CN111061231A - Weld assembly gap and misalignment feed-forward molten pool monitoring system and penetration monitoring method - Google Patents
Weld assembly gap and misalignment feed-forward molten pool monitoring system and penetration monitoring method Download PDFInfo
- Publication number
- CN111061231A CN111061231A CN201911204810.8A CN201911204810A CN111061231A CN 111061231 A CN111061231 A CN 111061231A CN 201911204810 A CN201911204810 A CN 201911204810A CN 111061231 A CN111061231 A CN 111061231A
- Authority
- CN
- China
- Prior art keywords
- weld
- image
- molten pool
- welding
- misalignment
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 43
- 238000012544 monitoring process Methods 0.000 title claims abstract description 34
- 230000035515 penetration Effects 0.000 title claims abstract description 30
- 238000003466 welding Methods 0.000 claims abstract description 69
- 230000000007 visual effect Effects 0.000 claims abstract description 18
- 238000004891 communication Methods 0.000 claims abstract description 15
- 238000012549 training Methods 0.000 claims description 8
- 238000012706 support-vector machine Methods 0.000 claims description 7
- 238000012216 screening Methods 0.000 claims description 6
- 238000003066 decision tree Methods 0.000 claims description 5
- 239000013598 vector Substances 0.000 claims description 3
- 238000010309 melting process Methods 0.000 claims 1
- 238000012545 processing Methods 0.000 description 14
- 230000004927 fusion Effects 0.000 description 7
- 238000001914 filtration Methods 0.000 description 3
- 230000002159 abnormal effect Effects 0.000 description 2
- 238000007781 pre-processing Methods 0.000 description 2
- 229910000838 Al alloy Inorganic materials 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000010891 electric arc Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012805 post-processing Methods 0.000 description 1
- 238000003908 quality control method Methods 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B19/00—Programme-control systems
- G05B19/02—Programme-control systems electric
- G05B19/418—Total factory control, i.e. centrally controlling a plurality of machines, e.g. direct or distributed numerical control [DNC], flexible manufacturing systems [FMS], integrated manufacturing systems [IMS] or computer integrated manufacturing [CIM]
- G05B19/41885—Total factory control, i.e. centrally controlling a plurality of machines, e.g. direct or distributed numerical control [DNC], flexible manufacturing systems [FMS], integrated manufacturing systems [IMS] or computer integrated manufacturing [CIM] characterised by modeling, simulation of the manufacturing system
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B2219/00—Program-control systems
- G05B2219/30—Nc systems
- G05B2219/32—Operator till task planning
- G05B2219/32339—Object oriented modeling, design, analysis, implementation, simulation language
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/02—Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]
Landscapes
- Engineering & Computer Science (AREA)
- Manufacturing & Machinery (AREA)
- General Engineering & Computer Science (AREA)
- Quality & Reliability (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to a weld assembly gap and staggered edge feed-forward molten pool monitoring system and a penetration monitoring method. Compared with the prior art, the invention adds the visual sensing unit and the data acquisition and control unit on the basis of the conventional communication modules of the welding machine and the robot to realize the acquisition of welding process parameters, and uses the groove profile scanner to acquire and measure aiming at random assembly conditions to realize the accurate monitoring of the welding penetration state.
Description
Technical Field
The invention relates to the technical field of welding quality control, in particular to a weld pool monitoring system and a penetration monitoring method for welding assembly gap and misalignment feed-forward.
Background
Welding is a nonlinear, time-varying and strongly-coupled complex process, and for accurate monitoring of the welding process state, multiple sensors are required to acquire information, and multiple information fusion is performed to predict and evaluate the welding penetration state. Currently used sensing modes include weld pool visual sensing, arc sound sensing, electric signal sensing such as welding current and arc voltage, etc., for example, patent CN101224519B "arc welding robot welding monitoring system based on visual sensing". However, the existing methods have the following disadvantages: firstly, the method mainly aims at the passive vision of the acquisition mode of the molten pool image, which is easily interfered by arc light, and increases the difficulty and the accuracy of the molten pool image processing; second, weld assembly errors and thermal distortion during welding can affect the penetration state. At present, no method for monitoring the welding quality by combining the information of assembly gaps and misalignment exists.
