TKY pipe joint intersecting line welding locating and self-adaptive deviation rectifying method

Abstract

The invention relates to the technical field of ocean engineering automatic welding, in particular to a TKY pipe node intersecting line welding locating and self-adaptive deviation rectifying method which comprises the following steps that ocean environment data are collected, a sea wave slope reflection disturbance potential function is established based on the ocean environment data, and a laser scanning error distribution field is generated through calculation; on the basis of welding seam point cloud data generated through robot laser scanning, abnormal points in the welding seam point cloud data are screened in combination with the laser scanning error distribution field, and a to-be-corrected point cloud data set is formed; defining a point cloud correction target energy functional, and iteratively optimizing the to-be-corrected point cloud data set in combination with a sea wave slope reflection disturbance potential function to obtain corrected welding seam point cloud data; and executing the robot to perform automatic welding based on the corrected weld point cloud data. According to the TKY pipe joint intersecting line welding locating and self-adaptive deviation correcting method, the actual situation that errors are generated by welding seam point cloud due to sunlight reflected by the sea surface is avoided.

Description

Technical Field

[0001] The present invention relates to the technical field of offshore engineering automated welding, and specifically, to a method for finding the intersection line and self-adaptive deviation correction of TKY pipe joint intersection line welding. Background Technique

[0002] The method for finding the intersection line and self-adaptive deviation correction of TKY pipe joint intersection line welding aims to improve the weld recognition accuracy and reduce the point cloud error in the marine environment. By constructing a disturbance potential function of the sea wave inclined plane reflection and combining with the laser scanning error distribution field, the point cloud data is corrected, and the weld positioning accuracy and the trajectory adjustment stability of the welding robot under the interference of sea waves are controlled, so as to realize the self-adaptive correction of the weld point cloud data in the dynamic marine environment.

[0003] The existing methods for finding the intersection line and self-adaptive deviation correction of TKY pipe joint intersection line welding usually have difficulty in distinguishing the influence of the dynamic inclination of sea waves on the laser scanning error caused by the specular reflection of sunlight. And due to the fact that in the marine environment, the sea wave fluctuations form a specific inclined plane that reflects sunlight into the weld seam, the strong sunlight will fade the clarity of the weld seam edge, resulting in errors when the welding robot scans the weld point cloud in real time. Therefore, a method for finding the intersection line and self-adaptive deviation correction of TKY pipe joint intersection line welding is provided. Summary of the Invention

[0004] The purpose of the present invention is to provide a method for finding the intersection line and self-adaptive deviation correction of TKY pipe joint intersection line welding to solve the problem that in the marine environment, the sea wave fluctuations form a specific inclined plane that reflects sunlight into the weld seam, and the strong sunlight will fade the clarity of the weld seam edge, resulting in errors when the welding robot scans the weld point cloud in real time as proposed in the above background technique.

[0005] To achieve the above purpose, the present invention aims to provide a method for finding the intersection line and self-adaptive deviation correction of TKY pipe joint intersection line welding, including the following steps:

[0006] S1. Use a sensor to collect marine environment data, establish a disturbance potential function of the sea wave inclined plane reflection based on the marine environment data, and calculate and generate a laser scanning error distribution field;

[0007] S2. Based on the weld point cloud data generated by the robot laser scanning, and combining with the laser scanning error distribution field, screen out the abnormal points in the weld point cloud data to form a point cloud data set to be corrected;

[0008] S3. Define a point cloud correction target energy functional, and combine it with the disturbance potential function of the sea wave inclined plane reflection to iteratively optimize the point cloud data set to be corrected to obtain the corrected weld point cloud data;

[0009] S4. Based on the corrected weld point cloud data, execute the robot for automatic welding.

[0010] As a further improvement of this technical solution, the marine environment data includes wave height, wave tilt angle, wave period, wave propagation speed, solar azimuth angle, solar intensity, and seawater refractive index;

[0011] The wave slope reflection perturbation potential function is constructed based on the nonlinear wave equation and considering the influence of the seawater wave - formed slope reflecting sunlight on the welding point, and is used to describe the dynamic interference caused by the seawater wave reflecting sunlight during the scanning of the welding point.

[0012] As a further improvement of this technical solution, in S1, the nonlinear wave equation is used to establish the wave slope reflection perturbation potential function and calculate the laser scanning error distribution field. The specific method steps are as follows:

[0013] S1.1: Based on the marine environment data, define and construct the wave slope reflection perturbation potential function;

[0014] S1.2: Based on the wave slope reflection perturbation potential function, by calculating the seawater wave tilt angle and the solar light reflection angle, obtain the laser scanning error distribution field;

[0015] Among them, the laser scanning error distribution field is used to quantitatively describe the laser scanning error caused by the seawater wave slope reflecting sunlight, and is used to identify and correct the error point data.

[0016] As a further improvement of this technical solution, in S1.1, based on the marine environment data, define and construct the wave slope reflection perturbation potential function as follows:

[0017] Define the wave slope reflection perturbation potential function as Φ(x, y, z, t):

[0018]

[0019] Among them, t is time; x is the horizontal coordinate of the sea - wave surface; y is the coordinate of the sea - wave surface perpendicular to the propagation direction; z is the height coordinate of the sea - wave surface; is the second - order partial derivative of the wave slope reflection perturbation potential function with respect to time; is the Laplace operator; c is the wave propagation speed; γ is the high - order nonlinear coupling coefficient; η is the nonlinear damping coefficient; is the modulus of the wave potential energy gradient; is the partial derivative of the wave slope reflection perturbation potential function with respect to time;

[0020] In S1.2, based on the wave slope reflection perturbation potential function, by calculating the seawater wave tilt angle and the solar light reflection angle, obtain the laser scanning error distribution field as follows:

[0021]

[0022] Where, ψ(x, y, z, t) is the laser scanning error distribution field; is the partial derivative of the potential function of the wave oblique reflection disturbance with respect to height; is the partial derivative of the wave oblique reflection disturbance potential function with respect to time; κ is the wave tilt angle influence coefficient; δ is the time dynamic influence coefficient.

