

[0045] Inner boundary point sequence filtering: For the inner boundary point sequence, those that satisfy condition d will be selected. avg -T2 <d i <d avg The candidate bevel boundary point of +T2 is included in the inner boundary point sequence, where T2 is the second preset offset, which is preferably 4mm~8mm, for example 6mm, and T2>T1.

[0046] S300. Based on the inner and outer boundary point sequences, circular arc fitting is performed in a two-dimensional plane perpendicular to the main axis, and then mapped back to three-dimensional space to generate a welding trajectory for robotic welding operations. The specific trajectory generation process is as follows: S301. Establish a two-dimensional projection plane perpendicular to the branch pipe axis, and project the inner boundary point sequence and the outer boundary point sequence onto the two-dimensional projection plane to obtain the two-dimensional projection point set {Q}. k}

[0047] S302. Establish the welding coordinate system {W x W y W z} is defined as follows: Where × represents the vector cross product. This coordinate system uses the cylinder axis as the y-axis and the cone axis as the z-axis to ensure that angle calculations are performed within the cross-section of the cylinder. For each point Q in the set of two-dimensional projection points... k Generated from an arbitrary 3D projection point P, its projection onto a cross section perpendicular to the cylinder axis is: Using the reference projection point Q0 as a reference, determine its two-dimensional local coordinates (x, y, y) in the welding coordinate system. k ,y k ) and the corresponding angle parameter θ k Angular parameter θ k The calculation formula is: At the same time, determine each point Q k Two-dimensional local coordinates (x) in the welding coordinate system k ,y k ): S303. The least squares method is used to fit the circular trajectory. For each target angle θ, within the angle range... Internal selection point set {θ k Q k The parameters (A, ϕ, C, D) of the circular arc are solved by minimizing the following objective function: Where A is the radius of the fitted circular arc. Let be the phase angle, and C and D be the offsets of the center of the circle in the x and y directions, respectively.

[0048] S304. Based on the arc parameters obtained in step S303, generate the first trajectory (P) representing the outer side of the weld bead in the two-dimensional projection plane. outer (θ)), representing the second trajectory (P) on the inner side of the weld bead. inner (θ)) and the robot path trajectory (R(θ)) representing the weld center. Specifically: Outer weld point (far from the center of the cylinder): Where A1>r c , indicating the outer weld bead, Z c It is the z-axis coordinate of the two-dimensional projection plane in the welding coordinate system.

[0049] Inner weld point (near the center of the cylinder): A2 <r c , indicating the inner weld bead.

[0050] Robot path point (weld center): S305. The two-dimensional coordinates of the first trajectory, the second trajectory, and the robot path trajectory generated in step S304 are reverse-mapped to three-dimensional space to form the final three-dimensional welding trajectory used for robot welding operations. The actual robot movement path is defined, and the weld width is: .

[0051] In some embodiments, after step S305, this embodiment further includes a step of precisely correcting the welding trajectory: S306. For each path point R(θ) on the robot path trajectory generated in step S304, establish a ray pointing from the path point R(θ) to the interior of the main body. The parametric equation of the ray is: Pintersection =R(θ)+t·d c Where, d c Let t be a unit vector pointing to the interior of the cylinder, and t be a parameter.

[0052] S307. Calculate the intersection point P between the ray and the surface of the main cylindrical tube. intersection The specific method is as follows: S307a. Define an auxiliary vector: w = R(θ) - C c C c It is the central point for the supervisor.

[0053] S307b. Establish a quadratic equation in terms of parameter t: at 2 +bt+c=0, where: a=d c ·d c −(d·u c ) 2 b=2[w·d c −(w·u c )(d c ·u c )] c = w·w−(w·u) c ) 2 - Among them, u c r is the unit vector of the main axis. c denoted by the radius, and · denotes the vector dot product operation.

[0054] S307c. Solve the quadratic equation and select the smallest non-negative solution t.

[0055] S307d. Substitute the t obtained in step S307c into the ray parameter equation to obtain the intersection point coordinates P. intersection =R(θ)+t·d c .

[0056] S308. The intersection point P obtained by calculating using step S307. intersection The second trajectory, i.e., the trajectory point corresponding to angle θ in the inner weld bead trajectory, is replaced to correct the inner weld bead trajectory, making it fit the surface of the main cylindrical tube and ensuring that the intersection point is located inside the weld bead. The final generated and precisely corrected robot welding trajectory is then... Figure 2 The trajectory, clearly visible as an orange dashed line, can be directly used to guide welding robots in high-precision automated welding operations.

[0057] Secondly, embodiments of this application provide a welding trajectory generation system based on the geometric feature extraction of point cloud of manifold bevel, used to implement the method described in the first aspect above. This system mainly includes: The point cloud acquisition module is used to acquire point cloud data of the area where the main and branch pipes of the manifold intersect. This module can integrate one or more 3D vision sensors, such as structured light cameras and laser scanners, to acquire high-precision, high-density point cloud data.