Disclosure of Invention
The invention aims to overcome the defects of the prior art and provide a weld pool monitoring system and a penetration monitoring method for weld assembly gap and misalignment feed-forward.
The purpose of the invention can be realized by the following technical scheme:
the utility model provides a molten bath monitored control system of welding fit-up gap and wrong limit feedforward, includes industrial computer, welding unit, vision sensing unit, data acquisition and the control unit, vision sensing unit includes groove profile scanner and industrial camera, industrial camera has connect laser auxiliary light source, groove profile scanner with data acquisition is connected with the control unit, industrial camera with the industrial computer is connected, welding unit, industrial computer all with data acquisition is connected with the control unit respectively.
Preferably, the welding unit comprises a welding machine, a wire feeder, a robot and a communication module, wherein the robot and the welding machine are respectively connected with the communication module, and the wire feeder is connected with the welding machine.
Preferably, the data acquisition and control unit comprises a PLC module and a high-speed acquisition card, the PLC module is respectively connected with the communication module and the industrial personal computer, and the high-speed acquisition card is respectively connected with the communication module, the groove profile scanner and the industrial personal computer.
Preferably, the laser auxiliary light source and the industrial camera are adjusted to a certain angle, so that the laser can press the arc light and a clear molten pool image enters the industrial camera.
Preferably, the groove profile scanner is fixed right ahead in the welding direction and used for collecting gap and misalignment information of the weldment, and collected data are transmitted into the industrial personal computer through the high-speed collection card to be processed and stored.
A penetration monitoring method of the weld pool monitoring system adopting the welding assembly gap and the misalignment feed-forward comprises the following steps:
step S1: collecting a molten pool image through a visual sensing unit, screening the molten pool image based on gray value distribution, and reserving a clear image with weak arc light interference;
step S2: carrying out molten pool interesting region positioning and molten pool contour identification based on a visual attention mechanism on the obtained clear image, and extracting molten pool characteristics;
step S3: and combining the welding assembly gap and the misalignment information and the molten pool characteristics obtained in the step S2, and obtaining the current penetration state through the established welding penetration prediction model.
Preferably, the process of screening the molten pool image based on the gray value distribution includes:
setting a gray threshold, traversing each pixel point of a molten pool image to obtain a gray value, counting the pixel points of which the gray values exceed the gray threshold, judging the molten pool image of which the total counting value is greater than or equal to a set value as an unqualified image with overhigh arc intensity, discarding the unqualified image, and judging the molten pool image smaller than the set value as a clear image with weak arc interference and keeping the clear image.
Preferably, the process of positioning the region of interest of the molten pool based on the visual attention mechanism comprises the following steps:
(1) calculating the gradient size and direction of each pixel of the image;
(2) dividing the image into small cell areas, and counting a directional gradient histogram of each cell area;
(3) combining a plurality of cell areas into blocks, counting the direction gradient histogram information of each block, and decomposing an image into a plurality of characteristic vectors;
(4) inputting a linear support vector machine classifier to determine whether a region in a window is a target object;
(5) and identifying the region of interest of the molten pool through a trained support vector machine classifier.
Preferably, the molten pool profile identification process adopts a gradient boosting decision tree algorithm, and specifically includes:
initializing an average shape from training samples
Shape after next iteration through the iterator
The method comprises the following steps:
wherein, I represents an input image,
representing the coordinates of the contour points after the t-th iteration, gammatRepresenting an iterator based on the image and contour point coordinates; after the initial position passes through all the trees obtained by training, the final contour key point position is obtained by adding the average shape and the shape change values of all the passing leaf nodes, polynomial fitting of a molten pool contour curve is carried out on the extracted position coordinates of the molten pool contour key point, and geometric information of the molten pool contour is obtained according to the fitted result.
