Weld groove positioning method based on simplified geometric feature data, weld tracking system, welding equipment and application in welding process

Abstract

The present invention provides a weld groove positioning method based on the simplification of geometric feature data, a weld tracking system, welding equipment, and applications in the welding process, and relates to the field of automated welding. The present invention is oriented to the geometric characteristics of the contour curve of the weld groove, and proposes an improved groove feature point extraction method, which effectively improves the traditional Douglas Peucker algorithm and effectively applies it to weld groove positioning. It has been verified that this method can improve the extraction stability of key feature points of the weld. In addition, the present invention uses fitting interpolation technology to further calculate the coordinates of the feature points, improve the anti-interference ability, and at the same time calculates the geometric information such as the weld width and height, and provides it to the weld tracking system to assist in adjusting the welding process parameters, and finally realizes a high-precision welding guidance process. The improved algorithm proposed in the invention has strong scalability and can be applied to the recognition and positioning of common types of groove features, such as V-grooves, lap grooves, corner grooves, etc.

Description

Technical Field

[0001] The present invention relates to the field of automated welding, and in particular to a weld groove positioning method based on simplified geometric feature data, a weld tracking system, welding equipment, and applications in a welding process. Background Art

[0002] Welding is one of the most important material forming and processing technologies in modern manufacturing. The development of welding manufacturing technology is crucial for my country to become a manufacturing powerhouse. However, welding is a job characterized by poor working conditions, high levels of smoke and dust, high heat radiation, high risk, and repetitiveness. Furthermore, as the welding industry faces increasing challenges with multiple batches, small production runs, and a shortage of personnel, automating the welding process has become a key technology driving the development of the manufacturing industry.

[0003] Currently, intelligent welding systems based on dedicated welding machines or welding robots, combined with laser seam tracking sensors, are experiencing continuous development and widespread application. Seam tracking sensors empower welding systems with keen environmental awareness and intelligent analysis and processing capabilities, enabling a more flexible and intelligent welding manufacturing process. This is primarily achieved by integrating the non-contact, high-speed, high-precision, and real-time detection advantages of laser vision sensors into the automated welding process, and developing corresponding algorithms for seam recognition and real-time tracking to address the challenges of online guided welding. The stability and positioning accuracy of weld groove identification are fundamental and crucial aspects of this suite of technologies. This is because seam tracking sensors provide direct data for subsequent real-time perception of the spatial position and posture of the actual weld, enabling precise and accurate operational actions, directly impacting tracking performance. Consequently, research on weld groove recognition and positioning technologies continues to flourish.

[0004] Research has shown that current vision-based weld inspection systems have shortcomings such as low real-time performance and insufficient universality. In particular, due to varying surface roughness of welding materials, varying groove shapes, and varying reflective properties, weld features are not distinct, resulting in insufficient positioning accuracy for automated welding systems and unstable welding results. Currently, commonly used vision-based weld tracking sensors are mainly divided into two categories: 1) Image-processing-based weld feature recognition. This method directly identifies and locates weld features in images, lacking depth information and exhibiting a strong dependence on light and weak generalization capabilities. 2) Laser-based weld tracking systems. This method utilizes active visual measurement technology, exhibiting strong anti-interference and adaptability. The sensor outputs rich information, including width and depth. However, the positioning accuracy of the weld groove is significantly affected by the contour data processing capability. Summary of the Invention

[0005] Purpose of the invention: To propose a weld groove positioning method based on simplified geometric feature data, and further propose a weld tracking system and welding equipment that can implement the above positioning method, so as to solve the above problems existing in the prior art.

[0006] A first aspect of the present invention provides a weld groove positioning method based on simplified geometric feature data, the method comprising the following steps:

[0007] S1. Using a line laser measurement sensor to collect an original contour curve of the weld groove, where the original contour curve is formed by connecting a number of discrete points;

[0008] S2. Select the starting point A and the end point B of the original contour curve, and connect the starting point A and the end point B to form a line segment AB;

[0009] S3. Calculate the distances between all discrete points in the original contour curve and the line segment AB, select the discrete point with the largest distance to the line segment AB as point C, and record the distance between point C and the line segment AB as distance d.

