CN103252560A - Automatic weld tracking method based on laser vision sensing - Google Patents

Automatic weld tracking method based on laser vision sensing Download PDF

Info

Publication number
CN103252560A
CN103252560A CN2013100926138A CN201310092613A CN103252560A CN 103252560 A CN103252560 A CN 103252560A CN 2013100926138 A CN2013100926138 A CN 2013100926138A CN 201310092613 A CN201310092613 A CN 201310092613A CN 103252560 A CN103252560 A CN 103252560A
Authority
CN
China
Prior art keywords
welding gun
weld
tracking
laser
driver
Prior art date
Application number
CN2013100926138A
Other languages
Chinese (zh)
Other versions
CN103252560B (en
Inventor
莫胜撼
戴建树
喻宁娜
钟义广
何建辉
Original Assignee
广西机电职业技术学院
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 广西机电职业技术学院 filed Critical 广西机电职业技术学院
Priority to CN201310092613.8A priority Critical patent/CN103252560B/en
Publication of CN103252560A publication Critical patent/CN103252560A/en
Application granted granted Critical
Publication of CN103252560B publication Critical patent/CN103252560B/en

Links

Abstract

The invention relates to a weld tracking method in the technical field of welding, overcomes the defect that tracking precision is affected by lagged control caused by the fact that single feedback is used only to control a welding torch in the existing automatic weld tracking method, and provides an automatic weld tracking method based on laser vision sensing. The method includes: using a camera to continuously acquire weld image information, processing by a master control computer to obtain actual weld deviation to form feedback control, processing by the master control computer to obtain weld trajectory curvature (slope) information, and predicting a feedforward to form feedforward control; combining the feedback control and the feedforward control into feedforward feedback, calculating expected adjustment of the welding torch, and controlling the welding torch to move forward along a weld. The automatic weld tracking method based on laser vision sensing has the advantages that actual weld deviation of the welding torch can be calculated while a detection position of a sensor leads the welding torch, next weld direction can be predicted, and weld tracking precision and effect are improved greatly.

Description

Automatic seam tracking method based on laser-vision sensing

Technical field

The present invention relates to welding technology field, specifically is a kind of seam tracking system based on laser-vision sensing and method.

Background technology

Weld joint tracking is to realize the Automation of Welding problem demanding prompt solution.Laser-vision sensing is one of the most promising weld joint tracking sensing mode at present, extremely people's favor, and progressively in welding is produced, apply.Laser vision sensor and welding gun follower are fixed together, and sensor is installed in the place ahead of welding gun, for detection of welding seam position information.Sensor sends image information to image capturing system, and handle through image and to obtain welding seam position information, welding torch position and this position while welding are made comparisons, calculate the deviation that welding gun departs from weld seam, it is weld seam deviation, utilize this deviation to adjust follower as feeding back to controller, make welding gun aim at weld seam all the time, realize the purpose of weld joint tracking.

Adopt this approach sensor to carry out Weld Seam Tracking Control, because sensor and welding gun are fixed together, and the certain distance of the leading welding gun in sensor detection position, if directly weld seam deviation is calculated in the detected position of sensor, and adjusts welding gun with this deviation, when weld seam is more straight, this method is little to the tracking accuracy influence, when weld seam was oblique line or sweep, then error was very big, can't guarantee accurate tracking at all.The reason that this problem occurs be mistake will detect deviation be used as reality (really) weld seam deviation and handle.

At present the existing method of calculating the actual welds deviation is (as Jiang Lipei at " welding journal " (1998, Vol.19, " sensor detects the control model of formula seam tracking system in advance " of delivering No.3:184-189)), but, only rely on this single feedback to control the adverse consequences that welding gun can produce the control that lags behind, the precision that influence is followed the tracks of.