Preferably, the weld penetration prediction model is based on a LightGBM model, input parameters of which include welding current, voltage, wire feed speed, extracted weld puddle characteristics, and weld assembly gap and misalignment information, and output parameters of which include weld back face weld width.
Compared with the prior art, the invention has the following advantages:
1. the welding fusion penetration monitoring system is built in a modularized mode, a visual sensing unit and a data acquisition and control unit are added on the basis of a conventional communication module of a welding machine and a robot, welding process parameters are acquired, a groove profile scanner is used for acquiring and measuring aiming at random assembly conditions, acquired high-dynamic molten pool images are subjected to design preprocessing and a processing algorithm based on visual attention, the welding fusion penetration monitoring system has self-adaptability and high real-time performance, multi-parameters including molten pool image characteristics, assembly information and a welding process are fused to monitor and evaluate a welding fusion penetration state, and accurate monitoring of the welding fusion penetration state is achieved.
2. The auxiliary light source is used for collecting the image of the molten pool, and the arc light interference is filtered to the maximum extent, so that the collected image is stable and clear.
3. The method and the device have the advantages that the high-dynamic molten pool images are screened, and the high-dynamic molten pool images are screened in an algorithm level based on the image gray values, so that the possibility of abnormal values in the post-processing process is reduced, and the subsequent processing efficiency is improved.
4. And aiming at the unstable molten pool shape of the reserved molten pool image, positioning a molten pool area and extracting molten pool characteristics by adopting an algorithm of a visual attention mechanism, so that when the relative position of a camera and the molten pool is changed, the interested area can still be accurately positioned, the molten pool profile can be accurately extracted, the speed of an image processing flow is accelerated, the time consumption of the image processing flow is within 30 milliseconds, and the requirement of real-time performance is met.
Drawings
FIG. 1 is a block diagram of a molten bath monitoring system according to the present invention;
FIG. 2 is a flow chart of a penetration monitoring method of the present invention.
The labels in the figure are: 1 is an industrial personal computer; 2 is a high-speed acquisition card; 3 is a PLC module; 4, a welding machine; 5, a wire feeder; 6 is a communication module; 7 is a robot; a groove profile scanner 8; 9 is an industrial camera; and 10 is a laser auxiliary light source.
Detailed Description
The invention is described in detail below with reference to the figures and specific embodiments. The present embodiment is implemented on the premise of the technical solution of the present invention, and a detailed implementation manner and a specific operation process are given, but the scope of the present invention is not limited to the following embodiments.
Examples
As shown in FIG. 1, the application provides a weld pool monitoring system for welding assembly gap and misalignment feed-forward, which comprises an industrial personal computer 1, a welding unit, a visual sensing unit and a data acquisition and control unit. In this example, the system was applied to aluminum alloy GTAW welding under random assembly conditions.
The welding unit comprises a welding machine 4, a wire feeder 5, a robot 7 and a communication module 6, wherein the robot 7 and the welding machine 4 are respectively connected with the communication module 6, and arc striking signals controlled by the robot 7 and current conduction (WCR) signals fed back by the welding machine 4 are transmitted. The wire feeder 5 is connected to the welder 4.
The visual sensing unit comprises a groove profile scanner 8 and an industrial camera 9, the industrial camera 9 is connected with a laser auxiliary light source 10, the groove profile scanner 8 is connected with the data acquisition and control unit, the industrial camera 9 is connected with the industrial personal computer 1, and the welding unit and the industrial personal computer 1 are respectively connected with the data acquisition and control unit. In this embodiment, the industrial camera 9 is a Charge Coupled Device (CCD) camera. The laser auxiliary light source 10 and the industrial camera 9 are adjusted to a proper angle, so that the laser can press the arc light, a clear molten pool image enters the industrial camera 9, and the triggering time of the laser and the image acquisition time of the industrial camera 9 are synchronized through the synchronization device. The vision sensing unit is equipped with a set of narrow band filters (905nm) and a dimmer, and can filter most of arc light and is close to the central wavelength of the laser. The industrial camera 9 and the welding machine 4 are fixed through a clamp, and the clamp can adjust the image acquisition angle. The collected images are directly transmitted to the industrial personal computer 1 through the USB for processing and storage. The groove profile scanner 8 is fixed right ahead in the welding direction and used for collecting gap and misalignment information of the weldment, and collected data are transmitted into the industrial personal computer 1 through the high-speed collection card 2 to be processed and stored. Aiming at random welding assembly conditions, a linear structured light groove profile scanner 8 fixed at the front end of a welding gun is used for real-time acquisition and measurement, and gap and misalignment information is used as input parameters of a subsequent prediction model.