[0023] As a further improvement of the present technical solution, in S2, based on the weld point cloud data generated by robot laser scanning, abnormal points in the weld point cloud data are screened out in combination with the laser scanning error distribution field to form a point cloud data set to be corrected. The specific method steps are as follows:

[0024] S2.1. The weld point cloud data generated by robot laser scanning is:

[0025]

[0026] Among them, M is the total number of point data in the weld point cloud data; i is the point data index; A i is the horizontal coordinate of point data i; B i is the coordinate of point data i perpendicular to the propagation direction; C i is the height coordinate of point data i; (A i ,B i ,C i ) is the three-dimensional coordinate of point data i; Q(t) is the weld point cloud data collected at time t;

[0027] S2.2. Use the laser scanning error distribution field to filter out abnormal points in the weld point cloud data to form a point cloud data set to be corrected.

[0028] As a further improvement of this technical solution, in S2.2, the laser scanning error distribution field is used to screen out abnormal points in the weld point cloud data to form a point cloud data set to be corrected. The specific method steps are as follows:

[0029] S2.2.1. Calculate the theoretical error of each point in the weld point cloud data based on the laser scanning error distribution field:

[0030]

[0031] Among them, ∈ i is the theoretical error of point data i; λ1 is the effect of wave height change on the error; λ2 is the effect of time dynamic change on the error; is the tilt effect of the wave slope on the data point i; is the impact of the ocean wave disturbance on the point data i in the time dimension;

[0032] S2.2.2. Set the error threshold τ, and filter out the point data with errors greater than the error threshold τ to form a point cloud data set to be corrected:

[0033] Q error (t) = {(A i , B i , C i ) | ∈ i > τ};

[0034] Among them, Q error (t) is the point cloud data set to be corrected.

[0035] As a further improvement of this technical solution, the point cloud correction target energy functional is constructed based on the sea wave slope reflection perturbation potential function combined with the point cloud data set to be corrected, and is used to correct the error of the weld point cloud data caused by the reflection of sunlight by the sea water ripples.

[0036] As a further improvement of this technical solution, in S3, define the point cloud correction target energy functional, and combine it with the sea wave slope reflection perturbation potential function to iteratively optimize the point cloud data set to be corrected to obtain the corrected weld point cloud data. The specific method steps are as follows:

[0037] S3.1. Define the point cloud correction target energy functional;

[0038] S3.2. Use the gradient descent algorithm to calculate the gradient descent direction of the point cloud data to be corrected;

[0039] S3.3. Combine the sea wave slope reflection perturbation potential function to correct the error points and calculate the offset correction amount of the error points;

[0040] S3.4. Based on the offset correction amount of the error points, obtain the final correction formula;

[0041] S3.5. Iteratively execute S3.1 - S3.4 to optimize all the point cloud data to be corrected, and finally obtain the corrected weld point cloud data.

[0042] As a further improvement of this technical solution, in S3.1, define the point cloud correction target energy functional. The specific method is as follows:

[0043]

[0045] Among them, is the point cloud correction target energy functional; Q is the point cloud data set to be corrected, equivalent to Q error (t); (X i , Y i , Z i ) is the current coordinate of the point data to be corrected; (M i,N i ,P i ) is the standard point in the ideal weld model; w(|ψ(X i ,Y i ,Z i ,θ)|) suppresses the error term; Ω(Q) is the smoothing regularization term; α1 is the deviation weight coefficient between the point cloud data and the ideal weld model; α2 is the error field weight coefficient; α3 is the point cloud smoothing regularization weight coefficient; θ is the number of optimization iterations; K is the total number of error points in the point cloud data set to be corrected;

[0046] In the above S3.2, the gradient descent algorithm is used to calculate the gradient descent direction of the point cloud data to be corrected. The specific method is as follows:

[0047]

[0048] Among them, is the partial derivative in the gradient descent algorithm;

[0049]

[0050] Among them, λ is the number of iteration rounds; is the point cloud coordinate of the λ-th round of iteration; is the point cloud coordinate of the (λ + 1)-th round of iteration; β is the learning rate.

[0051] As a further improvement of this technical solution, in the above S3.3, the error points are corrected by combining the wave slope reflection perturbation potential function, and the offset correction amount of the error points is calculated. The specific method is as follows:

[0052]

[0053] Among them, Δ i is the offset correction amount of point data i; μ1 is the influence coefficient of seawater ripple perturbation in the horizontal direction; μ2 is the influence coefficient of sea wave perturbation in the time dimension; is the influence of the wave slope reflection perturbation potential function of the sea wave in the horizontal direction of the point cloud; is the influence of the wave slope reflection perturbation potential function of the sea wave in the time dimension;

[0054] In the above S3.4, based on the offset correction amount of the error points, the final correction formula is obtained. The specific method is as follows:

[0055]

[0056] In the above S3.5, S3.1 - S3.4 are iteratively executed to optimize all the point cloud data to be corrected, and finally the corrected weld point cloud data is obtained. The specific method is as follows:

[0057]

[0058] Among them, Q corrected (t) is the corrected weld point cloud data.

[0059] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0060] 1. In the TKY pipe joint intersection line welding positioning and adaptive deviation correction method, based on the joint modeling of the sea wave inclined plane reflection disturbance potential function and the laser scanning error distribution field, the point cloud error caused by the specular reflection interference of sunlight due to the dynamic inclination of sea waves is identified, and the weld point cloud data is corrected in real time.

[0061] 2. In the TKY pipe joint intersection line welding positioning and adaptive deviation correction method, by constructing a point cloud correction target energy functional, the point cloud data greatly affected by the dynamic interference of sea waves is adaptively optimized to ensure the smoothness of the point cloud data within a local range and the overall consistency, enabling the welding robot to accurately position on the actual weld track. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] Figure 1 It is the overall method flow chart of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0063] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0064] Embodiment:

[0065] Please refer to Figure 1 As shown, this embodiment provides a TKY pipe joint intersection line welding positioning and adaptive deviation correction method, including the following steps:

[0066] S1. Use a sensor to collect ocean environment data, establish a sea wave inclined plane reflection disturbance potential function based on the ocean environment data, and calculate and generate a laser scanning error distribution field;

[0067] The ocean environment data includes sea wave height, sea wave inclination angle, wave period, sea wave propagation speed, sunlight azimuth angle, sunlight intensity, and seawater refractive index;

[0068] The sea wave inclined plane reflection disturbance potential function is constructed based on the nonlinear wave equation and considering the influence of the inclined plane reflection of sunlight formed by sea waves on the welding point, and is used to describe the dynamic interference caused by the reflection of sunlight by sea waves when scanning the welding point.

