A method and system for accurate positioning of a weld joint in a welding process

Abstract

The application discloses a kind of welding process weld joint accurate positioning method and system, belong to the technical field of automated welding, method includes: by 3D profile scanning sensor scanning the region where weld is located, continuously obtains multiple frames of weld profile data;To each frame weld profile data, trend distribution calculation and feature point calculation are executed in turn, and profile feature point is extracted;Extract the center point of groove in each frame weld profile data, and arrange according to scanning order, generate weld center sequence point set;Trend distribution calculation and feature point calculation are executed in turn to weld center sequence point set, and the feature point and coordinates of weld joint are extracted;Weld joint feature point is mapped to corresponding 3D weld profile data, obtain the three-dimensional space coordinates of weld joint, according to the three-dimensional space coordinates of weld joint, the coordinate of arc starting point executed by welding gun is calculated.The application realizes the three-dimensional accurate positioning of weld joint, guarantees that welding gun arcs or is extinguished in optimal space position, improves welding quality and automated welding efficiency.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of automatic welding, in particular to a welding process weld joint accurate positioning method and system. BACKGROUND

[0002] In the field of automatic welding, accurate identification and positioning of the weld joint is the core key to ensure welding quality. The existing positioning of the weld joint mainly relies on the following two ways:

[0003] The first is manual comparison positioning. The welder visually observes and moves the welding torch to the weld joint for the arc starting process. This method relies heavily on the experience and subjective judgment of the operator, and there are differences between different operators and different states of the same operator, resulting in low positioning accuracy and poor repeatability. In addition, it takes a long time from the beginning of positioning to the end of positioning, and the arc starting position is prone to errors due to human subjective judgment, directly affecting the stability and controllability of the welding quality.

[0004] The second is automatic positioning based on sensors. Commonly used is to collect data such as image data and electrical signal data at the weld joint by sensors, and then use traditional image processing techniques (such as edge detection and texture analysis) or deep learning techniques for feature recognition to locate the weld joint position; the image data collected is usually a 2D planar image, which is analyzed for texture features by traditional image processing techniques, or the pixel position of the weld joint is directly distinguished by deep learning, but the depth information of the weld joint cannot be obtained, resulting in the inability of the execution mechanism to accurately control the height of the welding torch and causing uncontrollable welding quality.

[0005] Therefore, the current field of automatic welding still lacks a method that can efficiently and accurately locate the position of the weld joint from three-dimensional space to overcome the inherent defects of human subjectivity and lack of 2D visual depth information. SUMMARY

[0006] The purpose of the present application is to overcome the problems in the prior art and provide a welding process weld joint accurate positioning method and system that can accurately identify the position of the weld joint and make the welding torch arc starting process at the optimal spatial position to ensure welding quality and improve welding efficiency.

[0007] The purpose of the present application is achieved by the following technical solutions:

[0008] In a first aspect, a welding process weld joint accurate positioning method is provided, comprising the following steps:

[0009] S1. Scanning the area where the weld is located by a 3D profile scanning sensor to continuously obtain multiple frames of weld profile data;

[0010] S2. For each frame of the acquired weld contour data, the trend distribution calculation and the feature point calculation are sequentially performed until the scanning is completed; wherein the trend distribution calculation comprises:

[0011] The single frame of weld contour data is divided into a plurality of continuous regions, the slope of a straight line connecting the first and last points in each continuous region is calculated, and by comparing the slope of the straight line corresponding to each continuous region with the size of the preset threshold, the rising trend point set and the falling trend point set in the frame of weld contour data are identified;

[0012] The feature point calculation comprises:

[0013] The distance between all points in the rising trend point set and the straight line formed by the first and last points of the point set is calculated, the distance between all points in the falling trend point set and the straight line formed by the first and last points of the point set is calculated, and the point with the maximum or minimum distance is found as the contour feature point;

[0014] S3. The contour feature points corresponding to all frames of weld contour data are counted, the bevel center point in each frame of weld contour data is extracted, and the bevel center point is arranged in the scanning order to generate a sequence point set of weld center points;

[0015] S4. The trend distribution calculation and the feature point calculation are sequentially performed on the sequence point set of weld center points to extract the weld joint feature points and coordinates;

[0016] S5. The weld joint feature points are mapped to the corresponding 3D weld contour data to obtain the three-dimensional space coordinates of the weld joint, the coordinates of the arc starting point executed by the welding gun are calculated according to the three-dimensional space coordinates of the weld joint, and the arc starting process is output to the welding gun execution mechanism.