[0010] S4. Compare the lengths of line segment AC and line segment BC:

[0011] If the length difference between line segment AC and line segment BC is less than the predetermined interval, retain point A and point B in step S3 as feature points and output the distance d in step S3;

[0012] If the length difference between line segment AC and line segment BC is greater than or equal to the predetermined interval, it is determined that the current point C has a positioning error, and new points A', B', and C' are refitted, and the distance between point C' and line segment A'B' is output as the latest distance d;

[0013] S5. Set a threshold d according to the feature point extraction requirements. t , compare the distance d with the threshold d t Size:

[0014] If d<d t , then output points A and B as feature points, and the contour line processing is completed;

[0015] If d≥d t , then take point C as the boundary, divide the original contour curve into two parts, line segment AC and line segment CB, and repeat S1 to S4 for line segment AC and line segment CB respectively;

[0016] S6. After all line segments are processed, all feature points extracted from each segment are output in sequence as the final features of the weld groove contour.

[0017] In a further embodiment of the first aspect, the predetermined interval in step S4 is defined as follows: the length difference between line segment AC and line segment BC is more than one times.

[0018] In a further embodiment of the first aspect, if the difference in length between line segment AC and line segment BC exceeds one, and the length of line segment AC is less than the length of line segment CB:

[0019] Find the midpoint A' of line segment AC, and at the same time approximately find point B' in the middle of line segment CB, so that A'C ≈ CB'. Then re-execute step S2 with A' and B' as endpoints, replace C with the newly determined C', and record the distance d between point C' and line segment A'B'.

[0020] If the difference between the lengths of line segment AC and line segment BC is more than one, and the length of line segment AC is greater than the length of line segment CB:

[0021] Find the midpoint B' of line segment CB, and at the same time find the point A' approximately in the middle of line segment AC, so that CB'≈A'C. Then re-execute step S2 with A' and B' as endpoints, replace C with the newly determined C', and record the distance d between point C' and line segment A'B'.

[0022] A second aspect of the present invention provides a weld tracking system, the system comprising:

[0023] welding robots;

[0024] A line laser measurement sensor is installed at the end of the welding robot and is used to collect weld profile data in real time;

[0025] a contour data acquisition unit, configured to receive the weld contour data collected from the line laser measurement sensor, execute the weld groove positioning method disclosed in the first aspect and further embodiments thereof based on the weld contour data to extract key groove geometry information, and send the key groove geometry information to a control unit via TCP;

[0026] The control unit is used to control the welding robot to guide and track the weld in real time according to the key geometric information of the groove.

[0027] A third aspect of the present invention provides automated welding equipment, comprising the weld seam tracking system disclosed in the second aspect, a welding gun, and a welding machine. A welding robot is controlled by the control unit to achieve real-time guidance and tracking of the weld seam; the welding gun is mounted on the end effector of the welding robot; and the welding gun is connected to the welding machine.

[0028] The fourth aspect of the present invention proposes a computer-readable storage medium, which stores executable software. When the executable software runs on an electronic device, the electronic device executes the weld groove positioning method based on simplified geometric feature data as disclosed in the first aspect and its further embodiments.

[0029] The fifth aspect of the present invention proposes a computer program product. When the instructions in the computer program product are executed by the processor of an electronic device, the electronic device executes the weld groove positioning method based on geometric feature data simplification as disclosed in the first aspect and its further embodiments.

[0030] The sixth aspect of the present invention proposes the application of the weld groove positioning method based on geometric feature data simplification as disclosed in the first aspect and its further embodiments in the welding process, which is specifically used to detect weld feature points of grooves of various shapes and / or automatically generate weld positioning points; the groove can be a V-shaped groove, a lap groove, or a corner groove.

[0031] Beneficial Effects: The proposed method effectively improves the traditional Douglas-Peucker algorithm and applies it to weld groove location. This method has been proven to improve the stability of extracting key weld feature points. Furthermore, the proposed method uses fitting interpolation techniques to further calculate feature point coordinates, improving anti-interference capabilities. It also calculates geometric information such as weld width and height, providing it to the weld tracking system to assist in adjusting welding process parameters, ultimately achieving highly accurate welding guidance. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 Schematic diagram of the structure of the weld groove positioning platform in an embodiment of the present invention.

[0033] Figure 2 Schematic diagram of the points in step 1 in an embodiment of the present invention.