[0069] In this embodiment, sensors are used to collect marine environment data, specifically as follows:

[0070] A lidar is used to obtain weld point cloud data and to monitor the surface morphology of ocean waves in real time, including the wave tilt angle and wave height; an inertial measurement unit is used to detect the wave propagation speed; an ultrasonic ranging sensor is used to measure the wave height and period, which is installed near the robot, and the water surface height is measured using ultrasonic echoes to analyze the undulation of the waves; an ambient light sensor monitors the azimuth angle and light intensity of sunlight; an underwater optical refractive index sensor measures the seawater refractive index and monitors the optical refraction characteristics of seawater to calculate the laser beam path offset.

[0071] In step S1 of this embodiment, a non - linear wave equation is used to establish a disturbance potential function for the inclined surface reflection of ocean waves and to calculate the laser scanning error distribution field. The specific method steps are as follows:

[0072] S1.1: Based on the marine environment data, define and construct a disturbance potential function for the inclined surface reflection of ocean waves;

[0073] S1.2: Based on the disturbance potential function for the inclined surface reflection of ocean waves, obtain the laser scanning error distribution field by calculating the seawater wave tilt angle and the sunlight reflection angle;

[0074] Among them, the laser scanning error distribution field is used to quantitatively describe the laser scanning error caused by the inclined surface reflection of sunlight by seawater waves and is used to identify and correct the error point data.

[0075] In S1.1, based on the marine environment data, define and construct a disturbance potential function for the inclined surface reflection of ocean waves, specifically as follows:

[0076] The measured marine environment data are: wave propagation speed 1.5 m / s; wave height 1.2 m; wave period 8 s; sunlight azimuth angle 35°; sunlight intensity 600 W / m 2 ; seawater refractive index 1.33; non - linear damping coefficient 0.1; high - order non - linear coupling coefficient 0.02;;

[0077] Define the disturbance potential function for the inclined surface reflection of ocean waves as Φ(x, y, z, t), which satisfies the following non - linear wave equation:

[0078]

[0079] Among them, t is time; x is the horizontal coordinate of the wave surface; y is the coordinate perpendicular to the propagation direction of the wave surface; z is the height coordinate of the wave surface; is the second - order partial derivative of the disturbance potential function for the inclined surface reflection of ocean waves with respect to time; is the Laplace operator; is the modulus of the gradient of the wave potential energy; is the partial derivative of the disturbance potential function of the sea wave slope reflection with respect to time;

[0080] In S1.2, based on the disturbance potential function of the sea wave slope reflection, by calculating the sea wave tilt angle and the solar light reflection angle, the laser scanning error distribution field is obtained as follows:

[0081]

[0082] where ψ(x, y, z, t) is the laser scanning error distribution field; is the partial derivative of the disturbance potential function of the sea wave slope reflection with respect to height; is the partial derivative of the disturbance potential function of the sea wave slope reflection with respect to time; κ is the sea wave tilt angle influence coefficient; δ is the time dynamic influence coefficient;

[0083]

[0084] At the moment of the maximum wave tilt (t = 2 s), calculate its specific value:

[0085]

[0086] Sea wave tilt angle influence coefficient:

[0087]

[0088] Time dynamic influence coefficient:

[0089]

[0090] Substitute specific data for calculation:

[0091] ψ(x, y, z, t) = (1.65)×(0.848)+(1.03×10 - 6)×(1.2)×(0.665);

[0092] ψ(x, y, z, t) = 1.355 + 0.000000824;

[0093] Finally, it is calculated that: the maximum error distribution value is 1.355 m;

[0094] Since the error amplitude reaches 1.355 m, it will cause a large - scale deviation of the weld point cloud. Therefore, it is necessary to introduce the point cloud correction target energy functional to optimize the data collected by the welding robot and reduce its error to an acceptable range (±0.1 mm);

[0095] The error point screening threshold is set to τ = 0.5 m;

[0096] In this embodiment, in a marine environment, sea waves will form a dynamic inclined plane, which will cause an optical deviation of the laser scanning path. Let the perturbation potential function of the sea surface, that is, the perturbation potential function of the sea wave inclined plane reflection, be Φ(x, y, z, t), which satisfies the nonlinear wave equation. This nonlinear wave equation describes the time evolution of the sea surface morphology and is used to calculate the perturbation of the sea waves on the laser path. The error of laser scanning can be calculated from the sea surface inclination angle and the solar reflection angle to obtain the laser scanning error distribution field ψ(x, y, z, t).

[0097] In a marine environment, when a welding robot uses laser scanning to find the weld seam, the dynamic inclined plane formed by the sea wave fluctuations will cause the solar reflection to interfere with the laser scanning, resulting in a systematic deviation in the point cloud data. In standard water wave dynamics, sea waves can be regarded as a nonlinear wave system controlled by the Laplace equation and boundary conditions. However, due to factors such as turbulence, surface tension, and viscous damping in sea waves, the linear wave equation cannot accurately describe these complex phenomena. Therefore, it is necessary to introduce the nonlinear wave equation. Using the nonlinear wave equation can accurately simulate the dynamics of sea waves without relying on empirical filtering methods, thus having stronger generalization ability. By calculating the error distribution through the laser scanning error distribution field ψ(x, y, z, t), the point cloud data of the robot can be directly corrected adaptively to improve the weld seam positioning accuracy. Since the nonlinear term is considered, the model can adapt to extreme situations such as strong winds and waves and wave breaking.