The data acquisition and control unit comprises a PLC module 3 and a high-speed acquisition card 2, the PLC module 3 is respectively connected with the communication module 6 and the industrial personal computer 1, and the high-speed acquisition card 2 is respectively connected with the communication module 6, the groove profile scanner 8 and the industrial personal computer 1. The welding machine 4 sends a current conduction (WCR) signal to the PLC module 3 when the arc striking is successful, and the PLC module 3 outputs a signal for starting to collect to the industrial personal computer 1; and after the welding is finished, the WCR signal disappears, the collection is stopped, and the automatic control of the collection process is realized.
As shown in fig. 2, the present application proposes a penetration monitoring method of a weld pool monitoring system using the above weld assembly gap and misalignment feed-forward, comprising:
step S1: collecting a molten pool image through a visual sensing unit, screening the molten pool image based on gray value distribution, and reserving a clear image with weak arc light interference;
step S2: carrying out molten pool interesting region positioning and molten pool contour identification based on a visual attention mechanism on the obtained clear image, and extracting molten pool characteristics;
step S3: and combining the welding assembly gap and the misalignment information and the molten pool characteristics obtained in the step S2, and obtaining the current penetration state through the established welding penetration prediction model.
Extreme conditions during welding, such as strong arc and surface reflections, can greatly affect the quality of the weld pool image. Image processing of low quality images may cause failure of a preset image processing algorithm, even generation of abnormal values, and waste of processing time, affecting real-time performance.
In order to solve the above problem, step S1 proposes a high dynamic molten pool image preprocessing method, i.e. a molten pool image sequence filtering method based on gray-scale values. The process of screening the molten pool image based on the gray value distribution comprises the following steps: setting a gray threshold, traversing each pixel point of a molten pool image to obtain a gray value, counting the pixel points of which the gray values exceed the gray threshold, judging the molten pool image of which the total counting value is greater than or equal to a set value as an unqualified image with overhigh arc intensity, discarding the unqualified image, and judging the molten pool image smaller than the set value as a clear image with weak arc interference and keeping the clear image. In this embodiment, the acquired molten pool images are a group of continuous images with a frame rate of 20Hz and an interval of 0.05s, and 5 low-quality images with strong arc interference are separated between two clear high-quality images. 4 high-quality molten pool images are obtained through the filtration of the molten pool image filtration method, the images are reserved for subsequent processing, and low-quality images are directly discarded.
The welding pool image is characterized in that the area of the welding pool is small, most of the surrounding area has no effective information, and therefore the area of interest needs to be positioned first, so that the image pixels needing to be processed are reduced, and the speed and the accuracy of image processing are improved. The existing positioning of the region of interest of the molten pool depends on manually setting a fixed window after a welding platform is built, so that the relative position of a camera and a welding gun is required to be ensured not to be changed all the time, otherwise, the window of interest needs to be reset.
In order to solve the above problem, step S2 proposes a molten pool region-of-interest positioning process based on a visual attention mechanism, which includes:
(1) calculating the gradient size and direction of each pixel of the image;
(2) dividing the image into small cell areas, and counting a Histogram Of oriented gradients (HOG) Of each cell area;
(3) combining a plurality of cell areas into blocks, counting HOG information of each block, and decomposing an image into a plurality of feature vectors at the moment;
(4) inputting a linear Support Vector Machine (SVM) classifier to determine whether a region in a window is a target object;
(5) and identifying the region of interest of the molten pool through a trained support vector machine classifier.