[0034] Figure 3 Schematic diagram of the points in step 2 in an embodiment of the present invention.

[0035] Figure 4 Schematic diagram of the points in step 3 in the embodiment of the present invention.

[0036] Figure 5 Schematic diagram of the points in step 4 in the embodiment of the present invention.

[0037] Figure 6 Schematic diagram of the points in step 5 in the embodiment of the present invention.

[0038] Figure 7 This is a diagram showing the effect of lap groove recognition in an embodiment of the present invention.

[0039] Figure 8 This is a diagram showing the effect of butt joint recognition in an embodiment of the present invention.

[0040] Figure 9 This is a diagram showing the V-groove recognition effect in an embodiment of the present invention.

[0041] Figure 10 This is a diagram showing the U-shaped groove recognition effect in an embodiment of the present invention.

[0042] The reference numerals in the figure are as follows: welding robot 1 , line laser measurement sensor 2 , welding machine 3 , control unit 4 . DETAILED DESCRIPTION

[0043] In the following description, numerous specific details are provided to provide a more thorough understanding of the present invention. However, it will be apparent to those skilled in the art that the present invention may be practiced without one or more of these details. In other instances, certain technical features well known in the art have not been described to avoid confusion with the present invention.

[0044] According to research findings, the current vision-based weld detection system has shortcomings such as low real-time performance and insufficient universality. In particular, due to the different surface roughness of welding materials, different groove shapes, different reflective properties, and unclear weld characteristics, the positioning accuracy of the automated welding system is insufficient, resulting in unstable welding effects.

[0045] The present invention aims to address the aforementioned issues in groove location technology based on laser weld tracking sensors. By focusing on the geometric characteristics of weld groove contour curves and building on the essential principles of the Douglas-Peucker algorithm, an improved groove feature point extraction method is proposed. This method, which pertains to the field of automated welding, specifically addresses the detection of weld feature points and the automatic generation of weld location points for various groove shapes during welding. The improved algorithm proposed in the invention is highly scalable and can be applied to the identification and location of common groove types, such as V-grooves, lap joints, and fillet joints.

[0046] The Douglas–Peucker algorithm approximates a given curve as a series of points. This commonly used curve simplification algorithm can reduce the number of points in a given curve point set and extract the curve's characteristic points. The number of output points can be actively controlled by setting a threshold to meet different simplification requirements. While the Douglas–Peucker algorithm effectively reduces the number of data points, thereby reducing both data volume and computational complexity, research has found that the retained data points are those with more distinct geometric features. Based on this finding, by appropriately selecting the threshold (or dynamically calculating it based on the number of data points to be extracted), the N data points with the strongest characteristics can be extracted. However, the traditional Douglas–Peucker algorithm has certain drawbacks. When there is significant asymmetry between the data on the left and right sides of the true characteristic point, the extracted characteristic point may contain some error. This situation often occurs in weld groove profile data, such as lap grooves and V-grooves.

[0047] In order to solve the problem of identifying and locating the key feature points of various common grooves, the present invention adopts the following technical solutions:

[0048] Build the weld groove positioning platform:

[0049] The robot laser welding groove positioning platform is mainly composed of a six-degree-of-freedom welding robot 1, a line laser measurement sensor 2, a welding gun and a welding machine 3. Figure 1 As shown in the figure, the line laser measurement sensor 2 collects weld seam contour data in real time, extracts key groove geometry information based on the groove characteristics and positioning algorithm designed in R&D, and transmits this information via TCP to the robot control unit 4, enabling real-time guidance and tracking of the welding robot 1. This system is a typical closed-loop control system, with the line laser measurement sensor 2 and the weld seam tracking algorithm forming the feedback link, making it crucial for achieving a successful automated welding process.

[0050] Groove feature extraction method based on improved Douglas Puck:

[0051] The Douglas–Peucker algorithm approximates a given curve as a series of points. This commonly used curve simplification algorithm can reduce the number of points in a given curve point set and extract the curve's characteristic points. The number of output points can be actively controlled by setting a threshold to meet different simplification requirements. While the Douglas–Peucker algorithm effectively reduces the number of data points, thereby reducing data volume and computational complexity, research has found that the data points that remain are those with more distinct geometric features. Based on this finding, by appropriately selecting a threshold (or dynamically calculating it based on the number of data points to be extracted), it is possible to extract the N data points with the strongest characteristics.