[0058] The point cloud processing module is used to preprocess the point cloud data, extract geometric features, and separate bevel boundary points. This module can be implemented by software and runs on a general-purpose processor or dedicated hardware.

[0059] The trajectory planning module is used to generate welding trajectories based on the bevel boundary points. This module can also be implemented through software programs and run on general-purpose processors or dedicated hardware.

[0060] The control module includes a processor, such as a CPU, GPU, or FPGA, and a memory communicatively connected to the processor. The memory stores computer program instructions that, when executed by the processor, enable the system to coordinate with the point cloud acquisition module, point cloud processing module, and trajectory planning module to fully execute the method described in any embodiment of the first aspect.

[0061] The multidimensional space fitting mapping method and system for manifold bevel welding trajectory planning provided in this application achieves high-precision and robust extraction of the geometric features of main and branch pipes from complex point cloud data through a step-by-step random sampling consensus algorithm based on normal vector estimation. Furthermore, it combines theoretical intersection line guidance and adaptive distance statistical feature filtering to accurately separate the inner and outer boundaries of the bevel. Finally, it utilizes dimensionality reduction projection and least squares circular arc fitting techniques to automatically generate a high-precision, smooth three-dimensional welding trajectory. This solution effectively solves the problems of low recognition accuracy, poor anti-interference ability, and reliance on manual teaching in existing technologies under industrial site interference such as oil stains and oxide scale, effectively improving the efficiency and quality of automated manifold welding.

[0062] In the description of this application, it should be noted that the terms "vertical", "up", "down", "horizontal", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this application.

[0063] In the description of this application, it should also be noted that, unless otherwise expressly specified and limited, the terms "set," "install," "connect," and "link" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this application according to the specific circumstances.

[0064] Finally, it should be noted that the above descriptions are merely preferred embodiments of this application and are not intended to limit this application. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A method for planning the welding trajectory of a manifold bevel by fitting a multi-dimensional spatial mapping, characterized in that, Includes the following steps: S100. Obtain point cloud data of the area where the main pipe and branch pipe of the manifold intersect, and identify the cylindrical geometric parameters of the main pipe and the conical and cylindrical geometric parameters of the branch pipe bevel based on the point cloud data. S200. Extract the bevel boundary points of the intersecting region from the point cloud data, calculate the characteristic distance of each bevel boundary point relative to the surface of the main cylinder, and based on the statistical characteristics of the characteristic distance, separate the bevel boundary points into an inner boundary point sequence and an outer boundary point sequence. S300. Based on the inner boundary point sequence and the outer boundary point sequence, perform circular arc fitting in a two-dimensional plane perpendicular to the main axis, and then map it back to three-dimensional space to generate a welding trajectory for robot welding operations.

2. The multi-dimensional space fitting mapping manifold bevel welding trajectory planning method according to claim 1, characterized in that, Step S100 specifically includes: S101. Filter and denoise the point cloud data and estimate the normal vector; S102. Based on the normal vector information, use the random sampling consensus algorithm to perform a first fitting on the point cloud data, extract the cylindrical model parameters of the main pipe, the cylindrical model parameters include the axial direction, center position and radius, and separate the point cloud data into the main pipe point cloud and the remaining point cloud; S103. For the remaining point cloud, use the random sampling consensus algorithm to perform a second fitting, extract the conical model parameters and cylindrical model parameters of the branch pipe welding bevel, the conical model parameters include the vertex position, axial direction and half vertex angle.

3. The multi-dimensional space fitting mapping manifold bevel welding trajectory planning method according to claim 1, characterized in that, Before extracting the bevel boundary points in step S200, the following processing steps are also included: S201a. Based on the main pipe cylinder parameters identified in step S100, as well as the branch pipe cone and cylinder parameters, the discrete point sequence of the theoretical intersection line is obtained analytically by solving the simultaneous equations of the main pipe cylinder equation and the cone equation. S201b. In the point cloud data, perform boundary detection on the point cloud of the branch pipe bevel and extract an initial candidate boundary point set; S201c. Using the theoretical intersection line as a reference, the initial candidate boundary point set is spatially filtered, and points located within a preset distance range around the theoretical intersection line are retained as candidate bevel boundary points for subsequent separation.

4. The multi-dimensional space fitting mapping manifold bevel welding trajectory planning method according to claim 1 or 3, characterized in that, Step S200 specifically includes: S201. For each candidate bevel boundary point P i Calculate the candidate bevel boundary point P i Calculate the characteristic distance d relative to the surface of the main cylindrical tube. i S202. Calculate the characteristic distance d of all candidate bevel boundary points. i arithmetic mean : Where N is the total number of candidate bevel boundary points; S203. Based on the arithmetic mean A separation threshold is set to filter the candidate bevel boundary points and separate them into the inner boundary point sequence and the outer boundary point sequence.