And continuously extracting the contour of the molten pool aiming at the extracted interesting area of the molten pool. The current widely applied passive visual molten pool image acquisition technology is to utilize the arc light of the electric arc as the only light source to illuminate the molten pool, and the intensity of the arc light is influenced by a plurality of factors: current level, bath surface state. The existing molten pool contour extraction algorithm depends on fixed arc intensity, and the geometric information of the molten pool is difficult to accurately extract under the conditions of strong arc interference and arc intensity change.
The process of molten pool contour identification in the application adopts a gradient lifting decision tree algorithm, and can stably and quickly acquire molten pool contour information under the condition of strong arc light and arc light intensity change. The method comprises the steps that a Gradient Boosting Decision Tree (GBDT) is the combination of Gradient Boosting and a Boosting Tree, during model training, residual values of a last Tree are fitted, iteration is continuously carried out to eliminate residual errors, and results of all trees are accumulated to obtain a final result. The process of identifying the molten pool profile specifically comprises the following steps:
initializing an average shape from training samples
Shape after next iteration through the iterator
The method comprises the following steps:
wherein, I represents an input image,
representing the coordinates of the contour points after the t-th iteration, gammatRepresenting an iterator based on the image and contour point coordinates; after the initial position passes through all the trees obtained by training, the final contour key point position is obtained by adding the average shape and the shape change values of all the passing leaf nodes, polynomial fitting of a molten pool contour curve is carried out on the extracted position coordinates of the molten pool contour key point, and geometric information of the molten pool contour is obtained according to the fitted result. The processing speed of the image measured by the method in actual test is less than 30 milliseconds, the requirement of real-time processing is met, and compared with manual labeling, the processing error is extremely small.
In this embodiment, the welding penetration prediction model is based on a LightGBM model, which is an efficient implementation of a Gradient Boosting Decision Tree (GBDT), and has the characteristics of faster training speed, smaller occupied memory, and high accuracy. The model combines the welding process information obtained in the steps, the input parameters of the model comprise welding current, voltage, wire feeding speed, extracted molten pool characteristics and welding assembly gap and misalignment information, and the output parameters of the model comprise welding back fusion width, so that the welding fusion penetration state is evaluated.
Claims (10)
1. The utility model provides a molten bath monitored control system of welding fit-up gap and wrong limit feedforward, its characterized in that, includes industrial computer, welding unit, vision sensing unit, data acquisition and the control unit, vision sensing unit includes groove profile scanner and industrial camera, industrial camera has connect laser auxiliary light source, groove profile scanner with data acquisition is connected with the control unit, industrial camera with the industrial computer is connected, welding unit, industrial computer all with data acquisition is connected with the control unit respectively.
2. The weld puddle monitoring system for weld setup gap and misalignment feed-forward of claim 1 wherein the welding unit includes a welder, a wire feeder, a robot, a communication module, the robot and the welder each being connected to the communication module, the wire feeder being connected to the welder.
3. The weld assembly gap and misalignment feed-forward weld pool monitoring system according to claim 2, wherein the data acquisition and control unit comprises a PLC module and a high-speed acquisition card, the PLC module is respectively connected with the communication module and the industrial personal computer, and the high-speed acquisition card is respectively connected with the communication module, the groove profile scanner and the industrial personal computer.
4. A weld puddle monitoring system for weld setup gap and misalignment feed-forward of claim 1 where the laser assist light source is angled from the industrial camera so that the laser can press the arc light to get a clear puddle image into the industrial camera.
5. The weld pool monitoring system for welding assembly gap and misalignment feed-forward according to claim 1, wherein the groove profile scanner is fixed right ahead of the welding direction and used for collecting gap and misalignment information of a weldment, and collected data are transmitted to an industrial personal computer through a high-speed collection card to be processed and stored.