[0052] However, the traditional Douglas-Peucker algorithm has certain drawbacks. When the data on the left and right sides of the true feature point are significantly asymmetric, the extracted feature point may contain certain errors. This situation often occurs in weld groove profile data, such as lap grooves and V-grooves.

[0053] The present invention is oriented to the geometric characteristics of the contour curve of the weld groove and based on the essential idea of ​​the Douglas-Peucker algorithm to propose an improved groove feature point extraction method.

[0054] The specific steps of the proposed method are:

[0055] Step 1: If the original contour curve collected by the sensor is as follows Figure 2 As shown, select a curve that needs to be simplified, set the starting points of the curve as points A and B, and connect points A and B to form a straight line.

[0056] Step 2: Calculate the distances between all points in the contour curve and the straight line AB. After comparison, select the point with the largest distance to the straight line as point C. At the same time, record the distance between point C and AB as d. Figure 3 shown.

[0057] Step 3: If the distance between AC and BC is more than double, then the position error of point C determined in step 2 may exist and further optimization and adjustment is required, such as Figure 4 As shown. If the AC distance is less than CB, find the midpoint A' of the AC curve and approximately find a point B' in the middle of CB, so that A'C ≈ CB'. Then, repeat step 2 with A' and B' as endpoints, replacing C with the newly determined C', and record the latest distance d between point C' and A'B'.

[0058] Step 4: Set an appropriate threshold based on the feature point extraction requirements. Compare the distance d with the given threshold. If it is less than the threshold, output points A and B as feature points, and the contour line processing is completed.

[0059] Step 5: If the distance d is greater than the threshold, the contour line is divided into two parts, AC and CB, with point C as the boundary, and steps 1 to 4 are repeated for the two curve segments, such as Figure 5 shown.

[0060] Step 6: After all curve segments are processed, all feature points extracted from each segment are output in sequence as the final features of the weld groove contour, such as Figure 6 Indicated by the red dot.

[0061] Tracking results and analysis:

[0062] To verify the feasibility and accuracy of the proposed method, a weld seam tracking test platform was constructed. The experimental platform mainly consists of a Yaskawa DX200 control cabinet, a MOTOMAN-MA1440 general-purpose robot (with network communication and Motoplus functional modules enabled), a linear laser weld seam tracking sensor, various welding test pieces, and weld seam tracking control software.

[0063] In order to verify the feasibility and accuracy of this method, a laser vision sensor was used to collect the contour data of common weld grooves such as lap joints, butt joints, V-shaped welds, and U-shaped welds, and the algorithm was verified. The located groove feature points were marked prominently in the original data center. The experimental results are shown as follows: Figures 7-10 As shown. Among them, Figure 7 is the overlap groove profile data, Figure 8 is the butt joint groove profile data, Figure 9 It is the V-groove profile curve. Figure 10 The figure shows the U-groove profile data curve. It can be seen that the method of the present invention accurately extracts each groove feature point, and the positioning accuracy is precise. Of course, the positioning accuracy of the groove feature is also greatly affected by factors such as the acquisition accuracy of the line laser measurement sensor and the groove processing accuracy. Therefore, in actual welding applications, multiple factors should be comprehensively considered to achieve the optimal result.

[0064] The experimental results show that the method of the present invention can well solve the problem of weld groove feature extraction. The proposed method has very high scalability. After the groove features are extracted, the groove tracking position coordinates can be further calculated according to user needs, thereby achieving the purpose of guiding automated welding.

[0065] In the actual application of the weld groove location method disclosed in the above embodiment, its operating logic can be written into a program product, which is written into a storage medium and runs on an electronic device. More specific examples of the computer-readable storage medium mentioned in this embodiment may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. The computer-readable storage medium may include a data signal propagated in baseband or as part of a carrier wave, which carries readable program code. Such a propagated data signal may take a variety of forms, including but not limited to an electromagnetic signal, an optical signal, or any suitable combination of the above. The readable signal medium may also be any readable medium other than a readable storage medium, which can send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, device, or component.

[0066] Based on the contents disclosed in the above embodiments, the present invention has at least the following advantages:

[0067] Solve the general identification and location problems of many common weld groove features.