[0017] In some embodiments, the scanning of the area where the weld is located by the 3D contour scanning sensor comprises:

[0018] The welding equipment carrying the 3D contour scanning sensor is controlled to move at a uniform speed in a direction parallel to the weld, and the direction of the laser line emitted by the 3D contour scanning sensor is perpendicular to the weld.

[0019] In some embodiments, the trend distribution calculation specifically comprises:

[0020] The single frame of weld contour data is divided into K continuous regions with a length of L, and if the length of the last region is less than L, the remaining length is taken as the length value;

[0021] All the slopes of the straight lines corresponding to the continuous regions are traversed, and if it is detected that the absolute value of the slope of the straight line of a continuous region is greater than the threshold value, the index of the corresponding continuous region is recorded to the rising trend point set or the falling trend point set.

[0022] In some embodiments, the geometric relationship between the 3D profile scanning sensor and the welding torch is fixed when calculating the coordinates of the arc starting point performed by the welding torch.

[0023] In a second aspect, a precise positioning system for a weld joint in a welding process is provided, comprising:

[0024] a 3D profile scanning sensor for scanning a region where a weld is located, continuously obtaining a plurality of frames of weld profile data;

[0025] a weld joint positioning module in communication with the 3D profile scanning sensor, having a computer program stored therein, which, when executed by a processor, implements the following processes:

[0026] For each frame of weld profile data obtained, trend distribution calculation and feature point calculation are sequentially performed until the scanning is completed; wherein the trend distribution calculation comprises:

[0027] dividing a single frame of weld profile data into a plurality of continuous regions, calculating the slope of a straight line connecting the first and last points in each continuous region, and identifying the rising trend point set and the falling trend point set in the frame of weld profile data by comparing the slope of the straight line corresponding to each continuous region with a predetermined threshold value;

[0028] The feature point calculation comprises:

[0029] calculating the distance between all points in the rising trend point set and the straight line formed by the first and last points of the point set, calculating the distance between all points in the falling trend point set and the straight line formed by the first and last points of the point set, and finding the point with the maximum or minimum distance as the profile feature point;

[0030] counting the profile feature points corresponding to all frames of weld profile data, extracting the bevel center point in each frame of weld profile data, and arranging them in the scanning order to generate a sequence of weld center point sets;

[0031] performing the trend distribution calculation and feature point calculation on the sequence of weld center point sets in turn to extract the weld joint feature points and coordinates;

[0032] mapping the weld joint feature points to the corresponding 3D weld profile data to obtain the three-dimensional spatial coordinates of the weld joint, and calculating the coordinates of the arc starting point performed by the welding torch according to the three-dimensional spatial coordinates of the weld joint;

[0033] a welding torch execution mechanism in communication with the weld joint positioning module for receiving the coordinates of the arc starting point and driving the welding torch to move to the corresponding position to perform the arc starting process.

[0034] In some embodiments, the 3D profile scanning sensor is fixedly connected with the welding torch through a rigid mechanical connecting member, and a geometric relationship of the mechanical connecting member is pre-calibrated and input into the welding joint positioning module.

[0035] It should be further explained that the technical features corresponding to the above-mentioned various option embodiments can be combined or replaced with each other to form new technical solutions without conflict.

[0036] Compared with the prior art, the present application has the following advantages:

[0037] 1. The present application discards the limitation of traditional 2D image recognition lacking depth information, and directly uses a 3D profile scanning sensor to obtain complete three-dimensional profile data of a weld. Through trend distribution calculation and feature point calculation, three-dimensional spatial coordinates of a joint in the length direction (scanning direction), width direction and height direction of the weld are simultaneously analyzed, thereby guiding the welding torch to start or extinguish arc at an optimal three-dimensional spatial position, and fundamentally solving the welding quality problem caused by inaccurate height control of the welding torch.

[0038] 2. The entire positioning process of the present application is automatically completed by a sensor and an algorithm, and subjectivity, experience dependence and fatigue error of manual visual positioning are completely avoided. Through judgment by a set algorithm threshold and logic, the result has high repeatability and consistency, and significantly improves the stability and process controllability of the welding quality.