[0098] In this embodiment, κ is the influence coefficient of the sea wave inclination angle, and is calculated as: where n is the refractive index of seawater, ΔL is the optical path offset; δ is the time dynamic influence coefficient, and is calculated as: where λ L is the laser wavelength, and c is the propagation speed of the sea waves.

[0099] S2. Based on the weld seam point cloud data generated by the robot laser scanning, and combined with the laser scanning error distribution field, screen out the abnormal points in the weld seam point cloud data to form a point cloud data set to be corrected;

[0100] In this embodiment S2, based on the weld seam point cloud data generated by the robot laser scanning, and combined with the laser scanning error distribution field, screen out the abnormal points in the weld seam point cloud data to form a point cloud data set to be corrected. The specific method steps are as follows:

[0101] S2.1. The weld seam point cloud data generated by the robot laser scanning is:

[0102]

[0103] where M is the total number of point data in the weld seam point cloud data; i is the point data index; A i is the horizontal coordinate of the point data i; Bi is the coordinate of point data i perpendicular to the propagation direction; C i is the height coordinate of point data i; (A i , B i , C i ) is the three-dimensional coordinate of point data i; Q(t) is the weld seam point cloud data collected at time t;

[0104] Robot scanning parameters:

[0105] Scanning method: Laser vision scanning

[0106] Scanning device: LiDAR + vision sensor

[0107] Scanning trajectory: Line-by-line scanning generated by offline programming software

[0108] Scanning accuracy: 1 point is collected per 1 mm

[0109] Scanning range:

[0110] Horizontal direction A i : [-0.5 m, 0.5 m]; Vertical direction B i : [-0.5 m, 0.5 m]; Height direction C i : [-0.1 m, 0.1 m];

[0111] The robot program triggers the laser vision sensor to perform weld seam point cloud scanning. According to the scanning trajectory preset by the offline programming software, laser point cloud data on the weld seam surface is obtained point by point to form a complete three-dimensional weld seam point cloud data set Q(t).

[0112] In this embodiment, the weld seam point cloud data generated by robot laser scanning is a commonly used technical means in the prior art. The specific method for obtaining the weld seam point cloud data is as follows:

[0113] The robot program triggers the laser vision sensor to start positioning and scans the weld seam according to the laser vision scanning trajectory established by the offline programming software to obtain the weld seam point cloud data;

[0114] S2.2. Use the laser scanning error distribution field to screen out the abnormal points in the weld seam point cloud data to form a point cloud data set to be corrected.

[0115] In step S2.2 of this embodiment, the laser scanning error distribution field is used to screen out the abnormal points in the weld seam point cloud data to form a point cloud data set to be corrected. The specific method steps are as follows:

[0116] S2.2.1. Calculate the theoretical error of each point data in the weld seam point cloud data based on the laser scanning error distribution field:

[0117]

[0118] where, ∈ i is the theoretical error of point data i; λ1 is the influence of the change in sea wave height on the error; λ2 is the influence of the time dynamic change on the error; is the inclination influence of the sea wave slope on point data i; is the influence of the sea wave disturbance in the time dimension on point data i;

[0119] where, λ1 is the influence of the change in sea wave height on the error, with a value of 0.8; λ2 is the influence of the time dynamic change on the error, with a value of 1.2;

[0120] S2.2.2. Set the error threshold τ, and screen out the point data with errors greater than the error threshold τ to form the point cloud data set to be corrected:

[0121] Q error (t) = {(A i , B i , C i ) | ∈ i > τ};

[0122] where, Q error (t) is the point cloud data set to be corrected;

[0123] Set the error threshold τ to 0.5m; screen out the point cloud data with excessive errors; substitute the maximum error:

[0124] ∈ max = 1.355m > τ = 0.5m;

[0125] This point belongs to the abnormal point and should be added to the point cloud data set to be corrected.

[0126] S3. Define the point cloud correction target energy functional, and combine it with the sea wave slope reflection perturbation potential function to iteratively optimize the point cloud data set to be corrected to obtain the corrected weld point cloud data;

[0127] The point cloud correction target energy functional is constructed based on the sea wave slope reflection perturbation potential function combined with the point cloud data set to be corrected, and is used to correct the error of the weld point cloud data caused by the reflection of sunlight by the sea water ripples.

[0128] In this embodiment S3, define the point cloud correction target energy functional, and combine it with the sea wave slope reflection perturbation potential function to iteratively optimize the point cloud data set to be corrected to obtain the corrected weld point cloud data. The specific method steps are as follows:

[0129] S3.1. Define the point cloud correction target energy functional;

[0130] S3.2. Use the gradient descent algorithm to calculate the gradient descent direction of the point cloud data to be corrected;

[0131] S3.3. Combine the wave slope reflection perturbation potential function to correct the error points and calculate the offset correction amount of the error points;

[0132] S3.4. Obtain the final correction formula based on the offset correction amount of the error points;

[0133] S3.5. Iteratively execute S3.1 - S3.4 to optimize all the point cloud data to be corrected, and finally obtain the corrected weld point cloud data.

[0134] In this embodiment, the purpose of the point cloud correction target energy functional is to construct an optimization framework so that the error point cloud data scanned by the robot can be automatically corrected and approach the true weld form. Due to the wave characteristics of the sea, the weld point cloud data may have local offsets, distortions, or outliers. The point cloud correction target energy functional quantifies the errors and uses gradient descent optimization to make the error points converge towards the true weld trajectory, thereby eliminating the laser scanning errors;

[0135] In traditional point cloud correction methods, simple interpolation or filtering is usually used to remove the error points without considering the dynamic interference characteristics of the sea waves. This method constructs the point cloud correction target energy functional based on the wave slope reflection perturbation potential function, which can be adaptively adjusted according to the sea wave dynamics and is more in line with the actual situation of the marine welding environment;

[0136] During the point cloud correction process, it is not enough to simply make each point converge to its ideal position. The overall consistency of the entire point cloud data also needs to be considered. Therefore, a smoothing regularization term Ω(Q) is added to the point cloud correction target energy functional to constrain the change trend of the point cloud data.

