CN203791808U - Intelligent welding robot based on machine vision - Google Patents
Intelligent welding robot based on machine vision Download PDFInfo
- Publication number
- CN203791808U CN203791808U CN201320700854.1U CN201320700854U CN203791808U CN 203791808 U CN203791808 U CN 203791808U CN 201320700854 U CN201320700854 U CN 201320700854U CN 203791808 U CN203791808 U CN 203791808U
- Authority
- CN
- China
- Prior art keywords
- machine vision
- welding robot
- robot based
- welding
- controller
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
Landscapes
- Manipulator (AREA)
Abstract
The utility model relates to the field of robots, in particular to an intelligent welding robot based on machine vision. The intelligent welding robot based on machine vision comprises a robot body and a weld joint tracking system. The weld joint tracking system comprises an image processing subsystem and a deviation rectification control subsystem. The image processing subsystem comprises a camera. The deviation rectification control subsystem comprises a controller. The controller is connected with a stepping motor through a stepping motor driver. An end effector is connected to an output shaft of the stepping motor. The camera is connected with the controller. The intelligent welding robot can automatically track a weld joint in real time with high accuracy and the requirements for the automatic welding technology are met. The welding robot has a strict requirement for accuracy of the motion track of the end effector, the posture of a welding gun on the whole track needs to be continuously adjusted in the spatial position welding process, a matched control unit is arranged, so that position and accuracy errors caused by assembly, deformation and the like are overcome, and high-quality automatic welding production is achieved.
Description
Technical field
The utility model relates to robot field, relates in particular to a kind of Intelligent welding robot based on machine vision.
Background technology
In welding robot welding process, due to the impact of environmental factor, as: strong arc light radiation, high temperature, flue dust, splash, groove situation, mismachining tolerance, fixture clamping precision, surface state and workpiece thermal deformation etc., the variation of actual welding condition tends to cause welding torch to depart from weld seam, thereby it is even failed to cause welding quality to decline.
To different weld task, in control technology, need to adopt different solutions.On six degree of freedom welding robot platform, utilize the relatively accurate feature of motion, often utilize the method butt welded seam position of binocular vision to position, and then butt welded seam trend carries out matching, adopt Kalman Filter Technology to plan the motion of manipulator; Disturb the impact on control system in order to reduce welding surroundings, adopt robust estimation method to improve the signal to noise ratio of metrical information.But in binocular vision method, generally need to demarcate in advance the inside and outside parameter of video camera, the problem of existence is that calibration process is loaded down with trivial details, and the demarcation of trick relation can not reach very high precision, after operation conditions change, often need again again to demarcate.The flexibility of binocular vision method is low, is not suitable for on-the-spot application.
Current vision control, being based upon vision measurement is on accurate, reliable basis.But, in the situation that there is more serious noise jamming, only rely on piece image to be often difficult to obtain clarification of objective, can not realize vision measurement.Because noise has randomness, so when thering is the sequence image of serious noise and add up, can eliminate influence of noise, obtain stable statistical property, and then can realize the measurement to target according to statistical property.For example, in welding process weldment image be subject to arc light and the interference ratio of splashing more serious, be directly difficult to extract weld seam from present image.The variation of weld seam wants slow a lot of with respect to arc light with splashing, and in consecutive frame image, the position of weld seam is substantially constant, but arc light has larger difference with splashing.Therefore, utilize the logical “and” of consecutive frame image can eliminate preferably arc light and the interference of splashing.Eliminating arc light and splashing in the image disturbing, often can extract multiple candidate's characteristics of weld seams.According to the statistical law of the characteristics of weld seam extracting in former some two field pictures, in these candidate feature, select the current characteristics of weld seam of conduct that probability is the highest, for vision closed-loop control, realize reliable and stable weld joint tracking.
For above problem, need the Intelligent welding robot based on machine vision that wants a kind of high accuracy real-time automatic tracking that can butt welded seam badly.
Utility model content
The purpose of this utility model is to provide a kind of Intelligent welding robot based on machine vision, and this Intelligent welding robot based on machine vision can realize the high accuracy real-time automatic tracking of the wide 1mm～15mm of opposite joint weld seam, meets the needs of automatic soldering technique.
For reaching this object, the utility model by the following technical solutions:
An Intelligent welding robot based on machine vision, comprises robot body and seam tracking system;
Described seam tracking system comprises image processing subsystem and correction RACS;
Described image processing subsystem comprises camera;
Described correction RACS comprises controller, and described controller is connected with stepper motor by stepper motor driver, on the output shaft of described stepper motor, is connected with end effector, and described camera is connected with controller.