[0039] 3. The present application creatively proposes a double-layer analysis architecture combining trend distribution calculation and feature point calculation. Firstly, a trend interval and key feature points (edge points, center points) representing a groove structure are quickly screened out at a single frame profile level, and then a center sequence reflecting the overall trend of the weld is integrated; and the macro sequence is analyzed again to sensitively capture feature mutations representing the joint position. This analysis idea from local to global and from geometric features to structural features is designed for identifying the specific target of the weld joint, and has clear algorithm logic and high calculation efficiency, which can meet the real-time or quasi-real-time processing demand of welding.

[0040] 4. The 3D profile scanning sensor of the present application is fixedly connected with the welding torch through a rigid mechanical connecting member, and a geometric relationship of the mechanical connecting member is pre-calibrated and input into the welding joint positioning module. The requirement for hardware is clear (3D profile scanning sensor + mechanical connection), and complex multi-sensor fusion or special mechanical structure is not required. Through accurate mechanical calibration, a stable conversion relationship from the sensor coordinate system to the welding torch execution coordinate system can be established, so that the present method can be conveniently integrated into existing automatic welding equipment, has low modification cost and strong practicality.

[0041] 5.The accurate automatic joint positioning reduces manual intervention and teaching time before welding, and good arc striking and extinguishing quality reduces subsequent finishing process, thereby comprehensively improving the efficiency and benefit of welding production. BRIEF DESCRIPTION OF DRAWINGS

[0042] Figure 1 A flow chart of the precise joint positioning method of a welding process of the application;

[0043] Figure 2 A structural schematic diagram of the 3D profile scanning sensor for data acquisition of the application;

[0044] Figure 3 A schematic diagram of the welding profile data trend distribution calculation of the application;

[0045] Figure 4 A feature calculation schematic diagram of the application;

[0046] Figure 5 A schematic diagram of the application for obtaining the coordinate range of the welding joint;

[0047] Figure 6 A schematic diagram of the laser line profile extraction result of the application;

[0048] Figure 7 A 3D schematic diagram of the laser line scanning welding profile of the application.

[0049] In the figure: 1-3D profile scanning sensor; 2-welding torch; 3-protective gas nozzle of the welding torch; 4-welding seam; 5-welding joint; 6-welding material; 7-mechanical connecting piece. DETAILED DESCRIPTION

[0050] The technical solutions of the application will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the application, rather than all the embodiments. The components of the embodiments of the application described and shown in the accompanying drawings can be arranged and designed in various different configurations. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor should fall within the scope of protection of the application.

[0051] It should be noted that the defects in the above prior art solutions are the results obtained by the inventors after practice and careful study. Therefore, the discovery process of the above problems and the solutions proposed by the embodiments of the application to the above problems should be the contributions made by the inventors to the application in the process of invention and creation, and should not be understood as technical content known to those skilled in the art.

[0052] In view of the technical problems pointed out in the background art, the embodiments provided by the application are as follows:

[0053] In an exemplary embodiment, a welding process weld joint accurate positioning method is provided, as shown in Figures 1-2 The method comprises the following steps:

[0054] S1. Scan the area where the weld 4 is located by the 3D profile scanning sensor 1, and continuously obtain multiple frames of weld profile data;

[0055] S2. For each frame of obtained weld profile data, sequentially perform trend distribution calculation and feature point calculation until the scanning is completed; wherein the trend distribution calculation comprises:

[0056] Divide the single frame of weld profile data into multiple continuous regions, calculate the slope of the straight line connecting the first and last points in each continuous region, and identify the rising trend point set and the falling trend point set in the frame of weld profile data by comparing the slope of each continuous region with the preset threshold value.

[0057] The feature point calculation comprises:

[0058] Calculate the distance between all points in the rising trend point set and the straight line formed by the first and last points of the point set, calculate the distance between all points in the falling trend point set and the straight line formed by the first and last points of the point set, and find the point with the maximum or minimum distance as the profile feature point.

[0059] S3. Statistically analyze the profile feature points corresponding to all frames of weld profile data, extract the bevel center point in each frame of weld profile data, and arrange them in the scanning order to generate a weld center sequence point set.

[0060] S4. Perform the trend distribution calculation and feature point calculation on the weld center sequence point set in turn, extract the weld joint feature points and coordinates.