[0068] Solve the problem of feature point positioning error when there is local asymmetry in the groove and the shape processing is non-standard.

[0069] The method is simple in principle and easy to implement, and can be widely used in industrial sites to meet the needs of high-speed and high-precision detection.

[0070] It is proposed to apply the Douglas Peucker algorithm for data reduction to the location of groove contour features to achieve rapid extraction of groove geometric features.

[0071] An improved method for feature point positioning using secondary optimization of short-distance symmetrical data is proposed to solve the positioning error problem caused by the significant asymmetry of the left and right data of the feature points.

[0072] It should be noted that the real-time groove positioning algorithm proposed in the above embodiment is not only applicable to the case where the welding robot described in the article is used as an actuator, but is also applicable to the case where an actuator with multiple degrees of freedom is used, such as a welding machine. Therefore, the tracking technology of the present invention has scalability in application objects. Any use of the tracking technology of the present invention, even if the actuator changes, should be considered within the scope of protection of the present invention.

Claims

1. A weld groove positioning method based on simplified geometric feature data, characterized in that: The steps include: S1. Using a line laser measurement sensor to collect an original contour curve of the weld groove, where the original contour curve is formed by connecting a number of discrete points; S2. Select the starting point A and the end point B of the original contour curve, and connect the starting point A and the end point B to form a straight line segment AB; S3. Calculate the distances between all discrete points in the original contour curve and the line segment AB, select the discrete point with the largest distance to the line segment AB as point C, and record the distance between point C and the line segment AB as distance d. S4. Compare the lengths of line segments AC and BC: If the length difference between line segment AC and line segment BC does not exceed one, retain point A and point B in step S3 as feature points and output the distance d in step S3; If the length difference between line segment AC and line segment BC exceeds one time, it is determined that the current point C has a positioning error, and new points A', B', and C' are refitted, and the distance between point C' and line segment A'B' is output as the latest distance d; S5, according to the feature point extraction requirements, set a threshold d t , compare the distance d with the threshold d t Size: If d<d t , then output points A and B as feature points, and the contour line processing is completed; If d≥d t , then the original contour curve is divided into two parts, curve segment AC and curve segment CB, with point C as the boundary, and the curve segment AC and curve segment CB are processed repeatedly from S1 to S4 respectively; S6. After all line segments are processed, all feature points extracted from each segment are output in sequence as the final features of the weld groove contour.

2. The weld groove positioning method based on geometric feature data simplification according to claim 1 is characterized in that: If the difference in length between line segment AC and line segment BC is more than one, and the length of line segment AC is less than the length of line segment BC: Find the midpoint A' of curve segment AC, and at the same time approximately find point B' in the middle of curve segment CB, so that line segment A'C ≈ line segment CB'. Then re-execute step S2 with A' and B' as endpoints, replace C with the newly determined C', and record the distance d between point C' and line segment A'B'.

3. The weld groove positioning method based on geometric feature data simplification according to claim 1 is characterized in that: If the difference in length between line segment AC and line segment BC is more than one, and line segment AC is longer than line segment BC: Find the midpoint B' of the curve segment CB, and at the same time approximately find the point A' in the middle of the curve segment AC, so that the line segment CB' ≈ the line segment A'C. Then re-execute step S2 with A' and B' as endpoints, replace C with the newly determined C', and record the distance d between point C' and line segment A'B'.

4. A weld tracking system, characterized in that: include: welding robots; A line laser measurement sensor is installed at the end of the welding robot and is used to collect weld profile data in real time; a contour data acquisition unit, configured to receive the weld contour data collected from the line laser measurement sensor, execute the weld groove positioning method according to any one of claims 1 to 3 based on the weld contour data to extract key groove geometry information, and send the key groove geometry information to a control unit via TCP; The control unit is used to control the welding robot to guide and track the weld in real time according to the key geometric information of the groove.

5. An automated welding equipment, characterized in that: The weld seam tracking system according to claim 4, further comprising: A welding gun is mounted on the end effector of the welding robot; the welding gun is connected to the welding machine.

6. Application of the weld groove positioning method based on geometric feature data simplification according to any one of claims 1 to 3 in a welding process, characterized in that: Specifically used to detect weld feature points of various groove shapes and / or automatically generate weld positioning points; The groove can be a V-shaped groove, a lap groove, or a corner groove.

Images