Summary of the invention

The present invention is directed to existing automatic seam tracking method relies on single feedback to control welding gun generation hysteresis control, this deficiency of precision that influence is followed the tracks of, a kind of laser-vision sensing seam tracking system is provided, under the situation of the leading welding gun in sensor detection position, it can calculate welding gun actual welds deviation, the trend of next step weld seam be can predict simultaneously, weld joint tracking precision and effect improved greatly.Technical scheme of the present invention is:

A kind of laser-vision sensing seam tracking system, comprise computer, welding gun, vision sensor, image acquisition units, motion control unit and driver element, vision sensor, image acquisition units, motion control unit and driver element are installed in the top of welding gun, image acquisition units is connected computer with motion control unit, and welding gun is driven by driver element; Welding gun the place ahead is provided with semiconductor laser, and the vision sensor below is provided with filter set.

The present invention sends a word line laser striped by semiconductor laser and slants on the weldment in welding gun the place ahead, absorb and send to image acquisition units through vision sensor, image acquisition units is stored in calculator memory with the image information that collects, if weld seam is a straight line, laser stripe is a broken line in the weld image that then collects; If weld seam is a curve, because weld groove place height is inconsistent, therefore laser stripe will be a broken line (broken line has reflected weld groove information) in the weld image that collects, this weld image image of machine is as calculated handled and recognizer, detect the weld seam center, calculate position deviation, the soldering joint automatic tracking algorithm goes out welding gun correction amount according to deviation, transmission mechanism property calculation, produce corresponding control signal by motion control unit, drive the welding gun motion through actuator unit, make welding gun aim at weld seam.

Driver element of the present invention comprises stepper motor and driver.What control module was exported is frequency pulse, carries out signal through driver and amplifies the motion of rear drive stepper motor drive mechanical transmission mechanism, moves thereby drive welding gun.

The present invention also provides a kind of automatic seam tracking method based on laser-vision sensing, may further comprise the steps:

(1) start-up system is set tracking parameter;

(2) begin to follow the tracks of, computer obtains weld image information through vision sensor and image acquisition units, calculates measured deviation, the line data of going forward side by side storage;

(3) computer calculates the actual welds deviation as feedback quantity, the line data of going forward side by side storage;

(4) computer calculates the feedforward amount by the weld seam slope prediction, the line data of going forward side by side storage;

(5) computer calculates welding gun and answers adjustment amount, and the line data of going forward side by side storage is by motion control unit output control signal, by the motion of driver element control welding gun;

(6) step (2) to step (5) circulation is carried out, and realizes soldering joint automatic tracking, finishes up to welding.

Described step (1) comprises the resolution ratio of the distance at setting laser fringe center and welding gun center, the relative moving speed of setting sampling period, setting welding gun and weldment and setting vision sensor.

The present invention obtains the actual welds deviation by Computer Processing and constitutes FEEDBACK CONTROL, obtains seam track curvature (slope) information by Computer Processing simultaneously, and then predicts that a feedforward amount constitutes FEEDFORWARD CONTROL; FEEDBACK CONTROL and FEEDFORWARD CONTROL in conjunction with constituting feedforward feedback, are calculated welding gun and answered adjustment amount, and the control welding gun advances along weld seam.

Below the inventive method is described further, step is as follows:

(1) start-up system, set tracking parameter: the first, the distance at setting laser fringe center (test point) and welding gun center, and guarantee that laser stripe overlaps at the picture centre line; The second, the initialization system sampling time; The 3rd, set the speed that weldment phase Butt welding gun moves; The 4th, the resolution ratio of setting camera review.

(2) begin to follow the tracks of, computer obtains weld image information through vision sensor and image acquisition units, calculates measured deviation , the line data of going forward side by side storage.

(3) computer calculates the actual welds deviation As feedback quantity, the line data of going forward side by side storage.

??????????????????(a)

In the formula, It is the actual adjustment amount of iT cycle welding gun.

(4) computer calculates the feedforward amount by the weld seam slope prediction, the line data of going forward side by side storage.Seam track exists ITSlope constantly Can be by ITThe increment of cycle on the y direction With the increment on the x direction The merchant calculate, therefore, by slope The welding gun adjustment amount of prediction is In actual applications, can select a bit of seam track curve in welding gun front and back, ask for their average increment Feedforward amount as the weld seam slope prediction.

During calculating, by formula (b) calculates respectively earlier Extremely Value.

??????????????(b)

Calculate the according to following formula (c) then ITMoment seam track average increment

???????????????????(c)

In the formula, NHop count (calculating by the sampling period number) for selected welding gun the place ahead seam track curve.