[0137] In step S3.1 of this embodiment, the point cloud correction target energy functional is defined as follows:

[0138]

[0139] Among them, is the point cloud correction target energy functional; Q is the point cloud data set to be corrected, equivalent to Q error (t); (X i , Y i , Z i ) is the current coordinate of the point data to be corrected; (M i , N i , P i ) is the standard point in the ideal weld model; w(|ψ(X i , Y i , Z i, λ)|) is the suppression error term; Ω(Q) is the smoothing regularization term; α1 is the deviation weight coefficient between the point cloud data and the ideal weld model; α2 is the error field weight coefficient; α3 is the point cloud smoothing regularization weight coefficient; θ is the number of optimization iterations; K is the total number of error points in the point cloud data set to be corrected;

[0140] Among them, α1 is the deviation weight coefficient between the point cloud data and the ideal weld model, with a value of 1.5; α2 is the error field weight coefficient, with a value of 0.8; α3 is the point cloud smoothing regularization weight coefficient, with a value of 0.5; θ is the number of optimization iterations, with a value of 30;

[0141] In this embodiment S3.2, the gradient descent algorithm is used to calculate the gradient descent direction of the point cloud data to be corrected. The specific method is as follows:

[0142] [[ID=1)]]

[0143] This formula is the calculation of gradient descent;

[0144] Among them, is the partial derivative in the gradient descent algorithm;

[0145] The correction direction of each error point is determined by its gradient. The update formula is as follows:

[0146]

[0147] Among them, λ is the number of iteration rounds; is the point cloud coordinate of the λ-th round of iteration; is the point cloud coordinate of the (λ + 1)-th round of iteration; β is the learning rate, with a value of 0.01.

[0148] In this embodiment S3.3, the error points are corrected by combining the wave slope reflection perturbation potential function, and the offset correction amount of the error points is calculated. The specific method is as follows:

[0149]

[0150] Among them, Δ i is the offset correction amount of point data i; μ1 is the influence coefficient of seawater ripple perturbation in the horizontal direction, with a value of 0.7; μ2 is the influence coefficient of sea wave perturbation in the time dimension, with a value of 0.9; is the influence of the wave slope reflection perturbation potential function of the point cloud in the horizontal direction; is the influence of the wave slope reflection perturbation potential function of the point cloud in the time dimension;

[0151] In this embodiment S3.4, based on the offset correction amount of the error points, the final correction formula is obtained. The specific method is as follows:

[0152]

[0153] In this embodiment S3.5, S3.1 - S3.4 are iteratively executed to optimize all the point cloud data to be corrected, and finally the corrected weld point cloud data is obtained. The specific method is as follows:

[0154]

[0155] Among them, Q corrected (t) is the corrected weld point cloud data.

[0156] In this embodiment, the finally obtained corrected weld point cloud data includes the point cloud data corrected from the point cloud data set to be corrected plus the point cloud data that originally meets the standards.

[0157] S4. Perform automatic welding by the robot based on the corrected weld point cloud data.

[0158] Application example:

[0159] Implementation process

[0160] S1. Use a sensor to collect ocean environment data, establish a wave slope reflection perturbation potential function based on the ocean environment data, and calculate and generate a laser scanning error distribution field;

[0161] The ocean environment data includes wave height, wave tilt angle, wave period, wave propagation speed, solar azimuth angle, solar intensity, and seawater refractive index;

[0162] The wave slope reflection perturbation potential function is constructed based on the nonlinear wave equation and considering the influence of seawater waves forming a slope to reflect sunlight on the welding point, and is used to describe the dynamic interference caused by the reflection of seawater waves on sunlight when scanning the welding point.

[0163] In this embodiment, using a sensor to collect ocean environment data is as follows:

[0164] Use a lidar to obtain the weld point cloud data and real - time monitor the surface morphology of the waves, including the wave tilt angle and wave height; use an inertial measurement unit to detect the wave propagation speed; use an ultrasonic ranging sensor to measure the wave height and period, installed near the robot, measure the water surface height using ultrasonic echo, and analyze the undulation of the waves; use an ambient light sensor to monitor the solar azimuth angle and light intensity; use an underwater optical refractive index sensor to measure the seawater refractive index and monitor the optical refraction characteristics of the seawater to calculate the laser light path offset.

[0165] In this embodiment S1, use the nonlinear wave equation to establish the wave slope reflection perturbation potential function and calculate the laser scanning error distribution field. The specific method steps are as follows:

[0166] S1.1. Define and construct a perturbation potential function for the reflection of ocean waves on the inclined surface based on ocean environmental data;

[0167] S1.2. Based on the perturbation potential function for the reflection of ocean waves on the inclined surface, calculate the inclination angle of seawater waves and the solar light reflection angle to obtain the laser scanning error distribution field;

[0168] Among them, the laser scanning error distribution field is used to quantitatively describe the laser scanning error caused by the reflection of solar light on the inclined surface of seawater waves, and is used to identify and correct the error point data.

[0169] In S1.1, based on ocean environmental data, define and construct a perturbation potential function for the reflection of ocean waves on the inclined surface, specifically as follows:

[0170] Measured ocean environmental data: wave propagation speed: 1.5 m / s; wave height: 1.2 m; wave period: 8 s; solar azimuth angle: 35°; solar intensity: 600 W / m 2 ; seawater refractive index: 1.33; nonlinear damping coefficient: 0.1; high-order nonlinear coupling coefficient: 0.02;

[0171] Define the perturbation potential function for the reflection of ocean waves on the inclined surface as Φ(x, y, z, t), which satisfies the following nonlinear wave equation:

[0172]

[0173] Among them, t is time; x is the horizontal coordinate of the ocean wave surface; y is the coordinate of the ocean wave surface perpendicular to the propagation direction; z is the height coordinate of the ocean wave surface; is the second-order partial derivative of the perturbation potential function for the reflection of ocean waves on the inclined surface with respect to time; is the Laplace operator; c is the wave propagation speed; γ is the high-order nonlinear coupling coefficient; η is the nonlinear damping coefficient; is the modulus of the potential energy gradient of ocean waves; is the partial derivative of the perturbation potential function for the reflection of ocean waves on the inclined surface with respect to time; <>

[0174] In S1.2, based on the perturbation potential function for the reflection of ocean waves on the inclined surface, calculate the inclination angle of seawater waves and the solar light reflection angle to obtain the laser scanning error distribution field, specifically as follows:

[0175]

[0176] Among them, ψ(x, y, z, t) is the laser scanning error distribution field; is the partial derivative of the perturbation potential function for the reflection of ocean waves on the inclined surface with respect to height; is the partial derivative of the perturbation potential function for the reflection of ocean waves on the inclined surface with respect to time; κ is the influence coefficient of the wave inclination angle; δ is the influence coefficient of time dynamics.