5. The multi-dimensional space fitting mapping manifold welding trajectory planning method according to claim 4, characterized in that, The feature distance d i The calculation method includes the following steps: a) Establish a spatial coordinate system such that the Z-axis of the coordinate system coincides with the central axis of the main pipe; b) For each candidate bevel boundary point P i Establish from the candidate slope boundary point P i A ray pointing along the axis of the branch pipe toward the surface of the main cylindrical pipe; c) Calculate the intersection point P between the ray and the surface of the main cylindrical tube. i ′; d) Calculate the candidate bevel boundary point P i Intersection point P i The absolute value of the coordinate difference between ′ and ′ in the Z-axis direction is taken as the feature distance d. i d i =∣P i,z -P i,z ′∣.

6. The multi-dimensional space fitting mapping manifold bevel welding trajectory planning method according to claim 4, characterized in that, The specific method for filtering the candidate bevel boundary points in step S203 is as follows: For the outer boundary point sequence, those that meet condition d are selected. i >d avg Candidate bevel boundary points of -T1 are assigned to the outer boundary point sequence, where T1 is the first preset offset; for the inner boundary point sequence, those satisfying condition d are... avg −T2 <d i <d avg The candidate bevel boundary point of +T2 is included in the inner boundary point sequence, where T2 is the second preset offset and T2>T1.

7. The multidimensional space fitting mapping manifold bevel welding trajectory planning method according to claim 6, characterized in that, The first preset offset T1 is 0.3mm~1.0mm, and the second preset offset T2 is 4mm~8mm.

8. The multi-dimensional space fitting mapping manifold bevel welding trajectory planning method according to claim 1, characterized in that, Step S300 specifically includes: S301. Establish a two-dimensional projection plane perpendicular to the branch pipe axis, and project the inner boundary point sequence and the outer boundary point sequence onto the two-dimensional projection plane to obtain the two-dimensional projection point set {Q}. k S302. Establish a welding coordinate system, for each point Q in the two-dimensional projection point set. k Determine its two-dimensional local coordinates (x) in the welding coordinate system. k ,y k ) and the corresponding angle parameter θ k ; S303. The least squares method is used to fit the circular arc trajectory by minimizing the objective function. The circular arc parameters (A, ϕ, C, D) are solved using this method. Where A is the radius of the fitted circular arc. Let C be the phase angle, and D be the offset of the center of the circle in the x and y directions, respectively. S304. Based on the arc parameters, generate a first trajectory representing the outer side of the weld bead, a second trajectory representing the inner side of the weld bead, and a robot path trajectory representing the center of the weld bead in the two-dimensional projection plane. S305. The first trajectory, the second trajectory, and the robot path trajectory are reverse-mapped to three-dimensional space to form the welding trajectory.

9. The multidimensional space fitting mapping manifold bevel welding trajectory planning method according to claim 8, characterized in that, Following step S305, the method further includes a step of precisely correcting the welding trajectory: S306. For each path point R(θ) on the robot's path trajectory, establish a ray pointing from the path point R(θ) into the interior of the main body, the parametric equation of the ray being: P intersection =R(θ)+t·d c , where d c Let t be a unit vector pointing to the interior of the cylinder, and t be a parameter. S307. Calculate the intersection point P between the ray and the surface of the main cylindrical tube. intersection The specific method is as follows: S307a. Define an auxiliary vector: w = R(θ) - C c C c The central point of the supervisor; S307b. Establish a quadratic equation in terms of parameter t: at 2 +bt+c=0, where: a=d c ·d c −(d c ·u c ) 2 b=2[w·d c −(w·u c )(d c ·u c )] c=w·w−(w·u c ) 2 − Among them, u c r is the unit vector of the main axis. c The radius is the main radius, and · denotes the vector dot product operation; S307c. Solve the quadratic equation and select the smallest non-negative solution t; S307d. Substituting t into the ray parameter equation, we obtain the intersection point coordinates P. intersection =R(θ)+t·d c ; S308. Using the intersection point P intersection Replace the trajectory point corresponding to angle θ in the second trajectory to correct the inner trajectory of the weld bead so that it fits the surface of the main cylindrical tube.

10. A multi-dimensional spatial fitting mapping manifold bevel welding trajectory planning system, characterized in that, include: The point cloud acquisition module is used to acquire point cloud data of the area where the main pipe and branch pipes of the collection pipe intersect; The point cloud processing module is used for preprocessing the point cloud data, extracting geometric features, and separating bevel boundary points; The trajectory planning module is used to generate a welding trajectory based on the bevel boundary points; A control module includes a processor and a memory communicatively connected to the processor, the memory storing computer program instructions that, when executed by the processor, cause the system to perform the method as described in any one of claims 1 to 9.

Images