(5) computer calculates welding gun and answers adjustment amount , the line data of going forward side by side storage.By the feedforward amount And feedback quantity Addition obtains welding gun and answers adjustment amount , namely

?????????????????????(d)

According to the characteristic output adjustment amount of mechanism, by control module output control information, by the motion of driver element control welding gun.

(6) step (2) to step (5) circulation is carried out, and realizes soldering joint automatic tracking, finishes up to welding.

The present invention is not limited to laser vision sensor; the leading welding gun of every employing sensor detects position while welding; can use this method to realize weld joint tracking; and embodiment does not limit the present invention in any form; all take to be equal to replace or technical scheme that the form of equivalent transformation obtains, all within protection scope of the present invention.

The invention has the beneficial effects as follows:

1. can calculate welding gun actual welds deviation and seam track slope, utilize both to constitute feedforward feedback control, improve weld joint tracking precision and effect greatly.

2, prediction regulation and control amount is arranged, feedback is timely, and welding gun control does not lag behind.

Description of drawings

Fig. 1 is the actual adjustment amount of answering of the present invention The calculating principle schematic.

The specific embodiment

Embodiment: from motion tracking one circular weld seam

(1) before the tracking beginning, installs weldment, start-up system.The distance at setting laser fringe center (test point) and welding gun center 40Mm, and guarantee that laser stripe overlaps at the picture centre line; The initialization system sampling time (identical with the IMAQ cycle) is 0.5sThe speed that setting weldment phase Butt welding gun moves is 20mm/s, and therefore, can calculate accumulation number n is 40/ (20*0.5)=4; The image resolution ratio of setting vision sensor is 10pixel/mm (pixel/millimeter).

(2) begin to follow the tracks of, computer obtains a broken line laser stripe image by vision sensor and image acquisition units, calculates measured deviation simultaneously and equals the difference that coordinate at zero point in the image deducts detected weld seam center position coordinates in image.Because measured deviation value unit is pixel (pixel), the welding gun adjustment amount is obtained by stepper motor output, and tracker is exported to the pulse signal that stepper motor is changeable frequency (unit is Hz), consider the tracker output signal, unified dimension is Hz, also should multiply by the K value when therefore, calculating the measured deviation value.

(3) computer calculates the actual welds deviation according to formula (a) As feedback quantity, the line data of going forward side by side storage.

(4) computer calculates the feedforward amount , get the 1 section weld seam in welding gun the place ahead, then formula (c) becomes

(5) computer calculates and answers adjustment amount , with formula (a) and (b) addition obtain welding gun adjustment amount computing formula and be , because accumulation number n is 4, so welding gun answers the adjustment amount computing formula finally to become

Computer carries out signal through driver and amplifies the motion of rear drive stepper motor drive mechanical transmission mechanism, thereby drive the welding gun motion by the pulse of control module output frequency.

(6) step (2) to step (5) circulation is carried out, and realizes soldering joint automatic tracking, finishes up to circular weld seam welding.

Satisfying under the same weld joint tracking precision conditions, the weld seam that utilizes this method to follow the tracks of, owing to have feedback quantity and feedforward amount simultaneously, the weld seam precision can be higher.

Claims (4)