[0061] S5. Map the weld joint feature points to the corresponding 3D weld profile data to obtain the three-dimensional space coordinates of the weld joint 5, calculate the arc starting point coordinates executed by the welding gun according to the three-dimensional space coordinates of the weld joint 5, and output to the welding gun execution mechanism to execute the arc starting process.

[0062] Specifically, as shown in Figure 2 The weld 4 needs to be scanned on the welding material 6 using the 3D profile scanning sensor 1 on the welding equipment to obtain the corresponding data, and the weld joint 5 is positioned by the weld joint accurate positioning method to guide the welding gun 2 on the welding equipment to perform the point arc welding to ensure the quality of the welded joint.

[0063] Exemplarily, in step S1, the scanning of the area where the weld is located by the 3D profile scanning sensor comprises:

[0064] The welding equipment carrying the 3D profile scanning sensor 1 is controlled to move at a constant speed in a direction parallel to the weld 4, to uniformly scan the area where the weld 4 is located at a constant speed, and to make the direction of the laser line emitted by the 3D profile scanning sensor 1 perpendicular to the weld 4.

[0065] During the scanning process, each weld profile data is processed in real time, including profile data trend distribution calculation and profile data feature point calculation; after the scanning is completed, the feature points of all weld profile data are counted to form a weld center sequence point set, and then trend distribution calculation and feature point calculation are performed on the weld center sequence point set, and finally the weld joint features and coordinates are obtained; the coordinates are mapped to the weld 3D profile data to obtain the coordinates of the arc starting point executed by the welding gun; the arc starting process is accurately executed to ensure the welding quality. The 3D profile scanning sensor 1 on the welding equipment is fixedly connected to the welding gun 2 through a rigid mechanical connecting piece 7, and the geometric relationship of the mechanical connecting piece 7 is pre-calibrated to convert the coordinates of the weld joint 5 positioned by the weld joint positioning method into the execution coordinates of the welding gun 2.

[0066] In step S2, the trend distribution calculation specifically includes:

[0067] A frame of weld profile data is obtained, as shown in Figure 6 , which is represented as a data sequence [[x0, y0], [x1, y1],..., [xn, yn]], where [xi, yi] represents a two-dimensional coordinate data point on the profile, xi represents the transverse position of a data point on the laser line, yi represents the height value corresponding to the point, i = [0, n], and the data is divided into K continuous regions, where the length of each continuous region is L, and if the length of the last region is less than L, the remaining length is taken as the length; the slope of the straight line formed by the first and last points in each continuous region is calculated, the trend range of the data is divided by judging the change relationship between the slope of the straight line of each continuous region and the threshold T (a positive value), and finally the trend distribution of the data is obtained by sorting. Specifically, all the slopes of the continuous regions are traversed, and if it is detected that the absolute value of the slope of a continuous region is greater than the threshold, the index of the corresponding continuous region is recorded to the rising trend point set or the falling trend point set, wherein if the slope of the straight line is <-T, the index of the corresponding continuous region is recorded to the falling trend point set, and if the slope of the straight line is > T, the index of the corresponding continuous region is recorded to the rising trend point set.

[0068] It should be noted that directly differentiating (calculating the slope) each point in the weld profile data will be extremely sensitive to noise. By calculating the slope Ki of the straight line connecting the first and last points in a region (e.g. containing 10-20 points), the overall average trend of the data in the region is obtained, effectively filtering out small fluctuations in the region.

[0069] It should be noted that the rising trend corresponds to the process of laser line entering the groove from the base material plane (the height value decreases because the groove is lower than the base material plane, but in the coordinate system with height as the y-axis, the data points are convex upward, that is, the y value increases first and then decreases, and the left rising edge is the rising trend). The falling trend corresponds to the process of laser line returning to the base material plane from the bottom of the groove (the height value increases, which is represented in the coordinate system as the right falling edge of the concave downward data points, that is, the falling trend). Accurate extraction of these trend intervals is a prerequisite for subsequent searching for groove edge points and center points in the correct interval.