As preferably, described robot body comprises vertical axis and transverse movement axle, and described stepper motor is arranged on described transverse movement axle, can drive end effector to move left and right.
As preferably, described end effector is welding gun.
As preferably, the number of described camera is 2, is arranged on respectively the front end both sides of welding gun.
As preferably, described camera and welding gun are rigidly connected.
As preferably, between described camera and welding gun, be provided with protective baffle plate.
As preferably, described controller is PLC.
As preferably, described correction RACS also comprises touch-screen, and described touch-screen is connected with described controller.
As preferably, described controller, touch-screen and stepping electric drive are arranged in a control cabinet.
As preferably, described correction RACS also comprises action button, and described action button is connected with described controller.
The beneficial effects of the utility model are: the utility model provides a kind of Intelligent welding robot based on machine vision, and it comprises robot body and seam tracking system; Described seam tracking system comprises image processing subsystem and correction RACS; Described image processing subsystem comprises camera; Described correction RACS comprises controller, and described controller is connected with stepper motor by stepper motor driver, on the output shaft of described stepper motor, is connected with end effector, and described camera is connected with controller.This Intelligent welding robot based on machine vision can realize the high accuracy real-time automatic tracking of the wide 1mm～15mm of opposite joint weld seam, meets the needs of automatic soldering technique.Welding robot has strict required precision to the movement locus of end effector, posture of welding torch when weld locus all needs continuous setup on whole track, there is the supporting control module such as sensing system, expert system, overcome the position and the trueness error that cause because of factors such as assembling, distortion, produce to realize real high-quality automatic welding.Correction RACS requires arc welding robot can detect in real time the deviation of weld seam according to the variation of welding condition, and adjusts path of welding and welding parameter, ensures the reliability of welding quality.
Brief description of the drawings
Fig. 1 is the structural representation of the Intelligent welding robot based on machine vision that provides of the utility model detailed description of the invention;
Fig. 2 is the structural representation of the seam tracking system that provides of the utility model detailed description of the invention.
Wherein:
1, robot body; 2, seam tracking system;
11, transverse movement axle; 12, vertical axis;
21, controller; 22, stepper motor driver; 23, stepper motor; 24, camera; 25, end effector; 26, touch-screen; 27, action button.
Detailed description of the invention
Further illustrate the technical solution of the utility model below in conjunction with accompanying drawing and by detailed description of the invention.
The Intelligent welding robot based on machine vision described in this programme by the industrial robot of typical configurations as Cartesian robot, revolute robot, parallel robot and automatically THE WELD SEAM TRACKING TECHNOLOGY combine, developing intellectual resource welding robot, and be applied in the welding sequence of the industries such as stainless steel, tubulation, container.With the example that is spliced into of 1mm～15mm sheet material in production process, realize the high accuracy real-time automatic tracking of the wide 1mm～15mm of opposite joint weld seam, meet the needs of automatic soldering technique.Specific works comprises that exploitation has the robot of multiple frees degree, obtains feedback signal by automatic seam tracking system, and robot adjusts the attitude of welding gun automatically according to signal, and the variation of following weld seam completes high-quality welding automatically.Robot has 2～5 frees degree as required, and load capacity is not less than 10kg, the tracking velocity 1m/s of seam tracking system, and tracking accuracy ± 0.2mm, follows the tracks of envelope width 1～15mm.Fig. 5 is configuration picture and the welding robot prototype of tracking system.
Fig. 1 is the structural representation of the Intelligent welding robot based on machine vision that provides of the utility model detailed description of the invention; Fig. 2 is the structural representation of the seam tracking system that provides of the utility model detailed description of the invention.As depicted in figs. 1 and 2, be somebody's turn to do the Intelligent welding robot based on machine vision, comprise robot body 1 and seam tracking system 2.Described seam tracking system 2 comprises image processing subsystem and correction RACS.Described image processing subsystem comprises camera 24.Described correction RACS comprises controller 21, and described controller 21 is connected with stepper motor 23 by stepper motor driver 22, is connected with end effector 25 on the output shaft of described stepper motor 23, and described camera 24 is connected with controller 21.Correction RACS requires arc welding robot can detect in real time the deviation of weld seam according to the variation of welding condition, and adjusts path of welding and welding parameter, ensures the reliability of welding quality.
In the present embodiment, as preferred version, described robot body comprises vertical axis 12 and transverse movement axle 11, and described stepper motor 23 is arranged on described transverse movement axle 11, can drive end effector 25 to move left and right.
In the present embodiment, as preferred version, described end effector 25 is welding gun.