6. A penetration monitoring method of a weld pool monitoring system adopting any one of the welding assembly gap and misalignment feed-forward of 1-5 is characterized by comprising the following steps:
step S1: collecting a molten pool image through a visual sensing unit, screening the molten pool image based on gray value distribution, and reserving a clear image with weak arc light interference;
step S2: carrying out molten pool interesting region positioning and molten pool contour identification based on a visual attention mechanism on the obtained clear image, and extracting molten pool characteristics;
step S3: and combining the welding assembly gap and the misalignment information and the molten pool characteristics obtained in the step S2, and obtaining the current penetration state through the established welding penetration prediction model.
7. The weld assembly gap and misalignment feed-forward weld puddle penetration monitoring method of claim 6, wherein said process of gray value distribution-based screening of the weld puddle image comprises:
setting a gray threshold, traversing each pixel point of a molten pool image to obtain a gray value, counting the pixel points of which the gray values exceed the gray threshold, judging the molten pool image of which the total counting value is greater than or equal to a set value as an unqualified image with overhigh arc intensity, discarding the unqualified image, and judging the molten pool image smaller than the set value as a clear image with weak arc interference and keeping the clear image.
8. The weld assembly gap and stagger feed-forward weld puddle penetration monitoring method of claim 6, wherein the visual attention mechanism-based weld puddle region-of-interest positioning process comprises:
(1) calculating the gradient size and direction of each pixel of the image;
(2) dividing the image into small cell areas, and counting a directional gradient histogram of each cell area;
(3) combining a plurality of cell areas into blocks, counting the direction gradient histogram information of each block, and decomposing an image into a plurality of characteristic vectors;
(4) inputting a linear support vector machine classifier to determine whether a region in a window is a target object;
(5) and identifying the region of interest of the molten pool through a trained support vector machine classifier.
9. The weld assembly gap and stagger feed-forward weld puddle penetration monitoring method according to claim 6, wherein the weld puddle profile identification process employs a gradient boosting decision tree algorithm, specifically comprising:
initializing an average shape from training samples
Shape after next iteration through the iterator
The method comprises the following steps:
wherein, I represents an input image,
representing the coordinates of the contour points after the t-th iteration, gammatRepresenting an iterator based on the image and contour point coordinates; after all the trees obtained by training are placed in the initial position, the final contour key point position is obtained by adding the average shape and the shape change values of all the passed leaf nodes, and the melting process is carried out on the extracted position coordinates of the contour key point of the molten poolAnd performing polynomial fitting on the pool profile curve to obtain the geometric information of the pool profile according to the fitting result.
10. The weld setup gap and misalignment feed-forward weld puddle penetration monitoring method of claim 6, wherein the weld penetration prediction model is based on a LightGBM model with input parameters including weld current, voltage, wire feed speed, extracted puddle characteristics, and weld setup gap and misalignment information and output parameters including weld back weld width.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911204810.8A CN111061231B (en) | 2019-11-29 | 2019-11-29 | Weld assembly gap and misalignment feed-forward molten pool monitoring system and penetration monitoring method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911204810.8A CN111061231B (en) | 2019-11-29 | 2019-11-29 | Weld assembly gap and misalignment feed-forward molten pool monitoring system and penetration monitoring method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111061231A true CN111061231A (en) | 2020-04-24 |