[0070] The specific algorithm flow of the method is as follows:

[0071] S1: The 3D profile scanning sensor 1 acquires a frame of weld profile data, and enters S2;

[0072] S2: Use the data trend calculation module to calculate the trend distribution of the frame of weld profile data, and enter S3;

[0073] S3: Use the feature calculation module to calculate the main feature points in the profile, and enter S4;

[0074] S4: Determine whether the scanning is finished, if yes, enter S5, otherwise return to S1;

[0075] S5: Count all profile feature points to obtain a weld center sequence point set, and enter S6;

[0076] S6: Use the data trend calculation module to calculate the trend distribution of the weld center sequence point set, and enter S7;

[0077] S7: Use the feature calculation module again to calculate the main feature points of the weld center sequence point set, and enter S8;

[0078] S8: Map the feature points to the profile 3D data, and enter S9;

[0079] S9: Calculate the weld joint 5 range coordinates and output to the welding gun 2 for execution, and end the flow.

[0080] The algorithm flow of the data trend calculation module is as follows:

[0081] S11: Acquire a data sequence [[x0, y0], [x1, y1],..., [xn, yn]], and enter S12;

[0082] S12: Divide the data sequence into K continuous regions, and the length of each region is L, and enter S13;

[0083] S13: Calculate the slope Ki of the straight line connecting the first and last points in each continuous region, i=[1, K], and enter S14;

[0084] S14: judge the size relation between straight line slope Ki and threshold value T, wherein i=1, m=1, n=1, BM=0, BN=0, enter S15A and S15B; wherein, BM represents the state flag of the beginning of the downward trend (0 / 1 is taken), BN represents the state flag of the beginning of the upward trend (0 / 1 is taken); BM / BN takes 0 when it means that the current is not in a downward / upward trend process being recorded, or the last downward / upward trend has ended, and BM / BN takes 1 when it means that a new downward / upward trend starting point has been detected and recorded;

[0085] S15A: judge whether Ki is less than threshold value -T, if yes, enter S16A, otherwise enter S16A1;

[0086] S16A: judge whether flag signal BM is equal to 1, if yes, enter S17A, otherwise enter S18A;

[0087] S16A1: flag signal BM is set to 1, enter S20A1;

[0088] S17A: subscript m++ is added, flag signal BM is set to 0, enter S18A;

[0089] S18A: record i value in point set Dm, enter S19A;

[0090] S19A: judge whether i is greater than K, if yes, enter S20A, otherwise enter S20A1;

[0091] S20A: arrange point set Dm (downward trend point set), enter S21;

[0092] S20A1: subscript i++ is added, return to S15A;

[0093] S15B: judge whether Ki is greater than threshold value T, if yes, enter S16B, otherwise enter S16B1;

[0094] S16B: judge whether flag signal BN is equal to 1, if yes, enter S17B, otherwise enter S18B;

[0095] S16B1: flag signal BN is set to 1, enter S20B1;

[0096] S17B: subscript n++ is added, flag signal BN is set to 0, enter S18B;

[0097] S18B: record i value in point set Un, enter S19B;

[0098] S19B: judge whether i is greater than K, if yes, enter S20B, otherwise enter S20B1;

[0099] S20B: Arrange the point set Un (the point set of the rising trend), and enter S21;

[0100] S20B1: subscript i++, return to S15B;

[0101] S21: Obtain the trend distribution point set Dm and Un of the data sequence, and enter S22;

[0102] S22: End the flow.

[0103] As shown in the figure, the figure is the trend result calculated by using the data trend calculation method proposed in the application for the collected weld contour data, including the rising trend calculation result of the contour data in the data (the left side of the figure, the rising trend points are marked) and the falling trend calculation result of the contour data (the right side of the figure, the falling trend points are marked). Figure 3 Figure 3 Figure 3

[0104] The algorithm flow of the feature calculation module is as follows:

[0105] S111: Start the flow, and enter S112;

[0106] S112: Obtain the target point set Dm and Un, and enter S113;

[0107] S113: Calculate the straight line formula L connected by the first and last points in the point set Dm and Un, and enter S114;

[0108] S114: Calculate the distance from all points in the point set Dm and Un to the straight line L, and enter S115;

[0109] S115: Select the point with the maximum or minimum distance as the feature point, and enter S116;

[0110] S116: Output the feature point coordinates, and enter S117;

[0111] S117: End the flow.