In the present embodiment, as preferred version, the number of described camera 24 is 2, is arranged on respectively the front end both sides of welding gun.
In the present embodiment, as preferred version, described camera 24 is rigidly connected with welding gun.
In the present embodiment, as preferred version, between described camera 24 and welding gun, be provided with protective baffle plate, to prevent the damage camera that splashes.
In the present embodiment, as preferred version, described controller 21 is PLC.PLC(Programmable Logic Controller) be programmable logic controller (PLC), it adopts a class programmable memory, for its internally stored program, actuating logic computing, sequential control, regularly, the user oriented instruction such as counting and arithmetical operation, and by numeral or analog pattern input/output various types of machinery of control or production process.
In the present embodiment, as preferred version, described correction RACS also comprises touch-screen 26, and described touch-screen 26 is connected with described controller 21.
In the present embodiment, as preferred version, described controller 21, touch-screen 26 and stepping electric drive 22 are arranged in a control cabinet.Control cabinet can move together with welding tractor.
In the present embodiment, as preferred version, described correction RACS also comprises action button 27, and described action button 27 is connected with described controller 21.
In welding initial point vision location, adopt two camera acquisition weld images that are arranged on welding gun both sides.Because welding gun is fixed with respect to the position of video camera, the image coordinate of welding gun rifle point in camera review remains unchanged.In the weld image gathering, first extract weld edge, then utilize weld edge to determine the image coordinate of welding initiation feature point.In the time that welding gun rifle point is not aimed at the initiation feature point of weld seam, both characteristic coordinates differences in two camera reviews, can utilize fuzzy control to determine the direction of motion and the amount of exercise of welding gun accordingly, and guiding welding gun rifle point arrives welding initiation feature point position.In the time that the image coordinate of the characteristic point image coordinate of welding gun rifle point and welding initiation feature point is identical, show that welding gun rifle point aimed at the initiation feature point of weld seam, to weld initial point and located, vision control procedure finishes.
In the welding initial point centering stage, welding gun does not also have the starting the arc, can utilize natural lighting condition to gather weld image.But the image obtaining under natural lighting condition, background more complicated, is subject to illumination effect large, and when video camera depart from weld seam and from weld seam away from time, be difficult to weld seam to be separated from background.Therefore, the initial pad location based on vision, should video camera from weld seam close to, can obtain obvious weld image time, carry out.Due to weld image, the image of weld seam and its periphery weldment can be different, utilizes conventional Boundary extracting algorithm as Canny operator, Laplace operator etc., is difficult to weld seam to split rapidly.
For addressing the above problem, carry out histogram equalization by the gradient image of butt welded seam image, the gray level of the pixel of reflecting edge information in gradient image is compressed in a less scope.Then,, according to the grey level distribution of gained gradient image after histogram equalization, determine the overall binary-state threshold of image.Taking window half-breadth as step-length in image sliding window, determine the local binarization threshold value of image in window.The combining image overall situation and local binary-state threshold, obtain the self-adaption binaryzation threshold value in window, and window inside gradient image is carried out to binaryzation.In addition, also designed a kind of method of iteration denoising, to solve the problem that has much noise in gained bianry image.
The vision system adopting in welding initial point centering process, adopt apery vision control principle, deviation according to the initiation feature point of weld seam in two video cameras with welding gun rifle point image coordinate, utilize fuzzy control to determine the direction of motion and the amount of exercise of welding gun, guiding welding gun rifle point arrives welding initiation feature point position.This vision control system does not need the inside and outside parameter of video camera to demarcate in advance, easy to use, and strong adaptability has solved a bottleneck problem that restricts vision system commercial Application.The welding initial point centering vision control system that this programme forms, has fault-tolerant ability, its strong robustness, and reliability is high.
In weld seam correction process, adopt a video camera to gather weld image in the front of welding position, extract the image coordinate of weld bead feature points.Starting the welding stage, gathering weld image N time, utilizing the statistical property of the weld bead feature points of extracting for N time, determining the desired image coordinate of weld bead feature points.In welding process, utilize the deviation statistics characteristic of present image coordinate and the desired image coordinate of weld bead feature points, the control law of structure vision control, realizes the tracking of welding gun rifle point butt welded seam.
Compared with the robot of the Intelligent welding robot based on machine vision described in this programme and the fixed-point operations such as common carrying, loading and unloading, assembling, welding robot has strict required precision to the movement locus of end effector (welding gun), and the posture of welding torch when welding of locus all needs continuous setup on whole track.For this reason, welding robot should have the supporting control module such as sensing system, expert system, overcomes the position and the trueness error that cause because of factors such as assembling, distortion, produces to realize real high-quality automatic welding.