[0177] In this embodiment, in a marine environment, seawater waves will form a dynamic inclined plane, which will cause an optical deviation of the laser scanning path. Let the perturbation potential function of the sea surface, that is, the perturbation potential function of the reflection of the sea wave inclined plane, be Φ(x, y, z, t), which satisfies the nonlinear wave equation. This nonlinear wave equation describes the time evolution of the sea surface morphology and is used to calculate the perturbation of the laser path by the sea waves. The error of laser scanning can be calculated from the inclination angle of the sea surface and the reflection angle of sunlight to obtain the laser scanning error distribution field ψ(x, y, z, t).

[0178] In a marine environment, when a welding robot uses laser scanning to search for weld seams, the dynamic inclined plane formed by the sea wave fluctuations will cause sunlight reflection to interfere with laser scanning, resulting in systematic deviations in the point cloud data. In standard water wave dynamics, sea waves can be regarded as a nonlinear wave system controlled by the Laplace equation and boundary conditions. However, due to factors such as turbulence, surface tension, and viscous damping in sea waves, the linear wave equation cannot accurately describe these complex phenomena. Therefore, it is necessary to introduce a nonlinear wave equation. Using the nonlinear wave equation can accurately simulate the dynamics of sea waves without relying on empirical filtering methods, thus having stronger generalization ability. By calculating the error distribution through the laser scanning error distribution field ψ(x, y, z, t), the point cloud data of the robot can be directly corrected adaptively to improve the weld seam positioning accuracy. Since the nonlinear terms are considered, the model can adapt to extreme situations such as strong winds and waves and wave breaking.

[0179] In this embodiment, κ is the influence coefficient of the sea wave inclination angle, and is calculated as: where n is the refractive index of seawater, ΔL is the optical path offset; δ is the time dynamic influence coefficient, and is calculated as: where λ L is the laser wavelength, and c is the propagation speed of sea waves.

[0180] S2. Based on the weld seam point cloud data generated by the robot laser scanning, and combined with the laser scanning error distribution field, abnormal points in the weld seam point cloud data are screened to form a point cloud data set to be corrected;

[0181] In this embodiment S2, based on the weld seam point cloud data generated by the robot laser scanning, and combined with the laser scanning error distribution field, abnormal points in the weld seam point cloud data are screened to form a point cloud data set to be corrected. The specific method steps are as follows:

[0182] S2.1. The weld seam point cloud data generated by the robot laser scanning is:

[0183]

[0184] where M is the total number of point data in the weld seam point cloud data; i is the point data index; A iis the horizontal coordinate of the point data i; B i is the coordinate of the point data i perpendicular to the propagation direction; C i is the height coordinate of the point data i; (A i , B i , C i ) is the three-dimensional coordinate of the point data i; Q(t) is the weld seam point cloud data collected at time t;

[0185] In this embodiment, the weld seam point cloud data generated by the robot laser scanning is a commonly used technical means in the prior art. The specific method for obtaining the weld seam point cloud data is as follows:

[0186] The robot program triggers the laser vision sensor to start searching for the position, and scans the weld seam according to the laser vision scanning trajectory established by the offline programming software to obtain the weld seam point cloud data;

[0187] S2.2. Use the laser scanning error distribution field to screen out the abnormal points in the weld seam point cloud data to form a point cloud data set to be corrected.

[0188] In step S2.2 of this embodiment, the laser scanning error distribution field is used to screen out the abnormal points in the weld seam point cloud data to form a point cloud data set to be corrected. The specific method steps are as follows:

[0189] S2.2.1. Calculate the theoretical error of each point data in the weld seam point cloud data based on the laser scanning error distribution field:

[0190]

[0191] where, ∈ i is the theoretical error of the point data i; λ1 is the influence of the sea wave height change on the error; λ2 is the influence of the time dynamic change on the error; is the inclination influence of the sea wave slope on the point data i; is the influence of the sea wave disturbance in the time dimension on the point data i;

[0192] S2.2.2. Set an error threshold τ, and screen out the point data with an error greater than the error threshold τ to form a point cloud data set to be corrected:

[0193] Q error (t) = {(A i , B i , C i ) | ∈ i > τ};

[0194] where, Q error (t) is the point cloud data set to be corrected.

[0195] S3. Define the point cloud correction target energy functional, and combine it with the wave slope reflection perturbation potential function to iteratively optimize the point cloud data set to be corrected, and obtain the corrected weld point cloud data;

[0196] The point cloud correction target energy functional is constructed based on the wave slope reflection perturbation potential function combined with the point cloud data set to be corrected, and is used to correct the error of the weld point cloud data caused by the reflection of sunlight by seawater ripples.

[0197] In this embodiment S3, define the point cloud correction target energy functional, and combine it with the wave slope reflection perturbation potential function to iteratively optimize the point cloud data set to be corrected, and obtain the corrected weld point cloud data. The specific method steps are as follows:

[0198] S3.1. Define the point cloud correction target energy functional;

[0199] S3.2. Use the gradient descent algorithm to calculate the gradient descent direction of the point cloud data to be corrected;

[0200] S3.3. Combine the wave slope reflection perturbation potential function to correct the error points, and calculate the offset correction amount of the error points;

[0201] S3.4. Based on the offset correction amount of the error points, obtain the final correction formula;

[0202] S3.5. Iteratively execute S3.1 - S3.4 to optimize all the point cloud data to be corrected, and finally obtain the corrected weld point cloud data.