[0112] As shown in the figure, after calculating the trend distribution of the collected weld contour data, the data feature points are calculated based on the trend distribution result in the figure. Figure 4 Figure 3 In the weld scene, the two feature points above can be defined as the groove edge points (see the two triangular marks in the left side of the figure), which are used to calculate the groove width or the misalignment amount, and the feature point below is the groove center point (see the bottom triangular mark in the right side of the figure), which is used for weld tracking or calculating the groove depth in combination with the groove edge points. Figure 4 Figure 4

[0113] ​​​​​​Further, referring to Figure 5 , the figure is a trend result calculated by the data trend distribution calculation method in the application on a collected weld contour data, and then the feature point calculation method is used to obtain the coordinate of the groove center point. The center point coordinate is continuously calculated for the subsequent weld contour data, and finally the feature data which can feedback the fluctuation of the weld is formed. The change of this data is gentle and regular. However, at the position of the weld joint, this data will start or end. More specifically, at the joint, the geometric feature (groove) of the weld changes from nothing to something or from something to nothing, resulting in discontinuity or dramatic change in trend in the spatial position of the center point sequence. Specifically, the data trend calculation method and the feature point calculation method are continuously used on the feature data, and in the weld center sequence point set, the feature mutation point representing the position of the weld joint 5 in the weld center sequence is identified, so as to obtain the coordinate range of the weld joint. The coordinate range of the weld joint is mapped to the 3D data to output the accurate weld joint coordinate, referring to Figure 7 .

[0114] The application uses quadratic trend analysis to sharply identify the singular point representing the beginning or end of the trajectory from the macro trajectory. This method can effectively separate the weld joint 5 from the continuous weld trajectory. Finally, by reversely mapping the feature points in the sequence to the original high-resolution single-frame 3D contour data, the accurate three-dimensional geometric size at the joint can be obtained, so as to output a spatial coordinate range for the welding gun to directly execute, realizing the closed loop from identification to positioning.

[0115] In another exemplary embodiment, a welding process weld joint accurate positioning system is provided, comprising:

[0116] A 3D contour scanning sensor 1 is used to scan the area where the weld 4 is located, and continuously obtain multiple frames of weld contour data;

[0117] A weld joint positioning module is in communication connection with the 3D contour scanning sensor 1, and a computer program is stored in the interior of the module. When the program is executed by the processor, the following process is realized:

[0118] For each frame of weld contour data obtained, the trend distribution calculation and the feature point calculation are sequentially performed until the scanning is completed; wherein the trend distribution calculation comprises:

[0119] The single-frame weld contour data is divided into multiple continuous regions, the slope of the straight line connecting the first and last points in each continuous region is calculated, and by comparing the slope of the straight line corresponding to each continuous region with the size of the preset threshold, the rising trend point set and the falling trend point set in the frame of weld contour data are identified;

[0120] The feature point calculation comprises:

[0121] Calculate the distance between all points in the rising trend point set and the line formed by the first and last points of the point set, and calculate the distance between all points in the falling trend point set and the line formed by the first and last points of the point set, find the point with the maximum or minimum distance, as the contour feature point;

[0122] Statistical all frame weld contour data corresponding to the contour feature points, extract the groove center point in each frame of weld contour data, and arrange them in the scanning order to generate the weld center sequence point set;

[0123] Perform the trend distribution calculation and feature point calculation on the weld center sequence point set in turn, extract the weld joint feature points and coordinates;

[0124] Map the weld joint feature points to the corresponding 3D weld contour data to obtain the three-dimensional space coordinates of the weld joint, and calculate the arc starting point coordinates executed by the welding gun according to the three-dimensional space coordinates of the weld joint;

[0125] The welding gun execution mechanism is in communication connection with the weld joint positioning module, and is used for receiving the arc starting point coordinates and driving the welding gun to move to the corresponding position to execute the arc starting process.

[0126] In some embodiments, the 3D contour scanning sensor 1 and the welding gun 2 are fixedly connected through a rigid mechanical connecting piece 7, and the geometric relationship of the mechanical connecting piece 7 is pre-calibrated and input to the weld joint positioning module.

[0127] The above specific embodiments are detailed descriptions of the present application, and cannot be considered as limiting the specific embodiments of the present application to these descriptions. For ordinary skilled persons in the technical field to which the present application belongs, a number of simple deductions and substitutions can be made without departing from the concept of the present application, and all of them should be considered as belonging to the protection scope of the present application.