In the situation that can not ensureing that weld image feature a hundred per cent is correct, how to realize reliably, follow the tracks of accurately weld seam, be the key of vision control strategy design.Utilize apery Intelligent Control Strategy, in conjunction with the statistical law of priori and the weld image feature of weld seam, can effectively solve the inaccurate difficult problem such as even follow the tracks of unsuccessfully of weld joint tracking that random disturbances causes.
The seam tracking system of this programme carries out vision control based on statistical information, even in the not bery reliable situation of visual sensing information, still can realize reliable weld joint tracking.The seam tracking system of this project has the advantages such as volume is little, reliability is high, easy to use, is suitable for the nonstandard application scenario of weld groove, is suitable for domestic industry application.
The utility model can not only reduce human cost, aspect hommization, also can make to free people is severe from environment, labour intensity is large industrial production; Aspect production assurance, can avoid the subjective factor in traditional manual operation process, product quality supervision introduced, make the class of product brand produce qualitative leap; Aspect productivity ratio, use after robot automation production equipment, than more accurately standard more of manual operations, action executing is quick and never tired, can greatly raise labour productivity.
Know-why of the present utility model has below been described in conjunction with specific embodiments.These are described is in order to explain principle of the present utility model, and can not be interpreted as by any way the restriction to the utility model protection domain.Based on explanation herein, those skilled in the art does not need to pay performing creative labour can associate other detailed description of the invention of the present utility model, within these modes all will fall into protection domain of the present utility model.
Claims (10)
1. the Intelligent welding robot based on machine vision, is characterized in that, comprises robot body (1) and seam tracking system (2);
Described seam tracking system (2) comprises image processing subsystem and correction RACS;
Described image processing subsystem comprises camera (24);
Described correction RACS comprises controller (21), described controller (21) is connected with stepper motor (23) by stepper motor driver (22), on the output shaft of described stepper motor, be connected with end effector (25), described camera (24) is connected with controller.
2. the Intelligent welding robot based on machine vision according to claim 1, it is characterized in that, described robot body (1) comprises vertical axis (12) and transverse movement axle (11), described stepper motor is arranged on described transverse movement axle, can drive end effector (25) to move left and right.
3. the Intelligent welding robot based on machine vision according to claim 1, is characterized in that, described end effector (25) is welding gun.
4. the Intelligent welding robot based on machine vision according to claim 3, is characterized in that, the number of described camera (24) is 2, is arranged on respectively the front end both sides of welding gun.
5. the Intelligent welding robot based on machine vision according to claim 4, is characterized in that, described camera (24) is rigidly connected with welding gun.
6. the Intelligent welding robot based on machine vision according to claim 5, is characterized in that, between described camera (24) and welding gun, is provided with protective baffle plate.
7. the Intelligent welding robot based on machine vision according to claim 1, is characterized in that, described controller (21) is PLC.
8. the Intelligent welding robot based on machine vision according to claim 1, is characterized in that, described correction RACS also comprises touch-screen (26), and described touch-screen is connected with described controller.
9. the Intelligent welding robot based on machine vision according to claim 8, is characterized in that, described controller, touch-screen and stepping electric drive are arranged in a control cabinet.
10. the Intelligent welding robot based on machine vision according to claim 1, is characterized in that, described correction RACS also comprises action button (27), and described action button is connected with described controller.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201320700854.1U CN203791808U (en) | 2013-11-06 | 2013-11-06 | Intelligent welding robot based on machine vision |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201320700854.1U CN203791808U (en) | 2013-11-06 | 2013-11-06 | Intelligent welding robot based on machine vision |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN203791808U true CN203791808U (en) | 2014-08-27 |
Family
ID=51374077
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201320700854.1U Expired - Fee Related CN203791808U (en) | 2013-11-06 | 2013-11-06 | Intelligent welding robot based on machine vision |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN203791808U (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104493336A (en) * | 2014-12-25 | 2015-04-08 | 重庆天量电子科技有限公司 | Welded joint detecting and tracking system and method based on video analysis |
| CN104588837A (en) * | 2015-01-20 | 2015-05-06 | 柳州铁道职业技术学院 | Intelligent coupling system of arc welding robot |
| CN104690422A (en) * | 2015-01-11 | 2015-06-10 | 沈阳汇能机器人自动化有限公司 | Robot laser welding visual inspection programmed control system and implementing method thereof |
| CN108296608A (en) * | 2017-12-19 | 2018-07-20 | 天津市春信制冷净化设备有限公司 | A kind of high-level laboratory wall automatic soldering device |
| CN112475701A (en) * | 2020-11-21 | 2021-03-12 | 苏州卓罗智能科技有限公司 | Intelligent welding robot based on machine vision |
| CN114043080A (en) * | 2021-11-22 | 2022-02-15 | 吉林大学 | A kind of stainless steel intelligent laser welding processing method |
| CN115151367A (en) * | 2020-02-17 | 2022-10-04 | 株式会社神户制钢所 | Automatic welding system, automatic welding method, learning device, learned model generation method, learned model, estimation device, estimation method, and program |
-
2013
- 2013-11-06 CN CN201320700854.1U patent/CN203791808U/en not_active Expired - Fee Related
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104493336A (en) * | 2014-12-25 | 2015-04-08 | 重庆天量电子科技有限公司 | Welded joint detecting and tracking system and method based on video analysis |
| CN104690422A (en) * | 2015-01-11 | 2015-06-10 | 沈阳汇能机器人自动化有限公司 | Robot laser welding visual inspection programmed control system and implementing method thereof |
| CN104588837A (en) * | 2015-01-20 | 2015-05-06 | 柳州铁道职业技术学院 | Intelligent coupling system of arc welding robot |
| CN108296608A (en) * | 2017-12-19 | 2018-07-20 | 天津市春信制冷净化设备有限公司 | A kind of high-level laboratory wall automatic soldering device |
| CN115151367A (en) * | 2020-02-17 | 2022-10-04 | 株式会社神户制钢所 | Automatic welding system, automatic welding method, learning device, learned model generation method, learned model, estimation device, estimation method, and program |
| CN115151367B (en) * | 2020-02-17 | 2024-01-12 | 株式会社神户制钢所 | Automatic welding system, automatic welding method, learning device, neural network system, and estimation device |
| CN112475701A (en) * | 2020-11-21 | 2021-03-12 | 苏州卓罗智能科技有限公司 | Intelligent welding robot based on machine vision |
| CN114043080A (en) * | 2021-11-22 | 2022-02-15 | 吉林大学 | A kind of stainless steel intelligent laser welding processing method |
| CN114043080B (en) * | 2021-11-22 | 2024-01-26 | 吉林大学 | Intelligent laser welding treatment method for stainless steel |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN203791808U (en) | Intelligent welding robot based on machine vision | |
| CN210046133U (en) | Welding seam visual tracking system based on laser structured light | |
| CN104384765B (en) | Based on the automatic soldering method of threedimensional model and machine vision | |
| CN106392267B (en) | A kind of real-time welding seam tracking method of six degree of freedom welding robot line laser | |
| CN104588838B (en) | A kind of automatic tracing control of welding gun on-line amending system and control method thereof | |
| CN206263418U (en) | A kind of real-time seam tracking system of six degree of freedom welding robot line laser | |
| CN106181162B (en) | A kind of real-time weld joint tracking detection method based on machine vision | |
| Gu et al. | Autonomous seam acquisition and tracking system for multi-pass welding based on vision sensor | |
| CN113172307A (en) | Industrial robot system of visual module based on laser and visible light fusion | |
| Shen et al. | Real-time seam tracking technology of welding robot with visual sensing | |
| CN113146172B (en) | Multi-vision-based detection and assembly system and method | |
| CN112191993B (en) | Argon arc welding seam tracking system and method | |
| CN107999955A (en) | A kind of six-shaft industrial robot line laser automatic tracking system and an automatic tracking method | |
| CN103231162A (en) | Device and method for visual detection of welding quality of robot | |
| CN108527332A (en) | A kind of seam track off-line calibration method based on structured light vision sensor | |
| CN104175031A (en) | Welding robot with automatic correction function and welding method employing same | |
| CN101770710A (en) | Laser-vision sensing assisted remote teaching method for remote welding | |
| CN103567607A (en) | Welding-seam tracking method | |
| CN106271081A (en) | Three coordinate rectangular robot line laser seam tracking system and trackings thereof | |
| Chen et al. | A robust visual servo control system for narrow seam double head welding robot | |
| CN115091028B (en) | Welding method and device based on laser real-time tracking, electronic equipment and medium | |
| CN113269762A (en) | Screen defect detection method, system and computer storage medium | |
| CN114654465A (en) | Welding seam tracking and extracting method based on line laser structure optical vision sensing | |
| CN108788467A (en) | A kind of Intelligent Laser welding system towards aerospace structural component | |
| CN113828892A (en) | HDR image-based molten pool center identification system and weld joint tracking method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20140827 Termination date: 20141106 |
|
| EXPY | Termination of patent right or utility model |