1. laser-vision sensing seam tracking system, comprise computer, welding gun, vision sensor, image acquisition units, motion control unit and driver element, vision sensor, image acquisition units, motion control unit and driver element are installed in the top of welding gun, image acquisition units is connected computer with motion control unit, and welding gun is driven by driver element; It is characterized in that: welding gun the place ahead is provided with semiconductor laser, and the vision sensor below is provided with filter set.
2. laser-vision sensing seam tracking system according to claim 1, it is characterized in that: described driver element comprises stepper motor and driver.
3. automatic seam tracking method based on laser-vision sensing is characterized in that: may further comprise the steps:
(1) start-up system is set tracking parameter;
(2) begin to follow the tracks of, computer obtains weld image information through vision sensor and image acquisition units, calculates measured deviation, the line data of going forward side by side storage;
(3) computer calculates the actual welds deviation as feedback quantity, the line data of going forward side by side storage;
(4) computer calculates the feedforward amount by the weld seam slope prediction, the line data of going forward side by side storage;
(5) computer calculates welding gun and answers adjustment amount, and the line data of going forward side by side storage is by motion control unit output control signal, by the motion of driver element control welding gun;
(6) step (2) to step (5) circulation is carried out, and realizes soldering joint automatic tracking, finishes up to welding.
4. the automatic seam tracking method based on laser-vision sensing according to claim 3 is characterized in that: step (1) comprise setting laser fringe center and welding gun center distance, set the sampling period, set the resolution ratio of welding gun and relative moving speed and the setting vision sensor of weldment.
CN201310092613.8A 2013-03-22 2013-03-22 Based on the automatic seam tracking method of laser-vision sensing CN103252560B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201310092613.8A CN103252560B (en) 2013-03-22 2013-03-22 Based on the automatic seam tracking method of laser-vision sensing

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201310092613.8A CN103252560B (en) 2013-03-22 2013-03-22 Based on the automatic seam tracking method of laser-vision sensing

Publications (2)

Publication Number Publication Date
CN103252560A true CN103252560A (en) 2013-08-21
CN103252560B CN103252560B (en) 2016-01-06

Family

ID=48956960

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201310092613.8A CN103252560B (en) 2013-03-22 2013-03-22 Based on the automatic seam tracking method of laser-vision sensing

Country Status (1)

Country Link
CN (1) CN103252560B (en)

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104741790A (en) * 2015-03-30 2015-07-01 中国石油天然气集团公司 Pipeline all-position laser and arc combined welding system
CN105195888A (en) * 2015-10-09 2015-12-30 航天工程装备(苏州)有限公司 Two-dimensional laser tracking compensation method for stirring friction welding plane
CN105750774A (en) * 2016-05-06 2016-07-13 烟台艾森信息技术股份有限公司 Three-dimension intelligent laser visual tracking welding system and using method
CN105855668A (en) * 2016-05-27 2016-08-17 廊坊智通机器人系统有限公司 Linear online seam tracking method for arc welding robot during welding
CN106392267A (en) * 2016-11-28 2017-02-15 华南理工大学 Line laser real-time welding seam tracking method for six-freedom-degree welding robot
CN107350649A (en) * 2016-05-10 2017-11-17 费希尔控制产品国际有限公司 Prediction algorithm from the welding deformation that flange is added to butt welding or socket weld ends valve body moulding
CN107378246A (en) * 2016-05-16 2017-11-24 大族激光科技产业集团股份有限公司 A kind of method and system for correcting laser welding track
CN109598760A (en) * 2018-12-07 2019-04-09 北京博清科技有限公司 Image processing method and image processing apparatus

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH0550240A (en) * 1991-08-27 1993-03-02 Amada Co Ltd Automatic welding device and its welding condition control method
US5347101A (en) * 1994-02-07 1994-09-13 Mcdermott International, Inc. Automatic tracking system for pipeline welding
CN1600488A (en) * 2004-10-21 2005-03-30 上海交通大学 Method for autonomous tracking welding line
CN1782659A (en) * 2004-12-02 2006-06-07 中国科学院自动化研究所 Welding seam tracking sight sensor based on laser structure light
CN1935437A (en) * 2006-09-29 2007-03-28 南昌大学 Bending weld seam tracking method based on CCD

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH0550240A (en) * 1991-08-27 1993-03-02 Amada Co Ltd Automatic welding device and its welding condition control method
US5347101A (en) * 1994-02-07 1994-09-13 Mcdermott International, Inc. Automatic tracking system for pipeline welding
CN1600488A (en) * 2004-10-21 2005-03-30 上海交通大学 Method for autonomous tracking welding line
CN1782659A (en) * 2004-12-02 2006-06-07 中国科学院自动化研究所 Welding seam tracking sight sensor based on laser structure light
CN1935437A (en) * 2006-09-29 2007-03-28 南昌大学 Bending weld seam tracking method based on CCD