Claims

1. A method for accurate positioning of a weld joint in a welding process, characterized in that The method comprises the following steps: S1. Scanning the area where the weld is located by a 3D profile scanning sensor, and continuously acquiring multiple frames of weld profile data; S2. For each frame of acquired weld profile data, sequentially performing trend distribution calculation and feature point calculation until the scanning is completed; wherein the trend distribution calculation comprises: dividing the single frame of weld profile data into multiple continuous regions, calculating the slope of the straight line connecting the first and last points in each continuous region, and identifying the rising trend point set and the falling trend point set in the frame of weld profile data by comparing the slope of the straight line corresponding to each continuous region with the size of the preset threshold value; the feature point calculation comprises: calculating the distance between all points in the rising trend point set and the straight line formed by the first and last points of the point set, calculating the distance between all points in the falling trend point set and the straight line formed by the first and last points of the point set, and finding the point with the maximum or minimum distance as the profile feature point; S3. Statistics of all frames of weld profile data corresponding to the profile feature points, extracting the groove center point in each frame of weld profile data, and arranging them in the scanning order to generate a weld center sequence point set; S4. For the weld center sequence point set, sequentially performing the trend distribution calculation and the feature point calculation, and extracting the weld joint feature points and coordinates; S5. Mapping the weld joint feature points to the corresponding 3D weld profile data to obtain the three-dimensional space coordinates of the weld joint, calculating the arc starting point coordinates executed by the welding gun according to the three-dimensional space coordinates of the weld joint, and outputting to the welding gun execution mechanism to execute the arc starting process.

2. The method of precisely positioning a weld joint in a welding process of claim 1, wherein, The scanning of the area where the weld is located by the 3D profile scanning sensor comprises: controlling the welding equipment carrying the 3D profile scanning sensor to move uniformly in a direction parallel to the weld, and making the laser line direction of the 3D profile scanning sensor perpendicular to the weld.

3. The method of precisely positioning a weld joint in a welding process of claim 1 wherein, The trend distribution calculation specifically comprises: dividing the single frame of weld profile data into K continuous regions with a length of L, and if the length of the last region is less than L, taking the remaining length as its length value; traversing the slopes of the straight lines corresponding to all continuous regions, and if it is detected that the absolute value of the slope of the straight line of a continuous region is greater than the threshold value, recording the index of the corresponding continuous region to the rising trend point set or the falling trend point set.

4. The method of precisely positioning a weld joint in a welding process of claim 1 wherein, When calculating the arc starting point coordinates executed by the welding gun, the geometric relationship between the 3D profile scanning sensor and the welding gun is fixed.

5. A system for accurate positioning of a weld joint in a welding process, characterized in that It comprises: a 3D profile scanning sensor for scanning the area where the weld is located, and continuously acquiring multiple frames of weld profile data; a weld joint positioning module in communication connection with the 3D profile scanning sensor, which internally stores a computer program, and the program is executed by the processor to realize the following process: for each frame of acquired weld profile data, sequentially performing trend distribution calculation and feature point calculation until the scanning is completed; wherein the trend distribution calculation comprises: dividing the single frame of weld profile data into multiple continuous regions, calculating the slope of the straight line connecting the first and last points in each continuous region, and identifying the rising trend point set and the falling trend point set in the frame of weld profile data by comparing the slope of the straight line corresponding to each continuous region with the size of the preset threshold value; the feature point calculation comprises: The distance between all points in the rising trend point set and the straight line formed by the first and last points of the point set is calculated, the distance between all points in the falling trend point set and the straight line formed by the first and last points of the point set is calculated, the point with the maximum or minimum distance is found out as the contour feature point; The contour feature points corresponding to the weld contour data of all frames are counted, the center points of the bevel in each frame of weld contour data are extracted, and the center points are arranged in the scanning order to generate a weld center sequence point set; The trend distribution calculation and feature point calculation are sequentially performed on the weld center sequence point set, and the weld joint feature points and coordinates are extracted; The weld joint feature points are mapped to the corresponding 3D weld contour data to obtain the three-dimensional space coordinates of the weld joint, and the arc starting point coordinates executed by the welding gun are calculated according to the three-dimensional space coordinates of the weld joint; The welding gun execution mechanism is in communication connection with the weld joint positioning module, is used for receiving the arc starting point coordinates, and drives the welding gun to move to the corresponding position to execute the arc starting process.

6. A system for accurately positioning a weld joint in a welding process as defined in claim 5, wherein, The 3D contour scanning sensor and the welding gun are fixedly connected through a rigid mechanical connecting piece, and the geometric relationship of the mechanical connecting piece is pre-calibrated and input into the weld joint positioning module.

Images