| CN111061231B CN111061231B (en) | 2021-11-30 |
Family
ID=70299395
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911204810.8A Active CN111061231B (en) | 2019-11-29 | 2019-11-29 | Weld assembly gap and misalignment feed-forward molten pool monitoring system and penetration monitoring method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111061231B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112453751A (en) * | 2020-11-27 | 2021-03-09 | 上海交通大学 | Pipeline all-position welding back surface online monitoring method based on visual sensing |
| CN114178697A (en) * | 2021-12-20 | 2022-03-15 | 上海交通大学 | Pulse laser micro spot welding system and welding method for motor stator lamination |
| CN116021119A (en) * | 2023-03-29 | 2023-04-28 | 中建安装集团有限公司 | Petrochemical process pipeline robot welding assembly misalignment detection system and method |
| CN116597391A (en) * | 2023-07-18 | 2023-08-15 | 武汉纺织大学 | Synchronous on-line monitoring method for weld surface morphology and penetration state |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US4532408A (en) * | 1982-10-22 | 1985-07-30 | The Ohio State University | Apparatus and method for viewing molten pools in arc welding |
| JPS6117366A (en) * | 1984-07-03 | 1986-01-25 | Mitsubishi Heavy Ind Ltd | Method and device for automatic welding |
| CN101127083A (en) * | 2007-09-06 | 2008-02-20 | 江苏大学 | Welding image identification method |
| CN102139398A (en) * | 2011-04-15 | 2011-08-03 | 中铁山桥集团有限公司 | Orthotropic plate U-shaped rib angle weld joint double-side welding technique |
| CN103464869A (en) * | 2013-08-26 | 2013-12-25 | 江苏科技大学 | Rotating arc narrow gap MAG (metal active gas) welding seam offset identification device and method based on visual sensing |
| CN103521890A (en) * | 2013-10-12 | 2014-01-22 | 王晓宇 | Device and method for double-faced double-arc vertical welding near-infrared vision sensing and penetration control |
| CN105458490A (en) * | 2015-12-30 | 2016-04-06 | 哈尔滨工业大学 | Real-time monitoring system and monitoring method for judging welding type of laser deep penetration welding by using high-speed camera |
| CN105741306A (en) * | 2016-03-03 | 2016-07-06 | 上海交通大学 | Penetration state determination method based on small hole characteristic on back side |
| CN106041258A (en) * | 2016-08-17 | 2016-10-26 | 江苏北人机器人系统股份有限公司 | Intelligent robot welding system |
| CN106624266A (en) * | 2016-12-31 | 2017-05-10 | 东莞职业技术学院 | Weld seam deviation and penetration state monitoring method for automobile welding |
| CN108152291A (en) * | 2017-11-30 | 2018-06-12 | 清华大学 | End seam welding incomplete fusion shortcoming real-time detection method based on weld pool dynamics feature |
-
2019
- 2019-11-29 CN CN201911204810.8A patent/CN111061231B/en active Active
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US4532408A (en) * | 1982-10-22 | 1985-07-30 | The Ohio State University | Apparatus and method for viewing molten pools in arc welding |
| JPS6117366A (en) * | 1984-07-03 | 1986-01-25 | Mitsubishi Heavy Ind Ltd | Method and device for automatic welding |
| CN101127083A (en) * | 2007-09-06 | 2008-02-20 | 江苏大学 | Welding image identification method |
| CN102139398A (en) * | 2011-04-15 | 2011-08-03 | 中铁山桥集团有限公司 | Orthotropic plate U-shaped rib angle weld joint double-side welding technique |
| CN103464869A (en) * | 2013-08-26 | 2013-12-25 | 江苏科技大学 | Rotating arc narrow gap MAG (metal active gas) welding seam offset identification device and method based on visual sensing |
| CN103521890A (en) * | 2013-10-12 | 2014-01-22 | 王晓宇 | Device and method for double-faced double-arc vertical welding near-infrared vision sensing and penetration control |
| CN105458490A (en) * | 2015-12-30 | 2016-04-06 | 哈尔滨工业大学 | Real-time monitoring system and monitoring method for judging welding type of laser deep penetration welding by using high-speed camera |
| CN105741306A (en) * | 2016-03-03 | 2016-07-06 | 上海交通大学 | Penetration state determination method based on small hole characteristic on back side |
| CN106041258A (en) * | 2016-08-17 | 2016-10-26 | 江苏北人机器人系统股份有限公司 | Intelligent robot welding system |
| CN106624266A (en) * | 2016-12-31 | 2017-05-10 | 东莞职业技术学院 | Weld seam deviation and penetration state monitoring method for automobile welding |