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
席峰等: ""基于视觉传感的焊缝跟踪控制系统"", 《仪表技术与传感器》, no. 5, 30 May 2006 (2006-05-30) *
莫胜撼等: ""塔吊臂架角钢外角接焊缝跟踪系统"", 《电焊机》, vol. 39, no. 9, 20 September 2009 (2009-09-20), pages 29 - 33 *

Cited By (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104741790A (en) * 2015-03-30 2015-07-01 中国石油天然气集团公司 Pipeline all-position laser and arc combined welding system
CN105195888A (en) * 2015-10-09 2015-12-30 航天工程装备(苏州)有限公司 Two-dimensional laser tracking compensation method for stirring friction welding plane
CN105750774A (en) * 2016-05-06 2016-07-13 烟台艾森信息技术股份有限公司 Three-dimension intelligent laser visual tracking welding system and using method
CN107350649A (en) * 2016-05-10 2017-11-17 费希尔控制产品国际有限公司 Prediction algorithm from the welding deformation that flange is added to butt welding or socket weld ends valve body moulding
CN107378246A (en) * 2016-05-16 2017-11-24 大族激光科技产业集团股份有限公司 A kind of method and system for correcting laser welding track
CN105855668A (en) * 2016-05-27 2016-08-17 廊坊智通机器人系统有限公司 Linear online seam tracking method for arc welding robot during welding
CN106392267A (en) * 2016-11-28 2017-02-15 华南理工大学 Line laser real-time welding seam tracking method for six-freedom-degree welding robot
CN106392267B (en) * 2016-11-28 2018-09-14 华南理工大学 A kind of real-time welding seam tracking method of six degree of freedom welding robot line laser
CN109598760A (en) * 2018-12-07 2019-04-09 北京博清科技有限公司 Image processing method and image processing apparatus

Also Published As

Publication number Publication date
CN103252560B (en) 2016-01-06

Similar Documents

Publication Publication Date Title
US9197810B2 (en) Systems and methods for tracking location of movable target object
CN104457569B (en) A kind of large-scale composite board geometric parameter vision measuring method
CN103900532B (en) Degree of depth image capturing device, its bearing calibration and method for measurement
CN104942404B (en) Dual wavelength binocular vision welding seam tracking method and tracking system
CN101493682B (en) Generating device of processing robot program
CN104571112B (en) Pilotless automobile lateral control method based on turning curvature estimation
JP2018527669A (en) Method, apparatus and unmanned aerial vehicle for terrain following flight of unmanned aerial vehicles
CN103273171B (en) Visual tracking monitoring system and method in automatic corrugated thin plate fillet weld welding
CN101614445B (en) Method for improving accuracy of control of automatic sun track following of heliostat
RU2404036C2 (en) Laser welding head
CN1289252C (en) Method for autonomous tracking welding line
US20050021170A1 (en) Method of controlling the welding of a three-dimensional structure
US9491448B2 (en) Laser videogrammetry
CN103934571A (en) Thick plate robot welding system and multilayer multiple-pass weld real-time tracking and planning method
US6741364B2 (en) Apparatus for determining relative positioning of objects and related methods
US9715730B2 (en) Three-dimensional measurement apparatus and robot system
CN102303190B (en) Method for visually tracking plane abut-jointed weld beam by linear laser
CN102615390A (en) Swing arc-based multi-layer and multi-channel weld tracking system and identification method thereof
US20090179021A1 (en) Welding robot
EP0678066A1 (en) A system and method for tracking a feature on an object using a redundant axis
CN104400279B (en) Pipeline space weld seam based on CCD identifies the method with trajectory planning automatically
CN104959706B (en) The welding seam tracking method of plate sheet welding swing type edge effect capacitance sensor
CN103480991B (en) Thin steel plate narrow welding joint online visual inspection and control device
CN103600155B (en) Real-time monitoring device for welding robot welding process
CN106181162B (en) A kind of real-time weld joint tracking detection method based on machine vision

Legal Events

Date Code Title Description
PB01 Publication
C06 Publication
SE01 Entry into force of request for substantive examination
C10 Entry into substantive examination
GR01 Patent grant
C14 Grant of patent or utility model
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20160106

Termination date: 20190322

CF01 Termination of patent right due to non-payment of annual fee