[0203] In this embodiment, the purpose of the point cloud correction target energy functional is to construct an optimization framework so that the error point cloud data scanned by the robot can be automatically corrected and approach the true weld shape. Due to the wave characteristics of the sea, the weld point cloud data may have local offsets, distortions, or outliers. The point cloud correction target energy functional quantifies the error and uses gradient descent optimization to make the error points converge towards the true weld trajectory, thereby eliminating the laser scanning error;

[0204] In traditional point cloud correction methods, simple interpolation or filtering is usually used to remove error points without considering the dynamic interference characteristics of the waves. This method constructs the point cloud correction target energy functional based on the wave slope reflection perturbation potential function, which can be adaptively adjusted according to the sea waves and is more in line with the actual situation of the marine welding environment;

[0205] In the process of point cloud correction, it is not enough to simply make each point converge to its ideal position. The overall consistency of the entire point cloud data also needs to be considered. Therefore, a smoothing regularization term Ω(Q) is added to the point cloud correction target energy functional to constrain the change trend of the point cloud data.

[0206] In this embodiment S3.1, define the point cloud correction target energy functional. The specific method is as follows:

[0207]

[0208] Among them, is the point cloud correction target energy functional; Q is the point cloud dataset to be corrected, equivalent to Q error (t); (X i , Y i , Z i ) is the current coordinate of the point data to be corrected; (M i , N i , P i ) is the standard point in the ideal weld model; w(|ψ(X i , Y i , Z i , θ)|) suppresses the error term; Ω(Q) is the smoothing regularization term; α1 is the deviation weight coefficient between the point cloud data and the ideal weld model; α2 is the error field weight coefficient; α3 is the point cloud smoothing regularization weight coefficient; θ is the number of optimization iterations; K is the total number of error points in the point cloud dataset to be corrected;

[0209] In this embodiment S3.2, the gradient descent algorithm is used to calculate the gradient descent direction of the point cloud data to be corrected. The specific method is as follows:

[0210]

[0211] This formula is the calculation of gradient descent;

[0212] Among them, is the partial derivative in the gradient descent algorithm;

[0213] The correction direction of each error point is determined by its gradient. The update formula is as follows:

[0214]

[0215] Among them, λ is the number of iteration rounds; is the point cloud coordinate of the λ-th round of iteration; is the point cloud coordinate of the (λ + 1)-th round of iteration; β is the learning rate.

[0216] In this embodiment S3.3, the error points are corrected by combining the sea wave slope reflection perturbation potential function, and the offset correction amount of the error points is calculated. The specific method is as follows:

[0217]

[0218] Among them, Δ i is the offset correction amount of point data i; μ1 is the influence coefficient of sea water ripple perturbation in the horizontal direction; μ2 is the influence coefficient of sea wave perturbation in the time dimension; It is the influence of the wave slope reflection disturbance potential function in the horizontal direction of the point cloud; It is the influence of the wave slope reflection disturbance potential function in the time dimension;

[0219] In this embodiment S3.4, based on the offset correction amount of the error points, the final correction formula is obtained. The specific method is as follows:

[0220]

[0221] In this embodiment S3.5, S3.1 - S3.4 are iteratively executed to optimize all the point cloud data to be corrected, and finally the corrected weld point cloud data is obtained. The specific method is as follows:

[0222]

[0223] Among them, Q corrected (t) is the corrected weld point cloud data.

[0224] In this embodiment, the finally obtained corrected weld point cloud data includes the point cloud data corrected from the point cloud data set to be corrected plus the point cloud data that originally meets the standard.

[0225] S4. Perform automatic welding by the robot based on the corrected weld point cloud data.

[0226] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments. The above embodiments and the descriptions in the specification are only preferred examples of the present invention and are not used to limit the present invention. Without departing from the spirit and scope of the present invention, the present invention will have various changes and improvements, and these changes and improvements all fall within the scope of the present invention claimed.

Claims

1. A method for finding the intersection line welding position and adaptive deviation correction of a TKY pipe joint, characterized in that It includes the following steps: S1. Collect marine environment data using sensors, establish a wave slope reflection perturbation potential function based on the marine environment data, and calculate and generate a laser scanning error distribution field; S2. Based on the weld point cloud data generated by the robot's laser scanning, and in combination with the laser scanning error distribution field, screen out the abnormal points in the weld point cloud data to form a point cloud data set to be corrected; S3. Define a point cloud correction target energy functional, and in combination with the wave slope reflection perturbation potential function, iteratively optimize the point cloud data set to be corrected to obtain the corrected weld point cloud data; S4. Based on the corrected weld point cloud data, execute the robot for automatic welding.

2. The TKY pipe joint intersection line welding position searching and adaptive deviation correction method according to claim 1, wherein: The marine environment data includes wave height, wave tilt angle, wave period, wave propagation speed, solar azimuth angle, solar intensity, and seawater refractive index; The wave slope reflection perturbation potential function is constructed based on the nonlinear wave equation and considering the influence of the seawater wave forming a slope to reflect sunlight on the welding point, and is used to describe the dynamic interference caused by the reflection of seawater waves on sunlight when scanning the welding point.

3. The TKY pipe joint intersection line welding position finding and adaptive deviation correction method according to claim 2, characterized in that: In S1, based on the marine environment data, establish a wave slope reflection perturbation potential function and calculate the laser scanning error distribution field. The specific method steps are as follows: S1.

1. Based on the marine environment data, define and construct a wave slope reflection perturbation potential function; S1.

2. Based on the wave slope reflection perturbation potential function, obtain the laser scanning error distribution field by calculating the seawater wave tilt angle and the solar light reflection angle; Among them, the laser scanning error distribution field is used to quantitatively describe the laser scanning error caused by the reflection of sunlight by the seawater wave slope, and is used to identify and correct the error point data.

4. The TKY pipe joint intersection line welding position finding and adaptive deviation correction method according to claim 3, characterized in that: In S1.1, based on the marine environment data, define and construct a wave slope reflection perturbation potential function, specifically as follows: Define the wave slope reflection perturbation potential function as Φ(x, y, z, t): where t is time; x is the horizontal coordinate of the sea surface; y is the coordinate perpendicular to the propagation direction of the sea surface; z is the height coordinate of the sea surface; is the second-order partial derivative of the reflected perturbation potential function of the sea surface slope with respect to time; is the Laplace operator; c is the sea wave propagation speed; γ is the high-order nonlinear coupling coefficient; ζ is the nonlinear damping coefficient; is the magnitude of the potential energy gradient of the sea wave; is the partial derivative of the reflected perturbation potential function of the sea surface slope with respect to time; In S1.2, based on the wave slope reflection perturbation potential function, obtain the laser scanning error distribution field by calculating the seawater wave tilt angle and the solar light reflection angle, specifically as follows: where, ψ(x, y, z, t) is the laser scanning error distribution field; is the partial derivative of the reflected disturbance potential function of the sea wave slope with respect to height; is the partial derivative of the reflected disturbance potential function of the sea wave slope with respect to time; κ is the sea wave tilt angle influence coefficient; δ is the time dynamic influence coefficient.

5. The TKY pipe joint intersection line welding position finding and adaptive deviation correction method according to claim 4, wherein: In S2, based on the weld point cloud data generated by the robot's laser scanning, and in combination with the laser scanning error distribution field, screen out the abnormal points in the weld point cloud data to form a point cloud data set to be corrected. The specific method steps are as follows: S2.

1. The weld point cloud data generated by the robot's laser scanning is: Where M is the total number of point data in the weld point cloud data; i is the point data index; A i is the horizontal coordinate of point data i; B i is the coordinate of point data i perpendicular to the propagation direction; C i is the height coordinate of point data i; (A i ,B i ,C i ) is the three-dimensional coordinate of point data i; Q(t) is the weld point cloud data collected at time t; S2.

2. Use the laser scanning error distribution field to screen out the abnormal points in the weld point cloud data to form a point cloud data set to be corrected.

6. The method for finding the intersection line of the TKY pipe joint and self-adaptive deviation correction during welding according to claim 5, characterized in that: In S2.2, use the laser scanning error distribution field to screen out the abnormal points in the weld point cloud data to form a point cloud data set to be corrected. The specific method steps are as follows: S2.2.

1. Calculate the theoretical error of each point data in the weld point cloud data based on the laser scanning error distribution field; where, ∈ i is the theoretical error of point data i; λ1 is the influence of sea wave height change on the error; λ2 is the influence of time dynamic change on the error; is the inclination influence of the sea wave slope on point data i; is the influence of sea wave disturbance in the time dimension on point data i; S2.2.

2. Set an error threshold τ, screen out the point data with an error greater than the error threshold τ to form a point cloud data set to be corrected; Q error (t) = {(A i , B i , C i ) | ∈ i > τ}; Among them, Q error (t) is the point cloud data set to be corrected.

7. The method for finding the intersection line of a TKY pipe joint and self-adaptive deviation correction during welding according to claim 6, characterized in that: The point cloud correction target energy functional is constructed based on the wave slope reflection perturbation potential function in combination with the point cloud data set to be corrected, and is used to correct the error of the weld point cloud data caused by the reflection of sunlight by the seawater ripples.

8. The TKY pipe joint intersection line welding position finding and adaptive deviation correction method according to claim 7, characterized in that: In S3, a point cloud correction target energy functional is defined, and combined with the wave slope reflection perturbation potential function, the point cloud data set to be corrected is iteratively optimized to obtain the corrected weld point cloud data. The specific method steps are as follows: S3.1: Define the point cloud correction target energy functional; S3.2: Use the gradient descent algorithm to calculate the gradient descent direction of the point cloud data to be corrected; S3.3: Combine the wave slope reflection perturbation potential function to correct the error points and calculate the offset correction amount of the error points; S3.4: Based on the offset correction amount of the error points, obtain the final correction formula; S3.5: Iteratively execute S3.1 - S3.4 to optimize all the point cloud data to be corrected, and finally obtain the corrected weld point cloud data.

9. The method for finding the intersection line welding position and self-adaptive deviation correction of the TKY pipe joint according to claim 8, characterized in that: In S3.1, the point cloud correction target energy functional is defined. The specific method is as follows: Among them, is the point cloud correction target energy functional; Q is the point cloud data set to be corrected, equivalent to Q error (t); (X i , Y i , Z i ) is the current coordinate of the point data to be corrected; (M i , N i , P i ) is the standard point in the ideal weld model; w(|ψ(X i , Y i , Z i , θ)|) suppresses the error term; Ω(Q) is the smoothing regularization term; α1 is the deviation weight coefficient between the point cloud data and the ideal weld model; α2 is the error field weight coefficient; α3 is the point cloud smoothing regularization weight coefficient; θ is the number of optimization iterations; K is the total number of error points in the point cloud data set to be corrected; In S3.2, the gradient descent algorithm is used to calculate the gradient descent direction of the point cloud data to be corrected. The specific method is as follows: Among them, is the partial derivative in the gradient descent algorithm; Among them, λ is the number of iteration rounds; is the point cloud coordinates of the λ-th round of iteration; is the point cloud coordinates of the (λ + 1)-th round of iteration; β is the learning rate.

10. The TKY pipe joint intersection line welding position finding and adaptive deviation correction method according to claim 9, characterized in that: In S3.3, the wave slope reflection perturbation potential function is combined to correct the error points and calculate the offset correction amount of the error points. The specific method is as follows: where, Δ i is the offset correction amount of point data i; μ1 is the influence coefficient of seawater ripple disturbance in the horizontal direction; μ2 is the influence coefficient of sea wave disturbance in the time dimension; is the influence of the sea wave slope reflection disturbance potential function in the horizontal direction of the point cloud; is the influence of the sea wave slope reflection disturbance potential function in the time dimension; In S3.4, based on the offset correction amount of the error points, the final correction formula is obtained. The specific method is as follows: In S3.5, iteratively execute S3.1 - S3.4 to optimize all the point cloud data to be corrected, and finally obtain the corrected weld point cloud data. The specific method is as follows: Among them, Q corrected (t) is the corrected weld point cloud data.

Images