| CN108152291A (en) * | 2017-11-30 | 2018-06-12 | 清华大学 | End seam welding incomplete fusion shortcoming real-time detection method based on weld pool dynamics feature |
Non-Patent Citations (2)
| Title |
|---|
| 刘新峰: "基于正面熔池图像和深度学习算法的PAW穿孔/熔透状态预测", 《中国博士学位论文全文数据库 工程科技I辑》 * |
| 褚慧慧: "基于视觉的焊缝质量检测技术研究", 《中国博士学位论文全文数据库 信息科技辑》 * |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112453751A (en) * | 2020-11-27 | 2021-03-09 | 上海交通大学 | Pipeline all-position welding back surface online monitoring method based on visual sensing |
| CN112453751B (en) * | 2020-11-27 | 2022-07-15 | 上海交通大学 | Pipeline all-position welding back surface online monitoring method based on visual sensing |
| CN114178697A (en) * | 2021-12-20 | 2022-03-15 | 上海交通大学 | Pulse laser micro spot welding system and welding method for motor stator lamination |
| CN114178697B (en) * | 2021-12-20 | 2022-12-13 | 上海交通大学 | Pulse laser micro spot welding system and welding method for motor stator lamination |
| CN116021119A (en) * | 2023-03-29 | 2023-04-28 | 中建安装集团有限公司 | Petrochemical process pipeline robot welding assembly misalignment detection system and method |
| CN116597391A (en) * | 2023-07-18 | 2023-08-15 | 武汉纺织大学 | Synchronous on-line monitoring method for weld surface morphology and penetration state |
| CN116597391B (en) * | 2023-07-18 | 2023-09-19 | 武汉纺织大学 | Synchronous on-line monitoring method for weld surface morphology and penetration state |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111061231B (en) | 2021-11-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111061231B (en) | Weld assembly gap and misalignment feed-forward molten pool monitoring system and penetration monitoring method | |
| CN109886298B (en) | Weld quality detection method based on convolutional neural network | |
| CN108320280B (en) | Molten pool image anomaly detection method based on visual definition and contour extraction | |
| CN103464869B (en) | The rotating the arc narrow gap MAG welding line deviation recognition device of view-based access control model sensing and method | |
| CN116423005B (en) | Tin soldering process optimization method and system for improving welding precision | |
| CN112183313A (en) | SlowFast-based power operation field action identification method | |
| CN106778439B (en) | A kind of lithium battery batch scan code system and method based on image procossing | |
| CN110648319B (en) | Equipment image acquisition and diagnosis system and method based on double cameras | |
| CN104020222A (en) | Welding quality detecting system of full-automatic ultrasonic aluminium wire press welder | |
| CN117589129B (en) | Gradient measurement method and system for vehicle-mounted flexible contact net positioner | |
| CN101169826A (en) | Row roller forming welding tube quality on-line monitoring and controlling system based on machine vision | |
| CN116493735B (en) | Real-time tracking method for motion splash in Wanwave-level ultra-high power laser welding process | |
| CN113762240A (en) | Method and system for predicting geometrical characteristics of cladding layer based on deep learning | |
| CN112215825A (en) | Quality analysis method and system based on machine vision in new energy battery manufacturing | |
| CN116989881A (en) | Contact line vibration measurement system and method based on line laser target identification | |
| CN102663781A (en) | Sub-pixel level welding center extraction method based on visual sense | |
| CN110837760B (en) | Target detection method, training method and device for target detection | |
| CN114723668A (en) | Cathode copper quality detection method and system based on area-array camera | |
| CN113077423A (en) | Laser selective melting pool image analysis system based on convolutional neural network | |
| EP3885069A1 (en) | Quality control of a laser machining process by means of machine learning | |
| CN115255565A (en) | Global pattern recognition based narrow gap welding notch edge vision sensing detection method and application | |
| KR20230101459A (en) | Apparatus for tracking and analyzing movements of welding weaving | |
| CN111790910B (en) | Method for feeding back and adjusting defects of fused forming part of laser powder bed | |
| CN111811437A (en) | Band saw blade welding quality detection method and device | |
| CN112132807B (en) | Weld joint region extraction method and device based on color similarity segmentation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |