WO2016107408A1 - 窄间隙焊缝偏差的红外视觉传感检测方法及装置 - Google Patents

窄间隙焊缝偏差的红外视觉传感检测方法及装置 Download PDF

Info

Publication number
WO2016107408A1
WO2016107408A1 PCT/CN2015/097552 CN2015097552W WO2016107408A1 WO 2016107408 A1 WO2016107408 A1 WO 2016107408A1 CN 2015097552 W CN2015097552 W CN 2015097552W WO 2016107408 A1 WO2016107408 A1 WO 2016107408A1
Authority
WO
WIPO (PCT)
Prior art keywords
image
groove
arc
wire
welding
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.)
Ceased
Application number
PCT/CN2015/097552
Other languages
English (en)
French (fr)
Inventor
王加友
朱杰
张才
杨茂森
苏娜
黎文航
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Jiangsu University of Science and Technology
Original Assignee
Jiangsu University of Science and Technology
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 Jiangsu University of Science and Technology filed Critical Jiangsu University of Science and Technology
Priority to US15/541,021 priority Critical patent/US9889528B2/en
Publication of WO2016107408A1 publication Critical patent/WO2016107408A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B23—MACHINE TOOLS; METAL-WORKING NOT OTHERWISE PROVIDED FOR
    • B23K—SOLDERING OR UNSOLDERING; WELDING; CLADDING OR PLATING BY SOLDERING OR WELDING; CUTTING BY APPLYING HEAT LOCALLY, e.g. FLAME CUTTING; WORKING BY LASER BEAM
    • B23K9/00—Arc welding or cutting
    • B23K9/16—Arc welding or cutting making use of shielding gas
    • B23K9/173—Arc welding or cutting making use of shielding gas and of a consumable electrode
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B23—MACHINE TOOLS; METAL-WORKING NOT OTHERWISE PROVIDED FOR
    • B23K—SOLDERING OR UNSOLDERING; WELDING; CLADDING OR PLATING BY SOLDERING OR WELDING; CUTTING BY APPLYING HEAT LOCALLY, e.g. FLAME CUTTING; WORKING BY LASER BEAM
    • B23K31/00—Processes relevant to this subclass, specially adapted for particular articles or purposes, but not covered by any single one of main groups B23K1/00 - B23K28/00
    • B23K31/12—Processes relevant to this subclass, specially adapted for particular articles or purposes, but not covered by any single one of main groups B23K1/00 - B23K28/00 relating to investigating the properties, e.g. the weldability, of materials
    • B23K31/125—Weld quality monitoring
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B23—MACHINE TOOLS; METAL-WORKING NOT OTHERWISE PROVIDED FOR
    • B23K—SOLDERING OR UNSOLDERING; WELDING; CLADDING OR PLATING BY SOLDERING OR WELDING; CUTTING BY APPLYING HEAT LOCALLY, e.g. FLAME CUTTING; WORKING BY LASER BEAM
    • B23K9/00—Arc welding or cutting
    • B23K9/12—Automatic feeding or moving of electrodes or work for spot or seam welding or cutting
    • B23K9/127—Means for tracking lines during arc welding or cutting
    • B23K9/1272—Geometry oriented, e.g. beam optical trading
    • B23K9/1274—Using non-contact, optical means, e.g. laser means
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B23—MACHINE TOOLS; METAL-WORKING NOT OTHERWISE PROVIDED FOR
    • B23K—SOLDERING OR UNSOLDERING; WELDING; CLADDING OR PLATING BY SOLDERING OR WELDING; CUTTING BY APPLYING HEAT LOCALLY, e.g. FLAME CUTTING; WORKING BY LASER BEAM
    • B23K9/00—Arc welding or cutting
    • B23K9/02—Seam welding; Backing means; Inserts
    • B23K9/0213—Narrow gap welding
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B23—MACHINE TOOLS; METAL-WORKING NOT OTHERWISE PROVIDED FOR
    • B23K—SOLDERING OR UNSOLDERING; WELDING; CLADDING OR PLATING BY SOLDERING OR WELDING; CUTTING BY APPLYING HEAT LOCALLY, e.g. FLAME CUTTING; WORKING BY LASER BEAM
    • B23K9/00—Arc welding or cutting
    • B23K9/02—Seam welding; Backing means; Inserts
    • B23K9/0216—Seam profiling, e.g. weaving, multilayer
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B23—MACHINE TOOLS; METAL-WORKING NOT OTHERWISE PROVIDED FOR
    • B23K—SOLDERING OR UNSOLDERING; WELDING; CLADDING OR PLATING BY SOLDERING OR WELDING; CUTTING BY APPLYING HEAT LOCALLY, e.g. FLAME CUTTING; WORKING BY LASER BEAM
    • B23K9/00—Arc welding or cutting
    • B23K9/02—Seam welding; Backing means; Inserts
    • B23K9/022—Welding by making use of electrode vibrations
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B23—MACHINE TOOLS; METAL-WORKING NOT OTHERWISE PROVIDED FOR
    • B23K—SOLDERING OR UNSOLDERING; WELDING; CLADDING OR PLATING BY SOLDERING OR WELDING; CUTTING BY APPLYING HEAT LOCALLY, e.g. FLAME CUTTING; WORKING BY LASER BEAM
    • B23K9/00—Arc welding or cutting
    • B23K9/09—Arrangements or circuits for arc welding with pulsed current or voltage
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B23—MACHINE TOOLS; METAL-WORKING NOT OTHERWISE PROVIDED FOR
    • B23K—SOLDERING OR UNSOLDERING; WELDING; CLADDING OR PLATING BY SOLDERING OR WELDING; CUTTING BY APPLYING HEAT LOCALLY, e.g. FLAME CUTTING; WORKING BY LASER BEAM
    • B23K9/00—Arc welding or cutting
    • B23K9/095—Monitoring or automatic control of welding parameters
    • B23K9/0956—Monitoring or automatic control of welding parameters using sensing means, e.g. optical
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01B—MEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
    • G01B11/00—Measuring arrangements characterised by the use of optical techniques
    • G01B11/02—Measuring arrangements characterised by the use of optical techniques for measuring length, width or thickness
    • G01B11/022—Measuring arrangements characterised by the use of optical techniques for measuring length, width or thickness by means of tv-camera scanning
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01B—MEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
    • G01B11/00—Measuring arrangements characterised by the use of optical techniques
    • G01B11/14—Measuring arrangements characterised by the use of optical techniques for measuring distance or clearance between spaced objects or spaced apertures
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G06T7/0004—Industrial image inspection
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/13—Edge detection
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/136—Segmentation; Edge detection involving thresholding
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/155—Segmentation; Edge detection involving morphological operators
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/70—Determining position or orientation of objects or cameras
    • G06T7/73—Determining position or orientation of objects or cameras using feature-based methods
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R19/00—Arrangements for measuring currents or voltages or for indicating presence or sign thereof
    • G01R19/0092—Measuring current only
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10048—Infrared image
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20024—Filtering details
    • G06T2207/20032—Median filtering
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20036—Morphological image processing
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30108—Industrial image inspection
    • G06T2207/30136—Metal
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30108—Industrial image inspection
    • G06T2207/30152—Solder

Definitions

  • the invention relates to the technical field of welding, in particular to an infrared vision sensing method and device for real-time detection of a deviation of a narrow gap weld of a rocking (or oscillating) or rotating arc.
  • the Chinese patent No. ZL201210325926.9 the patent document entitled “Narrow gap welding monitoring based on infrared vision sensing and weld deviation detection method", for the application of the narrow arc gap MAG welding of the arc, the welding torch is obtained in advance by the infrared camera.
  • the infrared camera captures the image of the narrow gap edge edge away from the arc side in real time during the welding process, the position information of the unilateral edge of the groove is extracted by image processing, and the same side slope is acquired in advance.
  • the weld boundary deviation is obtained by comparing the mouth boundaries.
  • the disadvantage is that the one-side edge position detection of the weld deviation is only applicable to the case where the groove gap is constant, and is not suitable for the common groove gap change occasion, so the application range is narrow.
  • the Chinese patent application number is 201310375369.6, the name is “Visual sensor-based rotating arc narrow gap MAG welding seam deviation identification device and method”, for the rotating arc narrow gap MAG welding application, using a charge coupled device (CCD) infrared camera,
  • CCD charge coupled device
  • the welding image obtained by rotating the arc to the left and right wall positions of the groove is obtained by triggering the sampling method, and the weld deviation is obtained according to the arc center and the groove center position information obtained by the image processing.
  • the disadvantages are as follows: 1 During the actual narrow gap welding process, the arc is generally asymmetrical with respect to its axis, especially when the arc is closer to the side edge of the groove, the arc often climbs along the sidewall of the groove, so According to the geometrical center of the arc shape extracted by the weld deviation identification method, it is difficult to accurately reflect the actual arc rotation center, thereby affecting the accuracy of the weld deviation detection, and also affecting its engineering practicability; The central detection method is only suitable for high frequency rotating arc narrow gap welding applications, not suitable for lower frequency swing (including shaking) arc narrow gap welding applications; 3 CCD camera, small dynamic range and slow response speed , affect the weld deviation detection accuracy and real-time.
  • the present invention proposes an infrared visual transmission suitable for narrow gap weld deviation of rocking or swinging and rotating arc.
  • the sensing method and device can obtain the weld deviation information by detecting the position change of the welding wire relative to the left and right side walls of the groove when the arc moves to the side walls of the groove, and has the advantages of high detection precision, strong environmental adaptability, practicability, etc. advantage.
  • the infrared visual sensing device for narrow gap weld deviation proposed by the invention comprises a narrow gap welding torch, an arc current sensor, a computer image processing system and an infrared camera system; one end of the bent conductive rod in the narrow gap welding torch is driven by the motor
  • the feeding mechanism is connected, the other end is connected with the straight conductive nozzle, the welding wire passing through the narrow gap welding torch extends into the groove to be welded to generate a welding arc;
  • the infrared camera system comprises a digital infrared camera and an infrared filter system;
  • computer image The processing system comprises an image acquisition card, a wire position information extraction module and a weld deviation value obtaining module which are sequentially connected, and the image acquisition card is connected to the infrared camera through a video line; the arc motion of the motor drive and the feeding mechanism is to the left of the groove,
  • the nearest position signals P L and P R of the right wall are simultaneously input into the infrared camera and the image acquisition card; one end of
  • the narrow gap welding torch integrally moves with the infrared camera system at the welding speed V w to the front of the groove, and triggers the infrared camera to capture the image of the welding area at the moment according to the arc position signal P L or P R , and sends the image to the computer through the image acquisition card.
  • the wire position information extraction module receives the welding image information from the image acquisition card, and processes the image to extract the right position of the wire to the slope.
  • the current distance X 1i of the left edge of the mouth and the current distance X 2i from the left position detection point of the welding wire to the right edge of the groove, and the current distance is sent to the weld deviation value obtaining module;
  • the wire position information extraction module adjusts the image intercepting window positioning point B 1i of the left side of the groove according to the lateral position change of the left edge line of the groove.
  • the horizontal coordinate value is taken to intercept the left side image of the groove without the arc interference.
  • the wire position information extraction module adjusts the abscissa value of the image capturing point B 2i of the image capturing window on the right side of the groove according to the lateral position change of the right edge line of the groove, and intercepts the image of the right side of the groove without arc interference.
  • the ordinate values of the L 2i and L 2(i+1) detection points on the left edge line of the groove are the same as the ordinate values of the L 3i detection point on the right edge line of the groove; in the right position of the wire axis of the wire
  • the L 1i and L 1(i+1) detection point ordinate values are the same as the L 4i detection point ordinate values on the left position wire axis; the wire position detection point to the global image left boundary distance L 1i , L 4i or
  • the final detection value of L 1(i+1) is the median or average value of the detected values of m (m ⁇ 1) different positions in the left and right intercepting windows of the welding wire image; the left and right edges of the groove
  • the final detected value of the distance L 2i , L 3i or L 2(i+1) to the left edge of the global image is the left and right image intercepting window, the left and right edge lines of the groove k (k ⁇ 1)
  • the median or average value of the detected values at different locations is the left and
  • the coordinate value of the highest point of the arc region is extracted by the global welding image processing, and the coordinate values of the positioning points of the left and right image intercepting windows of the welding wire are adjusted according to the change of the coordinate position of the highest point of the arc region;
  • the small window groove image intercepted by the left and right image intercepting windows of the groove and the small window wire image intercepted by the left and right image intercepting windows of the welding wire are processed;
  • the small intercepted window of the left and right image of the welding wire is taken small
  • the local adaptive threshold segmentation process is performed first, then the wire profile is extracted by the morphological corrosion of the full window image, and then the Canny edge detection algorithm is used to extract the wire skeleton, and finally the position of the wire axis of the wire is calculated.
  • the arc position signal P L or P R signal is valid, once the current sensor detects the pulsed arc first
  • the pulse welding arc base current signal i b comes, and the infrared camera is instantly triggered to collect the welding area image with the least interference from the arc arc light at the moment, and the welding image acquisition synchronized with the pulse arc base current period is realized.
  • the invention has the following beneficial effects:
  • the present invention extracts the deviation of the weld seam according to the position information of the left and right side walls of the welding wire with respect to the groove, and can effectively avoid the influence of the asymmetry of the arc shape on the detection precision;
  • the position information of the welding wire and the groove edge reflected by the invention are dynamic, and it is not necessary to establish a reference template in advance, and is suitable for the dynamic change of the groove gap;
  • the invention adjusts the position of the welding wire image and the image of the edge of the groove edge in real time according to the change of the position of the arc and the edge of the groove, thereby improving the environmental adaptability;
  • the pulse arc base value current synchronous detecting method adopted by the invention can collect the welding image with the smallest arc arc light interference, and further improve the welding deviation detection precision;
  • the invention is applicable to the narrow gap welding of rocking (or oscillating) arc with low arc motion frequency, and to the narrow arc gap welding of rotating arc with high arc motion frequency, and has wide application range.
  • Figure 1 is a schematic block diagram of a narrow gap weld deviation infrared visual sensing device
  • FIG. 2 is a schematic view showing a situation in which the weld seam is unbiased (ie, the torch is centered) in the circular arc-shaped arc welding of FIG. 1;
  • FIG. 3 is a schematic view showing the left deviation of the weld bead (ie, the right deviation of the welding torch) in the circular arc-shaped arc welding of FIG. 1;
  • Figure 4 is a schematic view showing the right deviation of the weld seam (i.e., the left deviation of the welding torch) in the circular arc-shaped arc welding of Figure 1;
  • Figure 5 is a schematic view of the weld seam without deviation (i.e., the torch is centered) in the rotating arc and the double-half-circumferential arc welding in Figure 1;
  • Figure 6 is a schematic view showing the left deviation of the weld seam (i.e., the right deviation of the welding torch) in the rotating arc and the double semi-circular rocking arc welding of Figure 1;
  • Figure 7 is a schematic view showing the right deviation of the weld seam (i.e., the left deviation of the welding torch) in the rotating arc and the double-half-circumferential arc welding in Figure 1;
  • Figure 8 is a schematic view showing the detection of weld deviation when the arc is deflected to the right side wall of the groove in Figure 1;
  • Figure 9 is a schematic diagram showing the detection of weld deviation when the arc is deflected to the left side wall of the groove in Figure 1;
  • Figure 10 is a flow chart of the weld deviation detection of Figure 1;
  • Figure 11 is a view showing an embodiment of the effect of extracting the axial position information of the wire during DC welding
  • Figure 12 is a view showing an embodiment of the effect of extracting the edge of the groove at the time of DC welding
  • Figure 13 is a view showing an embodiment of the effect of extracting the axial position information of the welding wire during pulse welding
  • Fig. 14 is a view showing an embodiment of the effect of extracting the groove edge information during pulse welding.
  • Figure 1 1 - arc; 2 - straight conductive nozzle; 3 - bending conductive rod; 4 - motor drive and feed mechanism; 5 - welding wire; 6 - wire feeder; 7 - workpiece; 8 - arc double semicircle Circumferential trajectory (or arc circular trajectory); 9—to be welded groove; 10—arc arc trajectory; 11—infrared camera; 12—infrared filter system; 13—arc current sensor; 14—welding Power; 15 - computer image processing system; 15-1 - image acquisition card; 15-2 - wire position information extraction module; 15-3 - weld deviation value seeking module.
  • V w welding speed
  • P L arc (in the direction perpendicular to the welding speed V w ) to the nearest position of the left side wall of the groove
  • P R arc (at the welding speed V w
  • a 1i O 2i A 2i arc-shaped rocking trajectory 10 of the arc with respect to the torch; O 1i — chord midpoint of the arc arc-shaped trajectory A 1i O 2i A 2i ; O 2i - the arc midpoint of the arc trajectory; O 3i - the projection point of the center of the torch at the bottom of the groove, which is also the center point of the arc relative to the circumferential trajectory 8 of the torch; O 4 O 5 - the center line of the groove; ⁇ - arc shaking Angle; A 1i - the closest position of the arc to the left side of the groove ⁇ L (in the direction perpendicular to the welding speed V w ); A 2i - the closest position of the arc to the right side of the groove ⁇ R (at the welding speed V w vertical direction); A 1i F i - the closest distance of the arc to the left side ⁇ L; A 2i E i - the closest distance of the arc to the right
  • the infrared visual sensing device for narrow gap weld deviation of the present invention is shown in FIG. 1 and mainly includes a narrow gap welding torch, an arc current sensor 13, a computer image processing system 15, an infrared camera system, and the like.
  • the narrow gap welding torch is further composed of a motor driving and feeding mechanism 4, a bending conductive rod 3, a straight conductive nozzle 2, a nozzle mechanism, etc., wherein one end of the bending conductive rod 3 is connected to the motor driving and feeding mechanism 4, and the other end
  • the direct contact nozzle 2 is connected; the motor drive and feed mechanism 4 in the narrow gap torch can be directly driven by the hollow shaft motor, or can be indirectly driven by the common motor through the transmission pair, and is preferably driven directly by the hollow shaft motor. formula.
  • An electro-optical (or electromagnetic) detecting device is mounted in the motor driving and feeding mechanism 4 to detect the arc 1 (in the direction perpendicular to the welding speed V w ) to the closest position signal PL of the left side wall and the right side wall of the groove 9 Or P R .
  • Power source by welding, narrow gap welding torch and the driving control system, before the weldability parameter set welding energy (arc current, arc voltage and the welding speed V w) 14 and arc arc rotation parameter or parameters shaking.
  • the welding wire 5 sent from the wire feeder 6 sequentially passes through the center hole of the motor driving and feeding mechanism 4, the bent conductive rod 3 and the straight conductive nozzle 2, and is sent out into the groove 9 to be welded to generate a welding arc 1;
  • One end of the power source 14 is connected to the motor driving and feeding mechanism 4, and is connected to the brush feeding mechanism of the motor driving and feeding mechanism 4.
  • the other end of the welding power source 14 is connected with the workpiece 7, so that the welding cable is not entangled. Welding feed.
  • the gas supply system provides an active or inert shielding gas to the weld zone where the arc 1 is located by means of a nozzle mechanism in the torch, but does not require a nozzle mechanism when welding with a self-shielded flux cored wire, nor does it provide a shielding gas to the weld zone.
  • the movement mode of the arc 1 relative to the welding torch is divided into two types: shaking (or oscillating) and rotating, including: reciprocating circular arc 10 shaking, unidirectional double semicircular trajectory 8 shaking (arc is off the slope)
  • the left and right side walls have the closest position A 1i and A 2i at the position), the circumferential trajectory rotation (the arc has no stop at the closest positions A 1i and A 2i from the left and right side walls of the groove); wherein the rotary arc motion
  • a common eccentric conductive nozzle type rotating arc and a conductive rod integrally rotating around the center of the torch are used.
  • the infrared camera system includes an infrared camera 11 and an infrared filter system 12, and the infrared filter system 12 is coaxially connected with the infrared camera 11, and is installed directly in front of the infrared camera 11.
  • the infrared camera 11 may be a charge coupled device (CCD) type or a complementary metal oxide semiconductor (CMOS) type, and preferably a CMOS type, and the infrared camera 11 is fixed directly in front of or behind the torch, so that the infrared camera 11 and the infrared camera 11 are to be soldered.
  • the bottom surface of the groove 9 is formed at an angle of 20 to 60°, which facilitates the infrared camera 11 to take in the image of the welding area of the groove 9 to be welded.
  • a digital infrared camera 11 with a zoom range of 18 to 45 mm and an aperture of 5.6 to 32 is selected, and an image of the welding area is obtained in an external trigger mode.
  • the infrared filter system 12 includes a narrow band filter, a neutral light reduction plate, and a protective lens.
  • the center wavelength of the narrow band filter is 700 to 1100 nm, and the transmittance of the neutral light reduction plate is 1 to 50%.
  • narrow-band filter is used to filter out optical radiation interference such as arc, smoke and splash.
  • Neutral dimming film can be used. The intensity of the radiation from the arc and the molten pool is adjusted so that a clear welded infrared image can be acquired.
  • the computer image processing system 15 mainly includes an image capturing card 15-1, a wire position information extracting module 15-2, and a weld deviation value obtaining module 15-3 which are sequentially connected, and is also provided with a general computer such as a display and a memory. hardware equipment.
  • the image capture card 15-1 is placed in the card slot of the entire computer image processing system 15 and connected to the infrared camera 11 via a video line.
  • Photoelectric mounted in the motor drive and the feed mechanism 4 (or electromagnetic) via the signal cable is connected to the detection means and the image acquisition card 15-1, respectively, and the infrared camera 11, and simultaneously outputs both the arc to 1 (with the welding speed V w In the vertical direction) to the left or right wall of the groove 9, the closest position signal P L or P R .
  • the infrared camera 11 captures the welding area image according to the triggering mode other than the arc position signal P L or P R , and sends the welding image signal to the computer image processing system 15 via the image capturing card 15-1, and undergoes image processing (specific image processing method) After seeing the details below, the weld deviation value output is extracted based on the position information of the welding wire relative to the left and right side walls of the groove for monitoring display and subsequent torch position adjustment.
  • the invention can effectively avoid the influence of the arc shape asymmetry on the weld deviation detection accuracy.
  • the rocking and rotating arc torch of FIG. 1 is changed to other forms of oscillating arc or rotating arc torch, the weld deviation sensing method and apparatus are equally applicable.
  • the connecting cable at one end of the welding power source 14 is passed through the detecting loop of the Hall type current sensor 13 or the shunt type current sensor 13 and then connected to the workpiece 7 and passed through
  • the current sensor 13 detects the arc current in a non-contact or contact manner, and the pulse base current signal i b outputted by the current sensor 13 and the arc position signal P L or P R are combined to act on the trigger signal input end of the infrared camera 11 .
  • the infrared camera 11 is triggered to collect the welding area image which is minimally disturbed by the arc arc at this moment, The welding image acquisition synchronized with the pulse arc base current i b is realized to improve the detection accuracy of the weld deviation during pulse arc welding.
  • the arc shaking parameters include a rocking frequency, a rocking amplitude (ie, a rocking angle), a rocking radius, and a point A 1i of the arc 1 near the left and right side walls of the groove 9 The residence time at A 2i , etc.
  • the O 2i symmetrical arc-shaped trajectory oscillates and tends to stay at a point (i.e., tens to hundreds of milliseconds) from the leftmost wall L of the groove 9 and the closest points A 1i and A 2i of the right side ⁇ R. Its shaking frequency is generally within a few hertz.
  • the chord length of the arc 1 arc-shaped trajectory A 1i O 2i A 2i Point O 1i is biased to the right side of the groove R, at this time, the left side of the weld (or the right deviation of the torch); when A 1i F i ⁇ A 2i E i , as shown in Figure 4, the arc 1 is arc-shaped
  • the midpoint O 1i of the chord length of the track A 1i O 2i A 2i is biased to the left side L of the groove, and this is the right deviation of the weld (or the left deviation of the torch).
  • the following describes the specific implementation steps of the infrared visual sensing detection method for the narrow gap weld deviation of the present invention by using the shaking or rotating arc narrow gap welding application as an example (the narrow gap weld deviation of other forms of moving arc)
  • the detection method is similar to this):
  • narrow gap welding torch drag mechanism (not shown) drives the entire narrow gap welding torch with infrared imaging systems, welding speed V w 9 is moved to the front groove; meantime, according to the motor drive and The arc 1 emitted by the feeding mechanism 4 moves to the leftmost wall (or right side wall) of the groove 9 to the nearest position signal P L (or P R ), triggering the infrared camera 11 to capture the image of the welding area at the moment, and passes through the image capturing card 15- 1 is sent to the computer image processing system 15.
  • the wire position information extracting module 15-2 receives the welding image information from the image capturing card 15-1, This image is processed. During image processing, the horizontal coordinate value of the positioning point B 1i of the image capturing window 17 on the left side of the groove is adjusted according to the change of the lateral position of the left edge line 21 of the groove, and the left edge image of the groove without the arc interference is intercepted, and the left edge of the groove is extracted.
  • the current distance L 2i of the left edge 21 of the groove to the left edge of the global image is obtained; the image of the right position of the wire is intercepted by the right image intercepting window 18 of the wire, and then the center of the wire 5 is extracted on the detecting line 23 to the global image.
  • the wire position information extracting module 15-2 receives the welding image information from the image capturing card 15-1, and then This image is processed. During image processing, the abscissa value of the positioning point B 2i of the image capturing window 20 on the right side of the groove is adjusted according to the change of the lateral position of the right edge line 22 of the groove, and the right side image of the groove without intercepting the arc is intercepted, and the right edge of the groove is extracted.
  • the current distance L 3i of the right edge 22 of the groove to the left edge of the global image is obtained; the image of the left position of the wire is intercepted by the image intercepting window 19 on the left side of the wire, and then the center of the wire 5 is extracted on the detecting line 24 to the global image.
  • the steps 2 and 3 may be reversed, and in the steps 2 and 3, the current distance L 1i or L of the center of the wire 5 to the left boundary of the global image may be extracted first. 4i , then find the current distance L 2i or L 3i of the bevel edge 21 or 22 to the left edge of the global image. And, when the arc 1 is moved again to the closest position to the right side wall of the groove, the wire position information extraction module 15-2 extracts the next distance from the left edge 21 of the groove to the left boundary of the global image in a similar manner as described in step 2.
  • the groove left side image capturing window 17 or the groove right side image capturing window 20 may be disposed on the groove side away from the arc 1, or may be disposed in The groove side closer to the arc 1 is preferably placed on the side of the groove away from the arc 1, see Figs. 8 and 9.
  • the global (ie, overall) welding image is sequentially subjected to median filtering, contrast stretching, global fixed threshold segmentation, and arc contour morphology etching treatment.
  • the highest point C 1 or C 2 position coordinate is (X C1i , Y C1i ) or (X C2i , Y C2i ), and determining the wire image interception according to the ordinate value of the highest point C 1 or C 2 of the arc region
  • the ordinate value of the positioning point D 1i or D 2i of the window 18 or 19 is (Y C1i + ⁇ Y) or (Y C2i + ⁇ Y), where ⁇ Y is an additional adjustment amount while adjusting the wire image intercepting window 18 or 19 positioning point D
  • the abscissa value of 1i or D 2i (the specific algorithm is detailed below).
  • the wire position detecting line 23 or 24 is located in the wire image intercepting window 18 or 19 at a position half or more of the height of the wire image capturing window 18 or 19.
  • the L 2i and L 2(i+1) detection point ordinate values on the groove left edge line 21 are the same as the L 3i detection point ordinate values on the groove right edge line 22, and the right position wire axis
  • the ordinate values of the L 1i and L 1(i+1) detection points on the heart line are the same as the ordinate values of the L 4i detection points on the left axis of the wire; if it is to maintain the ordinate value of the same detection point The same, and the next detection point (such as L 1 (i +1) and L 2 (i +1) detection point) ordinate position beyond the range of the image capture window of its image, then adjust to the next weld deviation detection
  • An arc motion (shake or rotation) period is detected once and then restored to an arc motion cycle for two inspections of the weld deviation.
  • the final detection value of the wire position detection point to the left edge distance L 1i , L 4i or L 1 (i+1) of the global image is the image intercepting window 18 on the left side of the wire or the image intercepting window 19 in the right side of the welding wire, and the welding wire 7
  • the median or average value of m (m ⁇ 1) different position detection values on the axis; the final detection value of the distance L 2i , L 3i or L 2 (i+1) of the groove edge line to the left edge of the global image The median or average value of k (k ⁇ 1) different position detection values in the image intercepting window 17 on the left side of the groove or in the image capturing window 20 on the right side of the groove, k (k ⁇ 1) on the groove edge line 21 or 22.
  • two inspections of the weld deviation during an arc motion (shake or rotation) cycle can be achieved by detecting two adjacent welding images. (without initial testing) to improve weld deviation Real-time detection.
  • the image of the edge region of the groove away from the arc side is intercepted by the small window for processing, which can effectively avoid the motion arc interference and improve the sensing accuracy of the weld deviation.
  • Image processing is the key to weld deviation detection.
  • the process of image processing to detect weld deviation is shown in Figure 10. It includes the following four steps:
  • the image processing workload of the image processing system 15 avoids the interference of the motion arc on the image detection of the groove; 3 the full window
  • Adaptive threshold segmentation that is, black and white processing of images to highlight the edge of the bevel; 4 edge edge extraction: using Canny edge operator (translated by Li Shiying, "Feature Extraction and Image Processing (No. 1 Second Edition), Electronic Industry Press, 2013, p. 103) Smoothing and filtering the window image after adaptive threshold segmentation to detect the change of the gray level of the molten pool and the background image after filtering out the noise.
  • welding wire contour image morphology corrosion the morphological corrosion method is used to process the partial contour segmentation of the wire contour image to remove image noise and highlight the wire contour
  • 4 wire skeleton extraction for the wire profile after morphological corrosion treatment Image, using the Canny edge operator to extract the wire skeleton edge line in a small window.
  • Extracting the center of the wire to the left edge of the global image extracting the wire axis from the wire skeleton and extracting the current distance L 1i or L 1(i+1) from the center of the wire 5 to the left boundary of the global image on the detection line 23 Or, on the detection line 24, the current distance L 4i from the center of the wire 5 to the left boundary of the global image is extracted.
  • the wire mesh interception window positioning point D 1i or D 2i the horizontal coordinate adaptive correction algorithm: the wire image intercepting window 18 or 19 positioning point D 1i or D 2i the abscissa value X d1i or X d2i , equal to its previous setting
  • Example 1 (taking shaking arc DC welding as an example)
  • Fig. 11 is a view showing an example of the effect of extracting the axial position information of the wire during DC welding.
  • 11(a) and (b) are respectively the global welding images of the arc 1 obtained by the computer image processing system 15 shaking to the left and right walls of the groove, and the welding wire and the arc are separated from the left side wall of the groove. Closer, indicating the left deviation of the welding torch, that is, the right deviation state of the welding seam relative to the welding torch (corresponding to the situation shown in Fig.
  • Fig. 12 is a view showing an example of the effect of extracting the groove edge information corresponding to the case shown in Fig. 11.
  • the image of the right side of the groove is used to intercept the window 20 and the slope.
  • the image of the small window of the edge of the groove obtained by the interception window 17 on the left side of the mouth is as shown in Fig.
  • the wire 5 is drawn by the wire (see FIGS. 11(u) and (v).
  • the median as the current distance L 2i or L 3i of the bevel edge 21 or 22 to the left edge of the global image.
  • the absolute error of the weld deviation sampling value ⁇ X si is ⁇ 0.15 mm for the case where the actual deviation of the weld seam continuously changes within ⁇ 1.0 mm, which further illustrates the proposed The effectiveness of the weld deviation bias infrared vision sensing method.
  • Example 2 (taking shaking arc pulse welding as an example)
  • Figure 13 is an example of the extraction effect of the axial position information of the welding wire during pulse welding.
  • the test conditions include: the neutral dimming plate transmittance is 30%, the digital infrared CMOS camera works in the external trigger mode, using pulse MAG arc welding, the average The welding current is 280A, the average arc voltage is 29V, the right deviation of the torch (or the left deviation of the weld) is 0.6mm, and the other test conditions are the same as those in Figure 11.
  • FIG. 13 (c) and (d) show the arc shape image after global image fixed threshold segmentation and morphological erosion processing; for the global pulse welding image after median filtering and contrast stretching, Figure 13 ( a) and (b) where the position of the white wire frame at the wire is determined (the position is adaptively determined according to the position of the highest point C 1 or C 2 of the arc region), and the wire area obtained by the wire image intercepting windows 19 and 18 is small.
  • the images are shown in Figures 13(e) and (f); after the locally adaptive threshold segmentation process is performed on the captured small-window image of the wire region, the obtained wire region image is shown in Figures 13(g) and (h), wherein The white regular area indicates the wire profile; the morphological corrosion of the wire profile image after partial threshold segmentation is performed, and the obtained wire profile image is shown in Figures 13(j) and (k); the wire extracted by the Canny edge operator
  • the image of the contour edge ie, the wire skeleton
  • the geometric center of the wire is obtained for the wire profile obtained after the edge extraction, and the obtained wire axis position image of the wire is as shown in the figure 13(u) and (v) are shown.
  • Fig. 14 is a view showing an example of the effect of extracting the groove edge information corresponding to the case shown in Fig. 13.
  • the groove image intercepting windows 20 and 17 are used for interception.
  • the obtained small window image of the bevel edge area is shown in Fig.
  • the obtained edge edge area image is as shown in the figure 14(c) and (d), where the white area indicates the pool area; the edge image extracted by the Canny edge operator is shown by the white lines in Figs. 14(e) and (f).
  • the wire 5 is drawn by the wire (see FIGS. 13(u) and (v).
  • the absolute error of the current sample value ⁇ X si of the weld deviation is ⁇ 0.15 mm, which further illustrates the welding.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Mechanical Engineering (AREA)
  • Plasma & Fusion (AREA)
  • General Physics & Mathematics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Theoretical Computer Science (AREA)
  • Quality & Reliability (AREA)
  • Optics & Photonics (AREA)
  • Geometry (AREA)
  • Length Measuring Devices By Optical Means (AREA)
  • Butt Welding And Welding Of Specific Article (AREA)

Abstract

一种窄间隙焊缝偏差的红外视觉传感检测方法及装置，该装置包括摇动（或旋转）电弧窄间隙焊炬、电弧电流传感器（3）、计算机图像处理系统（15）、红外摄像系统；红外摄像机（11）以外触方式采集电弧（1）摇动（或旋转）至坡口（9）左侧壁或右侧壁最近处时的焊接区域红外图像，经过计算机图像处理后，实时提取焊丝位置和坡口边缘信息，并根据焊丝相对于坡口左右侧壁位置的变化，求取焊缝偏差输出；脉冲电弧焊接时，通过电流传感器检测电弧电流脉冲基值期信号，可实现与脉冲电弧基值电流期同步的焊接图像采集。该装置构成简单、适用范围宽、焊缝偏差检测精度高、环境适应性好、抗干扰能力强。

Description

窄间隙焊缝偏差的红外视觉传感检测方法及装置 技术领域
本发明涉及焊接技术领域，特指一种对摇动(或摆动)或旋转电弧窄间隙焊缝偏差进行实时检测的红外视觉传感方法及装置。
背景技术
在实际窄间隙熔化极气体保护焊接过程中，往往受到坡口加工误差、工件装配误差、焊接热变形等因素的影响，导致坡口间隙大小和焊缝中心位置发生变化，出现焊炬中心偏离焊缝中心(即出现焊缝偏差)的情形。如果不对焊缝进行跟踪控制，焊缝偏差就会累积，那么，即使采用了基于电弧摆动或旋转工艺的坡口侧壁熔透控制技术，也将会影响窄间隙焊接质量。传感技术是实现焊缝跟踪的关键，焊缝跟踪传感方法分为接触式、非接触式和电弧传感式(即半接触式)三大类。作为非接触式传感方法的代表，视觉传感器因其信息量大、灵敏度高、适用坡口形式多、抗电磁干扰能力强等优点而受到特别重视。
中国专利号为ZL201210325926.9、名称为“基于红外视觉传感的窄间隙焊接监控及焊缝偏差检测方法”的专利文献，针对摇动电弧窄间隙MAG焊接应用，通过红外摄像机事先获取焊炬对中时的左右侧坡口基准边界信息，焊接过程中红外摄像机实时获取远离电弧侧的窄间隙坡口边缘图像后，通过图像处理提取坡口单侧边缘位置信息，并与事先获取的同侧基准坡口边界比较，求取焊缝偏差。其缺点是：焊缝偏差的坡口单侧边缘位置检测仅适用于坡口间隙恒定的情形，不适用于常见的坡口间隙变化场合，因此适用范围窄。
中国专利申请号为201310375369.6、名称为“基于视觉传感的旋转电弧窄间隙MAG焊焊缝偏差识别装置和方法”，针对旋转电弧窄间隙MAG焊接应用，采用电荷耦合器件(CCD)式红外摄像机，通过触发采样方式获取电弧旋转到坡口左、右侧壁位置处的焊接图像，并根据图像处理获取的电弧中心和坡口中心位置信息求取焊缝偏差。其缺点是：①在实际窄间隙焊接过程中，电弧关于其轴线一般呈不对称状态，特别是当电弧离坡口一侧边缘较近时常常会出现电弧沿坡口侧壁攀爬现象，因此根据所述焊缝偏差识别方法提取的电弧形态几何中心，难以准确地反映实际的电弧旋转中心，从而影响其焊缝偏差检测精度，也会影响其工程实用性；②所述焊缝偏差的电弧中心检测法，仅适用于高频率的旋转电弧窄间隙焊接应用，不适用于较低频率的摆动(含摇动)电弧窄间隙焊接应用场合；③采用的CCD摄像机，动态范围小，响应速度较慢，影响焊缝偏差检测精度和实时性。
发明内容
针对现有技术存在的焊缝偏差传感检测精度低、适用范围窄、工程实用性不强等缺点，本发明提出一种适用于摇动或摆动、旋转电弧的窄间隙焊缝偏差的红外视觉传感检测方法及装置，通过检测电弧运动至坡口两侧壁处时焊丝相对于坡口左右侧壁的位置变化来获取焊缝偏差信息，具有检测精度高、环境适应能力强、实用性好等优点。
本发明提出的窄间隙焊缝偏差的红外视觉传感检测装置，包括窄间隙焊炬、电弧电流传感器、计算机图像处理系统以及红外摄像系统；窄间隙焊炬中的折弯导电杆一端与电机驱动及馈电机构相连、另一端与直型导电嘴相接，穿过窄间隙焊炬的焊丝伸入待焊坡口中产生焊接电弧；红外摄像系统包括数字式红外摄像机及红外滤光系统；计算机图像处理系统包含依次连接的图像采集卡、焊丝位置信息提取模块和焊缝偏差值求取模块，图像采集卡通过视频线与红外摄像机相连；电机驱动及馈电机构发出的电弧运动至坡口左、右侧壁最近位置信号PL、PR同时输入红外摄像机和图像采集卡；焊接电源一端与电机驱动及馈电机构相连、另一端连接电缆穿过电流传感器的检测环或串接电流传感器后与工件相接；电弧电流传感器检测到的脉冲焊接电弧基值电流信号ib与电弧位置信号PL(或PR)共同作用在红外摄像机的图像拍摄触发信号输入端。
本发明提出的窄间隙焊缝偏差的红外视觉传感检测方法的技术方案是包括如下步骤：
1)窄间隙焊炬整体与红外摄像系统一起以焊接速度Vw向坡口前方移动，根据电弧位置信号PL或PR触发红外摄像机采集此刻焊接区域图像，并经图像采集卡送入计算机图像处理系统；
2)当电弧运动至离坡口左、右侧壁最近处时，焊丝位置信息提取模块接收到来自图像采集卡的焊接图像信息，并通过对该图像进行处理，提取焊丝右位置检测点至坡口左边缘当前距离X1i以及焊丝左位置检测点至坡口右边缘当前距离X2i，并将该当前距离送至焊缝偏差值求取模块中；
3)焊缝偏差值求取模块按式ΔXi＝(X1i－X2i)/2计算焊缝偏差的当前检测值ΔXi，然后以最近n(n≥1)次焊缝偏差检测值的中值或平均值作为焊缝偏差的当前采样值ΔXsi；若ΔXsi＝0则焊缝无偏差，若ΔXsi＞0则焊丝偏向坡口右侧，若ΔXsi＜0则焊丝偏向坡口左侧。
在所述步骤2)中，当电弧运动至离坡口右侧壁最近处时，焊丝位置信息提取模块根据坡口左边缘线的横向位置变化调整坡口左侧图像截取窗口定位点B1i的横坐标值，截取不受电弧干扰的坡口左侧图像，提取坡口左边缘后，求取坡口左边缘至全局图像左边界的当前距离L2i，并通过焊丝右侧图像截取窗口截取焊丝右位置图像后再提取焊丝中心至全局图像左边界的当前距离L1i，计算焊丝右位置检测点至坡口左边缘当前距离X1i＝(L1i－L2i)；当电弧运动至离坡口左侧壁最近处时，焊丝位置信息提取模块根据坡口右边缘线的横向位置变化调整坡口右侧图像截取窗口定位点B2i的横坐标值，截取不受电弧干扰的坡口右侧图像，提取坡口右边缘后，求取坡口右边缘至全局图像左边界的当前距离L3i，并通过焊丝左侧图像截取窗口截取焊丝左位置图像后再提取焊丝中心至全局图像左边界的当前距离L4i，计算焊丝左位置检测点至坡口右边缘当前距离X2i＝(L3i－L4i)；当电弧再次运动至离坡口右侧壁最近处时，先提取坡口左边缘至全局图像左边界的下次距离L2(i+1)及焊丝右位置检测点至全局图像左边界的下次距离L1(i+1)，再计算焊丝右位置检测点至坡口左边缘的下次距离X1(i+1)＝(L1(i+1)－L2(i+1))，计算出焊缝偏差的下次检测值ΔX(i+1)＝(X1(i+1)－X2i)/2，依此类推，实现在一个电弧运动周期内对焊缝偏差的两次检测。
在坡口左边缘线上的L2i和L2(i+1)检测点纵坐标值与在坡口右边缘线上的L3i检测点纵坐标值相同；在 右位置焊丝轴心线上的L1i和L1(i+1)检测点纵坐标值与在左位置焊丝轴心线上的L4i检测点纵坐标值相同；焊丝位置检测点至全局图像左边界距离L1i、L4i或L1(i+1)的最终检测值为焊丝图像左、右截取窗口内、焊丝轴心线上m(m≥1)个不同位置检测值的中值或平均值；坡口左、右边缘至全局图像左边界距离L2i、L3i或L2(i+1)的最终检测值，为坡口左、右侧图像截取窗口内、坡口左、右边缘线上k(k≥1)个不同位置检测值的中值或平均值。
在对所述图像进行处理时，先通过全局焊接图像处理提取电弧区域最高点坐标值，并根据电弧区域最高点坐标位置的变化调整焊丝左、右侧图像截取窗口定位点的坐标值；再分别对用坡口左、右侧图像截取窗口截取的小窗口坡口图像及用焊丝左、右侧图像截取窗口截取的小窗口焊丝图像进行处理；在对焊丝左、右侧图像截取窗口截取的小窗口图像进行处理时，先进行局部自适应阈值分割处理，然后通过全窗口图像的形态学腐蚀提取焊丝轮廓后，再利用Canny边缘检测算法提取焊丝骨架，最后计算出焊丝轴心线位置。
对于摇动(或旋转)电弧脉冲焊接场合，当电弧运动至坡口左侧壁或右侧壁最近处(电弧位置信号PL或PR信号有效)时，一旦电流传感器检测到脉冲电弧第一个脉冲焊接电弧基值电流信号ib来临，即刻触发红外摄像机，采集此刻受电弧弧光干扰最小的焊接区域图像，实现与脉冲电弧基值电流期同步的焊接图像采集。
本发明与现有技术相比，其有益效果是：
①与现有的电弧中心检测法相比，本发明根据焊丝相对于坡口左右侧壁位置信息来提取焊缝偏差量，可有效地避免电弧形态不对称对检测精度的影响；
②与现有的坡口单侧边缘位置检测法相比，本发明反映的焊丝和坡口边缘位置信息都是动态的，无需事先建立基准模板，适用于坡口间隙动态变化场合；
③对采集的焊接图像进行处理时，本发明根据电弧和坡口边缘位置的变化，实时调整焊丝图像和坡口边缘图像截取窗口的位置，提高了其环境适应性；
④根据相邻两幅焊接图像检测焊丝在焊接坡口内的相对位置，可实现在一个电弧运动周期内对焊缝偏差的两次检测，提高了焊缝偏差检测的实时性；
⑤对于摇动(或摆动、旋转)电弧脉冲焊接场合，本发明采用的脉冲电弧基值电流同步检测法，可采集到电弧弧光干扰最小的焊接图像，进一步提高了焊缝偏差检测精度；
⑥本发明既适用于电弧运动频率较低的摇动(或摆动)电弧窄间隙焊接场合，又可应用于电弧运动频率较高的旋转电弧窄间隙焊接，适用范围宽。
附图说明
下面结合附图和具体实施方式对本发明作进一步详细说明，但本发明保护范围不限于下述实施例，凡采用等同替换或等效变换形式获得的技术方案，均在本发明保护范围之内。
图1为窄间隙焊缝偏差红外视觉传感检测装置的原理框图；
图2为图1中圆弧形摇动电弧焊接时焊缝无偏差(即焊炬对中)情形示意图；
图3为图1中圆弧形摇动电弧焊接时焊缝左偏(即焊炬右偏)情形示意图；
图4为图1中圆弧形摇动电弧焊接时焊缝右偏(即焊炬左偏)情形示意图；
图5为图1中旋转电弧和双半圆周形摇动电弧焊接时焊缝无偏差(即焊炬对中)情形示意图；
图6为图1中旋转电弧和双半圆周形摇动电弧焊接时焊缝左偏(即焊炬右偏)情形示意图；
图7为图1中旋转电弧和双半圆周形摇动电弧焊接时焊缝右偏(即焊炬左偏)情形示意图；
图8为图1中电弧偏向坡口右侧壁时焊缝偏差检测原理图；
图9为图1中电弧偏向坡口左侧壁时焊缝偏差检测原理图；
图10为图1中焊缝偏差检测流程图；
图11为直流焊时焊丝轴心位置信息提取效果实施例图；
图12为直流焊时坡口边缘信息提取效果实施例图；
图13为脉冲焊时焊丝轴心位置信息提取效果实施例图；
图14为脉冲焊时坡口边缘信息提取效果实施例图。
图1中：1—电弧；2—直型导电嘴；3—折弯导电杆；4—电机驱动及馈电机构；5—焊丝；6—送丝机；7—工件；8—电弧双半圆周形摇动轨迹(或电弧圆周形旋转轨迹)；9—待焊坡口；10—电弧圆弧形摇动轨迹；11—红外摄像机；12—红外滤光系统；13—电弧电流传感器；14—焊接电源；15—计算机图像处理系统；15-1—图像采集卡；15-2—焊丝位置信息提取模块；15-3—焊缝偏差值求取模块。ib—脉冲基值电流，Vw—焊接速度，PL—电弧(在与焊接速度Vw垂直方向上)至坡口左侧壁最近位置信号，PR—电弧(在与焊接速度Vw垂直方向上)至坡口右侧壁最近位置信号。
图2至图7中：A1iO2iA2i—电弧相对于焊炬的圆弧形摇动轨迹10；O1i—电弧圆弧形摇动轨迹A1iO2iA2i的弦长中点；O2i—电弧摇动轨迹中点；O3i—焊炬中心在坡口底面的投影点，也是电弧相对于焊炬圆周形摇动轨迹8的圆心点；O4O5—坡口中心线；α—电弧摇动角度；A1i—电弧至坡口左侧璧L的最近位置点(在与焊接速度Vw垂直方向上)；A2i—电弧至坡口右侧璧R的最近位置点(在与焊接速度Vw垂直方向上)；A1iFi—电弧至左侧璧L的最近距离；A2iEi—电弧至右侧璧R的最近距离；EiFi与焊接速度Vw方向垂直；O3iA1i＝O3iA2i＝O3iO2i＝r，为电弧摇动半径或旋转半径。
图8至图9中中：17—坡口左侧图像截取窗口；18—焊丝右侧图像截取窗口；19—焊丝左侧图像截取窗口；20—坡口右侧图像截取窗口；21—坡口左边缘；22—坡口右边缘；23—焊丝右位置检测线；24—焊丝左位置检测线。C1—电弧至坡口右侧壁22最近时电弧区域最高点，C2—电弧至坡口左侧壁21最近时电弧区域最高点；D1i—焊丝右侧图像截取窗口18的当前定位点，D2i—焊丝左图像截取窗口19的当前定位 点；B1i—坡口左侧图像截取窗口17的当前定位点，B2i—坡口右图像截取窗口20的当前定位点；L1i—焊丝右位置中心至全局图像左边界的当前距离；L2i—坡口左边缘21至全局图像左边界的当前距离；L3i—坡口右边缘22至全局图像左边界的当前距离；L4i—焊丝左位置中心至全局图像左边界的当前距离；X1i—焊丝右位置检测点至坡口左边缘21的当前距离；X2i—焊丝左位置检测点至坡口右边缘22的当前距离；ΔXi—焊缝偏差的当前检测值。
具体实施方式
本发明窄间隙焊缝偏差的红外视觉传感检测装置如图1所示，主要包括窄间隙焊炬、电弧电流传感器13、计算机图像处理系统15、红外摄像系统等。窄间隙焊炬又由电机驱动及馈电机构4、折弯导电杆3、直型导电嘴2、喷嘴机构等组成，其中折弯导电杆3一端与电机驱动及馈电机构4相连、另一端与直型导电嘴2相接；窄间隙焊炬中的电机驱动及馈电机构4可以为空心轴电机直接驱动式，也可以为普通电机通过传动副间接驱动式，并且优选空心轴电机直接驱动式。电机驱动及馈电机构4中安装有光电(或电磁)检测装置，以检测电弧1(在与焊接速度Vw垂直方向上)至坡口9左侧壁与右侧壁距离最近位置信号PL或PR。通过焊接电源14、窄间隙焊炬及其拖动控制系统，可焊前设定焊接能量参数(电弧电流、电弧电压和焊接速度Vw)和电弧摇动参数或电弧旋转参数。
送丝机6送出的焊丝5依次穿过电机驱动及馈电机构4、折弯导电杆3和直型导电嘴2的中心孔后送出，伸入待焊坡口9中产生焊接电弧1；焊接电源14的一端与电机驱动及馈电机构4相连，连接电机驱动及馈电机构4中的电刷馈电机构，焊接电源14的另一端与工件7相接，从而实现在焊接电缆无缠绕下的焊接馈电。供气系统通过焊炬中的喷嘴机构向电弧1所在的焊接区域提供活性或惰性保护气体，但是当采用自保护药芯焊丝焊接时则不需要喷嘴机构，也无需向焊接区域提供保护气体。电弧1相对于焊炬的运动方式分为摇动(或摆动)式和旋转式两大类，具体包括：往复式圆弧形10摇动、单向双半圆周形轨迹8式摇动(电弧在离坡口左右侧壁最近位置点A1i和A2i处有停留)、圆周形轨迹式旋转(电弧在离坡口左右侧壁最近位置点A1i和A2i处无停留)；其中，旋转式电弧运动方式除了图1所示的折弯导电杆式旋转电弧以外，还包括常见的偏心导电嘴式旋转电弧、以及导电杆整体绕焊炬中心公转式旋转电弧等。
红外摄像系统包括红外摄像机11及红外滤光系统12，红外滤光系统12与红外摄像机11同轴相连，并安装在红外摄像机11的正前方。红外摄像机11可以为电荷耦合器件(CCD)式或互补金属氧化物半导体(CMOS)式，并且优选CMOS式，将红外摄像机11固定在焊炬的正前方或正后方，使红外摄像机11与待焊坡口9底面成20～60°的夹角，便于红外摄像机11摄取到待焊坡口9的焊接区域图像。选用变焦范围为18～45mm、光圈为5.6～32的数字式红外摄像机11，在外触发方式下获取焊接区域图像。红外滤光系统12包括有窄带滤光片、中性减光片、防护镜，其中窄带滤光片中心波长为700～1100nm，中性减光片透过率为1～50％；防护镜用于防焊接飞溅，窄带滤光片用于滤除弧光、烟尘、飞溅等光辐射干扰，中性减光片可 调节来自于电弧和熔池辐射光强，以便能采集到清晰的焊接红外图像。
计算机图像处理系统15主要包含依次连接的图像采集卡15-1、焊丝位置信息提取模块15-2和焊缝偏差值求取模块15-3，此外还配套有显示器、存储器等普通计算机应有的硬件设备。图像采集卡15-1置于整个计算机图像处理系统15的卡槽内并通过视频线与红外摄像机11相连。安装在电机驱动及馈电机构4中的光电(或电磁)检测装置通过信号电缆分别与图像采集卡15-1和红外摄像机11相连，并同时向二者输出电弧1(在与焊接速度Vw垂直方向上)至坡口9左侧壁或右侧壁距离最近位置信号PL或PR。
红外摄像机11根据电弧位置信号PL或PR以外触发方式摄取焊接区域图像，并将该焊接图像信号经图像采集卡15-1送入计算机图像处理系统15中，经过图像处理(具体图像处理方法详见下文)后，根据焊丝相对于坡口左右侧壁位置信息来提取焊缝偏差值输出，供监控显示和后续焊炬位置调节使用。与现有的电弧中心检测法相比，本发明可有效地避免电弧形态不对称对焊缝偏差检测精度的影响。此外，若将图1中的摇动和旋转电弧焊炬改换成其他形式的摆动电弧或旋转电弧焊炬，那么焊缝偏差传感检测方法及装置同样适用。
在摇动(或旋转)电弧脉冲焊接时，将焊接电源14一端的连接电缆线穿过霍尔型电流传感器13的检测环或串接分流器式电流传感器13后再与工件7相接，并通过电流传感器13以非接触或接触方式检测电弧电流，电流传感器13输出的脉冲基值电流信号ib与电弧位置信号PL或PR相与后，共同作用在红外摄像机11的触发信号输入端，以便在电弧位置信号PL或PR有效期间，一旦电流传感器13检测到脉冲电弧1第一个基值期电流信号来临，即刻触发红外摄像机11，采集此刻受电弧弧光干扰最小的焊接区域图像，实现与脉冲电弧基值电流ib同步的焊接图像采集，以提高脉冲电弧焊接时焊缝偏差检测精度。
在采用往复式圆弧形摇动电弧进行窄间隙焊接时，电弧摇动参数包括摇动频率、摇动幅值(即摇动角度)、摇动半径、以及电弧1在坡口9左右侧壁最近位置点A1i和A2i处的停留时间等。如图2、3和4所示，在i时刻，电弧1围绕焊炬中心O3i、以摇动半径r(r＝O3iA1i＝O3iA2i)和摇动角度α作关于摇动轨迹中点O2i对称的圆弧形轨迹摇动，并往往在离坡口9左侧壁L、右侧璧R的最近点A1i、A2i处作一定时间(一般为数十至数百毫秒)停留，其摇动频率一般在数赫兹以内。当A1iFi＝A2iEi时，如图2所示，电弧1圆弧形摇动轨迹A1iO2iA2i的弦长中点O1i刚好处于坡口中线O4O5上，此时处于焊缝无偏差(即焊炬对中)状态；当A1iFi＞A2iEi时，如图3所示，电弧1圆弧形摇动轨迹A1iO2iA2i的弦长中点O1i偏向坡口右侧R，此时为焊缝左偏(或焊炬右偏)状态；当A1iFi＜A2iEi时，如图4所示，电弧1圆弧形摇动轨迹A1iO2iA2i的弦长中点O1i偏向坡口左侧L，此时为焊缝右偏(或焊炬左偏)状态。
在采用旋转电弧或单向双半圆周形摇动电弧进行窄间隙焊接时，如图5、6和7所示，在i时刻，电弧1围绕焊炬中心O3i、以旋转半径或摇动半径r(r＝O3iA1i＝O3iA2i)作单向圆周运动或单向双半圆周形摇动，其电弧旋转频率一般为数赫兹至100Hz左右；在作单向双半圆周形摇动时，与图2至图4所示往复式圆弧 形摇动电弧焊接情形类似，一般也需要在离坡口左、右侧壁最近点A1i、A2i处作一定时间停留。当A1iFi＝A2iEi时，如图5所示，电弧1运动轨迹的圆心点O3i刚好处于坡口中线O4O5上，此时处于焊缝无偏差(即焊炬对中)状态；当A1iFi＞A2iEi时，如图6所示，电弧1运动轨迹的圆心点O3i偏向坡口右侧R，此时为焊缝左偏(或焊炬右偏)状态；当A1iFi＜A2iEi时，如图7所示，电弧1运动轨迹的圆心点O3i偏向坡口左侧L，此时为焊缝右偏(或焊炬左偏)状态。
参见图1至图9，下面以摇动或旋转电弧窄间隙焊接应用为例，说明本发明窄间隙焊缝偏差的红外视觉传感检测方法的具体实施步骤(其他形式运动电弧的窄间隙焊缝偏差检测方法与此类似)：
①采集焊接区域图像。焊接电弧引燃后，窄间隙焊炬拖动机构(图中未示出)带动窄间隙焊炬整体与红外摄像系统一起，以焊接速度Vw向坡口9前方移动；同时，根据电机驱动及馈电机构4发出的电弧1运动至坡口9左侧壁(或右侧壁)最近位置信号PL(或PR)，触发红外摄像机11采集此刻焊接区域图像，并经图像采集卡15-1送入计算机图像处理系统15。
②提取焊丝在坡口右侧时的位置信息。参见图8，当电弧1运动至离坡口9右侧壁最近处(PR信号有效)时，焊丝位置信息提取模块15-2接收到来自图像采集卡15-1的焊接图像信息后，对该图像进行处理。图像处理时，根据坡口左边缘线21的横向位置变化调整坡口左侧图像截取窗口17定位点B1i的横坐标值，截取不受电弧干扰的坡口左侧图像，提取坡口左边缘21后，再求取坡口左边缘21至全局图像左边界的当前距离L2i；通过焊丝右侧图像截取窗口18截取焊丝右位置图像后，再在检测线23上提取焊丝5中心至全局图像左边界的当前距离L1i；然后，计算焊丝右位置检测点至坡口左边缘21当前距离X1i＝(L1i－L2i)，并将该X1i值送至焊缝偏差值求取模块15-3中。
③提取焊丝在坡口左侧时的位置信息。参见图9，当电弧1运动至离坡口9左侧壁最近处(PL信号有效)时，焊丝位置信息提取模块15-2接收到来自图像采集卡15-1的焊接图像信息后，对该图像进行处理。图像处理时，根据坡口右边缘线22的横向位置变化调整坡口右侧图像截取窗口20定位点B2i的横坐标值，截取不受电弧干扰的坡口右侧图像，提取坡口右边缘22后，再求取坡口右边缘22至全局图像左边界的当前距离L3i；通过焊丝左侧图像截取窗口19截取焊丝左位置图像后，再在检测线24上提取焊丝5中心至全局图像左边界的当前距离L4i；然后，计算焊丝左位置检测点至坡口右边缘22当前距离X2i＝(L3i－L4i)，并将该X2i值送至焊缝偏差值求取模块15-3中。
④求取焊缝偏差信息。焊缝偏差值求取模块15-3接收到焊丝位置信息提取模块15-2相邻两次提取的焊丝位置信息X1i与X2i后，按式ΔXi＝(X1i－X2i)/2计算焊缝偏差的当前检测值ΔXi，然后以最近n(n≥1)次焊缝偏差检测值(即ΔXi-(n-1)、ΔXi-(n-2)、…、ΔXi-1、ΔXi)的中值或平均值，作为焊缝偏差的当前采样值ΔXsi，并将该采样值ΔXsi输出至窄间隙焊炬拖动机构中的焊炬位置调节系统以调节焊炬位置；若ΔXsi＝0则焊缝无偏差(对应于图2和图5所示情形)，若ΔXsi＞0则焊丝5偏向坡口右侧(即表示焊缝相对于焊炬 左偏，对应于图3和图6所示情形))，若ΔXsi＜0则焊丝5偏向坡口左侧(即表示焊缝相对于焊炬右偏，对应于图4和图7所示情形))；重复上述步骤①至步骤④，直至焊接过程结束。
在所述焊缝偏差传感检测方法中，步骤②和步骤③顺序可以颠倒使用，并且在所述步骤②和步骤③中可以先提取焊丝5中心至全局图像左边界的当前距离L1i或L4i，然后再求取坡口边缘21或22至全局图像左边界的当前距离L2i或L3i。并且，当电弧1再次运动至离坡口右侧壁最近处时，焊丝位置信息提取模块15-2按照步骤②所述类似方法，在提取坡口左边缘21至全局图像左边界的下次距离L2(i+1)、以及焊丝右位置检测点至全局图像左边界的下次距离L1(i+1)后，计算出焊丝右位置检测点至坡口左边缘21的下次距离X1(i+1)＝(L1(i+1)－L2(i+1))，并将该X1(i+1)值送入焊缝偏差值求取模块15-3后，在焊缝偏差值求取模块15-3中按式ΔX(i+1)＝(X1(i+1)－X2i)/2求得焊缝偏差的下次检测值ΔX(i+1)；依此类推，从而实现在一个电弧运动(摇动或旋转)周期内对焊缝偏差的两次检测。
在所述焊缝偏差传感检测方法的步骤②和步骤③中，坡口左侧图像截取窗口17或坡口右侧图像截取窗口20可以设置在远离电弧1的坡口侧，也可设置在离电弧1较近的坡口侧，但优选设置在远离电弧1的坡口侧，参见图8和图9。在确定焊丝右侧图像截取窗口18或焊丝左侧图像截取窗口19位置时，对全局(即整体)焊接图像依次进行中值滤波、对比度拉伸、全局固定阈值分割、电弧轮廓形态学腐蚀处理后，提取电弧区域最高点C1或C2位置坐标为(XC1i，YC1i)或(XC2i，YC2i)，并根据电弧区域最高点C1或C2的纵坐标值，确定焊丝图像截取窗口18或19的定位点D1i或D2i的纵坐标值为(YC1i+ΔY)或(YC2i+ΔY)，其中ΔY为附加调节量，同时调整焊丝图像截取窗口18或19定位点D1i或D2i的横坐标值(具体算法详见下文)。
焊丝位置检测线23或24纵坐标位于焊丝图像截取窗口18或19内，并处于焊丝图像截取窗口18或19高度的一半及以上位置处。并且，在坡口左边缘线21上的L2i和L2(i+1)检测点纵坐标值与在坡口右边缘线22上的L3i检测点纵坐标值相同，在右位置焊丝轴心线上的L1i和L1(i+1)检测点纵坐标值与在左位置焊丝轴心线上的L4i检测点纵坐标值相同；如果为了保持与本次同类检测点纵坐标值相同，而出现下次检测点(如L1(i+1)和L2(i+1)检测点)纵坐标位置超出其所在图像截取窗口范围，那么在下次焊缝偏差检测时就调整为一个电弧运动(摇动或旋转)周期检测一次，然后可再恢复至一个电弧运动周期对焊缝偏差的两次检测。另外，焊丝位置检测点至全局图像左边界距离L1i、L4i或L1(i+1)的最终检测值，为焊丝左侧图像截取窗口18或焊丝右侧图像截取窗口19内、焊丝7轴心线上m(m≥1)个不同位置检测值的中值或平均值；坡口边缘线至全局图像左边界距离L2i、L3i或L2(i+1)的最终检测值，为坡口左侧图像截取窗口17或坡口右侧图像截取窗口20内、坡口边缘线21或22上k(k≥1)个不同位置检测值的中值或平均值。
由于在一个电弧运动周期内，可完整地获取两幅焊接图像，因此通过对相邻两幅焊接图像的检测，可实现在一个电弧运动(摇动或旋转)周期内对焊缝偏差的两次检测(不含初始检测)，以提高焊缝偏差 检测的实时性。另外，通过所述小窗口截取远离电弧侧的坡口边缘区域图像进行处理，可有效避免运动电弧干扰，提高了焊缝偏差的传感检测精度。
图像处理是焊缝偏差检测的关键，图像处理检测焊缝偏差的流程如图10所示，包括以下四大步骤：
A)对整幅焊接区域红外图像进行全局图像处理，并确定图像截取窗口位置。具体又包括：①图像滤波处理：针对数字式红外摄像机11采集到的实时焊接图像，采用中值法对整体图像进行平滑滤波，以减低焊接过程中弧光、飞溅、烟尘、外部环境干扰引起的图像噪声；②图像对比度拉伸：通过图像灰度值分析，对输入图像灰度进行分段线性变换，以压缩背景区域灰度，提升熔池和电弧部分像素灰度值，增大坡口侧壁与熔池图像边界处、以及焊丝与熔池图像边界处对比度，实现对图像对比度的拉伸处理，为后续阈值分割创造条件；③全局图像阈值分割：对全局图像进行固定阈值分割，即将包含不同灰度级的图像变为一幅二值图像，以简化图像层次，突出电弧区域轮廓；④提取电弧区域最高点位置：对电弧轮廓进行形态学腐蚀处理，以消除焊丝端部反光区域的影响，提取出电弧形态轮廓，并求出电弧区域最高点C1或C2位置坐标值为(XC1i，YC1i)或(XC2i，YC2i)；⑤自适应修正图像截取窗口位置：根据电弧区域最高点C1或C2的纵坐标值，确定焊丝图像截取窗口18或19的定位点D1i或D2i的纵坐标值为(YC1i+ΔY)或(YC2i+ΔY)，其中ΔY为附加调节量；同时，根据电弧最高点C1或C2横坐标位置的变化，可对焊丝图像截取窗口18或19定位点D1i或D2i的横坐标值Xd1i或Xd2i作适应性修正(具体算法见下文)；根据坡口边缘线21或22横向位置的变化，对坡口边缘图像截取窗口17或20定位点B1i或B2i的横坐标值Xb1i或Xb2i作适应性修正(具体算法见下文)，以提高所述红外视觉传感检测方法的环境适应性和工作可靠性。
B)对小窗口坡口图像进行处理，提取窄间隙坡口边缘位置信息。具体又包括：①判断电弧1至坡口9侧壁最近点位置：利用输入到计算机图像处理系统15中的电弧运动至坡口左或右侧壁位置信号PL或PR，也可根据电弧区域最高点C1或C2的横坐标值XC1i或XC2i，辨识并确认此时电弧运动至离坡口左侧壁最近还是离坡口右侧壁最近；②坡口图像截取：针对在上述步骤A)中经过中值滤波和对比度拉伸处理过的全局焊接图像，采用小窗口17或20截取远离电弧侧的坡口左侧壁边缘21或右侧壁边缘22区域图像，以减少计算机图像处理系统15的图像处理工作量，同时避免运动电弧对坡口图像检测的干扰；③全窗口图像自适应阈值分割：对所截取的全窗口坡口图像进行灰度分析，自动寻找熔池图像与背景图像灰度分布的峰谷点并定位阈值点后，对全窗口坡口图像进行大津法(阮秋琦等译，《数字图像处理(第三版)》，电子工业出版社，2011年，第479页)自适应阈值分割，即对图像进行黑白化处理，以突出坡口边缘；④坡口边缘提取：采用Canny边缘算子(李实英等译，《特征提取与图像处理(第二版)》，电子工业出版社，2013年，第103页)对经过自适应阈值分割处理后的窗口图像进行平滑和滤波处理滤除噪声后，检测熔池与背景图像灰度发生变化的位置，在小窗口内提取出熔池与坡口侧壁之间的边缘线21或22；⑤提取坡口边缘至全局图像左边界距离：求取坡口左边缘21至全局图像左边界的当前距离L2i或L2(i+1)，或者坡口右边缘 22至全局图像左边界的当前距离L3i。
C)对小窗口焊丝图像进行处理，提取焊丝位置信息。具体又包括：①焊丝图像截取：针对在上述步骤A)中经过中值滤波和对比度拉伸处理过的全局焊接图像，采用小窗口18或19截取焊丝图像；②焊丝图像局部自适应阈值分割：针对焊丝图像上下部位灰度值变化大的特点，采用局部自适应阈值分割法(阮秋琦等译，《数字图像处理(第三版)》，电子工业出版社，2011年，第83页)，在对所截取图像每个像素的邻域灰度分析的基础上，依次对不同邻域中的局部图像进行二值化处理，最终实现对全窗口图像的黑白化处理，以增强焊丝局部图像特征；③焊丝轮廓图像形态学腐蚀：采用形态学腐蚀方法，对局部阈值分割后的焊丝轮廓图像进行处理，以去除图像噪声，突出焊丝轮廓；④焊丝骨架提取：针对经过形态学腐蚀处理后的焊丝轮廓图像，采用Canny边缘算子在小窗口内提取出焊丝骨架边缘线。⑤提取焊丝中心至全局图像左边界距离：根据焊丝骨架提取出焊丝轴心线，并在在检测线23上提取焊丝5中心至全局图像左边界的当前距离L1i或L1(i+1)，或者在检测线24上提取焊丝5中心至全局图像左边界的当前距离L4i。
D)求取焊缝偏差信息。按式X1i＝(L1i－L2i)或X1(i+1)＝(L1(i+1)－L2(i+1))计算焊丝右位置检测点至坡口左边缘21的当前距离X1i或下次距离X1(i+1)，并按式X2i＝(L3i－L4i)计算焊丝左位置检测点至坡口右边缘22的当前距离X2i；然后，按式ΔXi＝(X1i－X2i)/2或ΔX(i+1)＝(X1(i+1)－X2i)/2计算焊缝偏差的当前检测值ΔXi或下次检测值ΔX1(i+1)，再以最近n次焊缝偏差检测值的中值或平均值，作为焊缝偏差的当前采样值ΔXsi或下次采样值ΔXs(i+1)输出。
在图10所示的焊缝偏差检测流程中，如将所述步骤B)和步骤C)顺序颠倒，并不影响检测结果。并且，对图像截取小窗口位置横坐标值进行自适应修正的具体算法为：
1)焊丝图像截取窗口定位点D1i或D2i的横坐标自适应修正算法：焊丝图像截取窗口18或19定位点D1i或D2i的横坐标值Xd1i或Xd2i，等于其前次设定值Xd1(i-1)或Xd2(i-1)与电弧最高点C1或C2横坐标位置最近两次检测值的变化量(XC1i－XC1(i-1))或(XC2i－XC2(i-1))之和，即Xd1i＝Xd1(i-1)+(XC1i－XC1(i-1))，或Xd2i＝Xd2(i-1)+(XC2i－XC2(i-1))。
2)坡口边缘图像截取窗口定位点B1i或B2i的横坐标自适应修正算法：坡口边缘图像截取窗口17或20定位点B1i或B2i的横坐标值Xb1i或Xb2i，等于其前次设定值Xb1(i-1)或Xb2(i-1)与坡口边缘线21或22位置最近两次检测值的变化量(L2(i-1)－L2(i-2))或(L3(i-1)－L3(i-2))之和，即Xb1i＝Xb1(i-1)+(L2(i-1)－L2(i-2))，或Xb2i＝Xb2(i-1)+(L3(i-1)－L3(i-2))。
以下提供本发明焊缝偏差传感检测方法的2个实施例。
实施例1(以摇动电弧直流焊接为例)
图11为直流焊时焊丝轴心位置信息提取效果实例图。试验条件包括：数字式CMOS红外摄像机11工作于外触发方式，其摄像角度为25°、光圈16、曝光时间2ms；窄带滤光片中心波长970nm、带宽25nm， 中性减光片透过率10％；平位置直流施焊，电弧电流为280A，电弧电压为29V，焊接速度Vw＝20.3cm/min，焊丝干伸长18mm，焊丝直径1.2mm，焊接保护气体Ar+20％CO2的流量为30L/min，Ⅰ型低碳钢焊接坡口间隙为13mm；在图11所对应时刻焊炬左偏(或焊缝右偏)0.5mm；电弧摇动频率为2.5Hz、摇动半径r＝6.5mm、摇动角度α＝70°，电弧在坡口两侧壁A1i和A2i处停留时间各为100ms。
图11(a)和(b)分别为计算机图像处理系统15获取的电弧1摇动至坡口左侧壁和右侧壁处开始停留瞬间的全局焊接图像，可见焊丝和电弧距离坡口左侧壁较近，说明焊炬左偏，也即焊缝相对于焊炬出现右偏状态(相应于图4所示情形)；针对电弧在坡口左右侧壁停留时的全局焊接图像，经过固定阈值分割(阈值点灰度值为240)和形态学腐蚀后，获得的电弧形态图像如图11(c)和(d)所示；针对经过中值滤波和对比度拉伸处理后的全局焊接图像，在图11(a)和(b)中焊丝处白线框所示位置(其位置由电弧区域最高点C1或C2位置自适应确定，其中纵坐标值附加调节量ΔY＝10个像素，以有效地避开电弧干扰)处，采用焊丝左侧图像截取窗口19和焊丝右侧图像截取窗口18截取得到的焊丝区域小窗口图像如图11(e)和(f)所示；对截取的焊丝区域小窗口图像进行局部自适应阈值分割处理后，获得的焊丝区域图像如图11(g)和(h)所示，其中白色的规则区域表示焊丝轮廓；对经过局部阈值分割处理后的焊丝轮廓图像进行形态学腐蚀，获得的焊丝轮廓图像如图11(j)和(k)所示，可见由于受电弧弧光干扰的影响，所提取的焊丝轮廓呈现上粗下细状态；采用Canny边缘算子提取的焊丝轮廓边缘(即焊丝骨架)图像，如图11(m)和(n)中的白线所示；针对边缘提取后获得的焊丝轮廓求取几何中心，获得的焊丝轴心线位置图像如图11(u)和(v)所示。
图12为与图11所示情形相对应的坡口边缘信息提取效果实例图。针对经过中值滤波和对比度拉伸处理后的全局焊接图像，在图11(a)和(b)中坡口边缘处白线框所示位置处，采用坡口右侧图像截取窗口20和坡口左侧图像截取窗口17截取得到的坡口边缘区域小窗口图像如图12(a)和(b)所示；对截取的小窗口图像采用大津法进行全窗口自适应阈值分割处理后，获得的坡口边缘区域图像如图12(c)和(d)所示，其中白色区域表示熔池区域；采用Canny边缘算子提取的坡口边缘图像，如图12(e)和(f)中的白线所示。
最后，根据图8和图9所示的焊缝偏差检测原理及算法，在焊丝图像截取窗口18或19高度一半以上位置内，以焊丝5轴心线(参见图11(u)和(v))上相邻间隔为4个像素点的5个(即m＝5)不同位置检测值的中值，作为焊丝中心至全局图像左边界的当前距离L1i或L4i；并且，在坡口边缘图像截取窗口17或20高度一半位置处，以坡口边缘线(参见图12(e)和(f))上相邻间隔为10个像素点的3个(即k＝3)不同位置检测值的中值，作为坡口边缘21或22至全局图像左边界的当前距离L2i或L3i。求得焊缝偏差的当前检测值ΔXi后，再以最近5次(即n＝5)焊缝偏差检测值的平均值作为焊缝偏差当前采样值ΔXsi，求得ΔXsi＝0.41mm(绝对检测误差＜0.1mm)，同时ΔXsi>0表示焊缝右偏(或焊炬左偏)。
另外，在同样的直流电弧焊接试验条件下，对于焊缝实际偏差在±1.0mm范围内连续变化的情形，焊缝偏差采样值ΔXsi的绝对误差均＜±0.15mm，从而进一步说明了所提出的焊缝偏差红外视觉传感检测方法的有效性。
实施例2(以摇动电弧脉冲焊接为例)
图13为脉冲焊时焊丝轴心位置信息提取效果实例图，试验条件包括：中性减光片透过率为30％，数字式红外CMOS摄像机工作于外触发方式，采用脉冲MAG电弧焊接，平均焊接电流为280A，平均电弧电压为29V，焊炬右偏(或焊缝左偏)0.6mm，其余试验条件与图11所涉及条件相同。
当电弧摇动至离坡口侧壁较近时，电弧在坡口左侧壁处停留信号PL或在坡口右侧壁处停留信号PR有效，此时一旦电弧电流传感器13检测到脉冲电弧1的第一个基值期电流信号ib来临，即刻触发红外摄像机11，采集此刻受电弧弧光干扰最小的焊接区域图像。相应地，计算机图像处理系统15获取的电弧摇动至坡口左侧壁和右侧壁时的脉冲焊接全局图像如图13(a)和(b)所示，可见焊丝和电弧距离坡口右侧壁较近，说明焊炬右偏(或焊缝相对于焊炬左偏，相应于图3所示情形)；另外，脉冲基值期电弧较小(尽管采用了透过率更大的减光片)，使得电弧弧光干扰减小。图13(c)和(d)所示为经过全局图像固定阈值分割和形态学腐蚀处理后的电弧形态图像；针对经过中值滤波和对比度拉伸处理后的全局脉冲焊接图像，在图13(a)和(b)中焊丝处白线框所示位置(其位置根据电弧区域最高点C1或C2位置自适应确定)处，采用焊丝图像截取窗口19和18截取得到的焊丝区域小窗口图像如图13(e)和(f)所示；对截取的焊丝区域小窗口图像进行局部自适应阈值分割处理后，获得的焊丝区域图像如图13(g)和(h)所示，其中白色的规则区域表示焊丝轮廓；对经过局部阈值分割处理后的焊丝轮廓图像进行形态学腐蚀，获得的焊丝轮廓图像如图13(j)和(k)所示；采用Canny边缘算子提取的焊丝轮廓边缘(即焊丝骨架)图像，如图13(m)和(n)中的白线所示；针对边缘提取后获得的焊丝轮廓求取焊丝几何中心，获得的焊丝轴心线位置图像如图13(u)和(v)所示。
图14为与图13所示情形相对应的坡口边缘信息提取效果实例图。针对经过中值滤波和对比度拉伸处理后的脉冲焊接全局图像，在图13(a)和(b)中坡口边缘处白线框所示位置处，采用坡口图像截取窗口20和17截取得到的坡口边缘区域小窗口图像如图14(a)和(b)所示；对截取的小窗口图像采用大津法进行全窗口自适应阈值分割处理后，获得的坡口边缘区域图像如图14(c)和(d)所示，其中白色区域表示熔池区域；采用Canny边缘算子提取的坡口边缘图像，如图14(e)和(f)中的白线所示。
最后，根据图8和图9所示的焊缝偏差检测原理及算法，在焊丝图像截取窗口18或19高度一半以上位置内，以焊丝5轴心线(参见图13(u)和(v))上相邻间隔为4个像素点的5个不同位置检测值的中值，作为焊丝中心至全局图像左边界的当前距离L1i或L4i；并且，在坡口边缘图像截取窗口17或20高度一半位置处，以坡口边缘线(参见图14(e)和(f))上相邻间隔为10个像素点的3个不同位置检测值的 中值，作为坡口边缘21或22至全局图像左边界的当前距离L2i或L3i。求得焊缝偏差的当前检测值ΔXi后，再以最近5次焊缝偏差检测值的平均值作为焊缝偏差当前采样值ΔXsi，求得ΔXsi＝－0.65mm(负号表示焊缝左偏)，检测绝对误差为0.05mm。
另外，在同样的脉冲电弧焊接试验条件下，对于焊缝实际偏差在±1.0mm范围内连续变化的情形，焊缝偏差当前采样值ΔXsi的绝对误差均＜±0.15mm，从而进一步说明了焊缝偏差的脉冲基值同步检测方法的有效性。

Claims (8)

  1. 一种窄间隙焊缝偏差的红外视觉传感检测装置，包括窄间隙焊炬、电弧电流传感器(13)、计算机图像处理系统(15)以及红外摄像系统；窄间隙焊炬中的折弯导电杆(3)一端与电机驱动及馈电机构(4)相连、另一端与直型导电嘴(2)相接，穿过窄间隙焊炬的焊丝(5)伸入待焊坡口(9)中产生焊接电弧(1)；红外摄像系统包括数字式红外摄像机(11)及红外滤光系统(12)，其特征是：
    计算机图像处理系统(15)包含依次连接的图像采集卡(15-1)、焊丝位置信息提取模块(15-2)和焊缝偏差值求取模块(15-3)，图像采集卡(15-1)通过视频线与红外摄像机(11)相连；电机驱动及馈电机构(4)发出的电弧(1)运动至坡口(9)左、右侧壁最近位置信号PL、PR同时输入红外摄像机(11)和图像采集卡(15-1)；焊接电源(14)一端与电机驱动及馈电机构(4)相连、另一端连接电缆线穿过电流传感器(13)的检测环或串接电流传感器(13)后与工件(7)相接；电弧电流传感器(13)检测到的脉冲焊接电弧基值电流信号ib与电弧位置信号PL或PR共同作用在红外摄像机(11)的图像拍摄触发信号输入端。
  2. 一种如权利要求1所述窄间隙焊缝偏差的红外视觉传感检测装置的检测方法，其特征是包括如下步骤：
    1)窄间隙焊炬整体与红外摄像系统一起以焊接速度Vw向坡口(9)前方移动，根据电弧位置信号PL或PR触发红外摄像机(11)采集此刻焊接区域图像，并经图像采集卡(15-1)送入计算机图像处理系统(15)；
    2)当电弧(1)运动至离坡口(9)左、右侧壁最近处时，焊丝位置信息提取模块(15-2)接收到来自图像采集卡(15-1)的焊接图像信息，并通过对该图像进行处理，提取焊丝右位置检测点至坡口左边缘(21)当前距离X1i以及焊丝左位置检测点至坡口右边缘(22)当前距离X2i，并将该当前距离送至焊缝偏差值求取模块(15-3)中；
    3)焊缝偏差值求取模块(15-3)按式ΔXi＝(X1i－X2i)/2计算焊缝偏差的当前检测值ΔXi，然后以最近n(n≥1)次焊缝偏差检测值的中值或平均值作为焊缝偏差的当前采样值ΔXsi；若ΔXsi＝0则焊缝无偏差，若ΔXsi＞0则焊丝(5)偏向坡口右侧，若ΔXsi＜0则焊丝(5)偏向坡口左侧。
  3. 根据权利要求2所述的检测方法，其特征是：步骤2)中，当电弧(1)运动至离坡口(9)右侧壁最近处时，焊丝位置信息提取模块(15-2)根据坡口左边缘线(21)的横向位置变化调整坡口左侧图像截取窗口(17)定位点B1i的横坐标值，截取不受电弧干扰的坡口左侧图像，提取坡口左边缘(21)后，求取坡口左边缘(21)至全局图像左边界的当前距离L2i，并通过焊丝右侧图像截取窗口(18)截取焊丝右位置图像后再提取焊丝(5)中心至全局图像左边界的当前距离L1i，计算焊丝右位置检测点至坡口左边缘(21)当前距离X1i＝(L1i－L2i)；当电弧(1)运动至离坡口(9)左侧壁最近处时，焊丝位置信息提取模块(15-2)根据坡口右边缘线(22)的横向位置变化调整坡口 右侧图像截取窗口(20)定位点B2i的横坐标值，截取不受电弧干扰的坡口右侧图像，提取坡口右边缘(22)后，求取坡口右边缘(22)至全局图像左边界的当前距离L3i，并通过焊丝左侧图像截取窗口(19)截取焊丝左位置图像后再提取焊丝(5)中心至全局图像左边界的当前距离L4i，计算焊丝左位置检测点至坡口右边缘(22)当前距离X2i＝(L3i－L4i)。
  4. 根据权利要求3所述的检测方法，其特征是：先提取焊丝(5)中心至全局图像左边界的当前距离L1i或L4i，然后再求取坡口左、右边缘(21、22)至全局图像左边界的当前距离L2i或L3i。
  5. 根据权利要求2或3所述的检测方法，其特征是：当电弧(1)再次运动至离坡口右侧壁最近处时，先提取坡口左边缘(21)至全局图像左边界的下次距离L2(i+1)及焊丝右位置检测点至全局图像左边界的下次距离L1(i+1)，再计算焊丝右位置检测点至坡口左边缘(21)的下次距离X1(i+1)＝(L1(i+1)－L2(i+1))，计算出焊缝偏差的下次检测值ΔX(i+1)＝(X1(i+1)－X2i)/2；依此类推，实现在一个电弧运动周期内对焊缝偏差的两次检测。
  6. 根据权利要求3和权利要求5所述的检测方法，其特征是：在坡口左边缘线(21)上的L2i和L2(i+1)检测点纵坐标值与在坡口右边缘线(22)上的L3i检测点纵坐标值相同；在右位置焊丝轴心线上的L1i和L1(i+1)检测点纵坐标值与在左位置焊丝轴心线上的L4i检测点纵坐标值相同；焊丝位置检测点至全局图像左边界距离L1i、L4i或L1(i+1)的最终检测值为焊丝图像左、右截取窗口(19、18)内、焊丝(5)轴心线上m(m≥1)个不同位置检测值的中值或平均值；坡口左、右边缘(21、22)至全局图像左边界距离L2i、L3i或L2(i+1)的最终检测值，为坡口左、右侧图像截取窗口(17、20)内、坡口左、右边缘线(21、22)上k(k≥1)个不同位置检测值的中值或平均值。
  7. 根据权利要求2或3所述的检测方法，其特征是：在对所述图像进行处理时，先通过全局焊接图像处理提取电弧区域最高点坐标值，并根据电弧区域最高点坐标位置的变化调整焊丝左、右侧图像截取窗口(19、18)定位点的坐标值；再分别对用坡口左、右侧图像截取窗口(17、20)截取的小窗口坡口图像及用焊丝左、右侧图像截取窗口(19、18)截取的小窗口焊丝图像进行处理；在对焊丝左、右侧图像截取窗口(19、18)截取的小窗口图像进行处理时，先进行局部自适应阈值分割处理，然后通过全窗口图像的形态学腐蚀提取焊丝轮廓后，再利用Canny边缘检测算法提取焊丝骨架，最后计算出焊丝轴心线位置。
  8. 根据权利要求3所述的检测方法，其特征是：对于摇动(或旋转)电弧脉冲焊接场合，当电弧(1)运动至坡口左侧壁或右侧壁最近处(电弧位置信号PL或PR信号有效)时，一旦电流传感器(13)检测到脉冲电弧第一个脉冲焊接电弧基值电流信号ib来临，即刻触发红外摄像机(11)，采集此刻受电弧弧光干扰最小的焊接区域图像，实现与脉冲电弧基值电流期同步的焊接图像采集。
PCT/CN2015/097552 2014-12-30 2015-12-16 窄间隙焊缝偏差的红外视觉传感检测方法及装置 Ceased WO2016107408A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US15/541,021 US9889528B2 (en) 2014-12-30 2015-12-16 Infrared vision sensing detection method and device for narrow-gap weld seam deviation

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201410833371.8 2014-12-30
CN201410833371.8A CN104551347B (zh) 2014-12-30 2014-12-30 窄间隙焊缝偏差的红外视觉传感检测方法及装置

Publications (1)

Publication Number Publication Date
WO2016107408A1 true WO2016107408A1 (zh) 2016-07-07

Family

ID=53068991

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2015/097552 Ceased WO2016107408A1 (zh) 2014-12-30 2015-12-16 窄间隙焊缝偏差的红外视觉传感检测方法及装置

Country Status (3)

Country Link
US (1) US9889528B2 (zh)
CN (1) CN104551347B (zh)
WO (1) WO2016107408A1 (zh)

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110009603A (zh) * 2019-03-14 2019-07-12 广州供电局有限公司 高压电缆绝缘检测方法及高压电缆维护方法
CN110942457A (zh) * 2019-11-30 2020-03-31 天津大学 基于数字图像处理技术的太阳能电池板缺陷检测方法
CN112270335A (zh) * 2020-09-04 2021-01-26 网络通信与安全紫金山实验室 一种搭接接头焊接质量缺陷预测的方法、系统及计算机可读存储介质
CN113649673A (zh) * 2021-09-23 2021-11-16 塔里木大学 一种用于镇压滚筒焊接的焊缝跟踪方法
CN113837991A (zh) * 2021-06-18 2021-12-24 腾讯云计算（北京）有限责任公司 图像处理方法、装置、设备及存储介质
CN113983959A (zh) * 2021-11-29 2022-01-28 兰州交通大学 不同v形坡口坡口面角下tig电弧形态辅助测量装置
CN116385434A (zh) * 2023-06-02 2023-07-04 同济检测(济宁)有限公司 一种用于预制梁裂缝的智能检测方法
CN117781912A (zh) * 2023-12-28 2024-03-29 宝鸡市东阳机械制造有限公司 一种螺旋齿轴的对齿测量方法及机床调整计算软件

Families Citing this family (58)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104551347B (zh) 2014-12-30 2016-05-25 江苏科技大学 窄间隙焊缝偏差的红外视觉传感检测方法及装置
CN105127632A (zh) * 2015-08-07 2015-12-09 中国石油天然气集团公司 一种焊枪控制方法及装置
US20170246701A1 (en) * 2016-02-29 2017-08-31 Lincoln Global, Inc. System and method for welding using rotating consumable
CN105798426B (zh) * 2016-04-07 2018-02-02 湘潭大学 一种电容式窄间隙焊缝跟踪传感器
CN106141382B (zh) * 2016-07-01 2018-07-06 江苏科技大学 焊接坡口边缘位置视觉传感检测方法
CN106180994B (zh) * 2016-08-10 2018-10-19 哈尔滨工业大学（威海） 一种旋摆式焊炬及焊接方法
CN107962278A (zh) * 2016-10-19 2018-04-27 南京理工大学 一种基于透红外滤光片的弧焊熔池图像采集装置
CN106624266B (zh) * 2016-12-31 2018-08-07 东莞职业技术学院 一种用于汽车焊接的焊缝偏差及熔透状态监测方法
US11065707B2 (en) * 2017-11-29 2021-07-20 Lincoln Global, Inc. Systems and methods supporting predictive and preventative maintenance
CN108526745A (zh) * 2018-06-27 2018-09-14 苏州瑞耀三维科技有限公司 一种基于磁光传感器的激光焊缝检测装置
TWI674422B (zh) * 2018-07-03 2019-10-11 皓琪科技股份有限公司 非接觸式光學量測方法與系統
CN108908403A (zh) * 2018-07-25 2018-11-30 武汉瑞博奥科技有限公司 基于红外视觉和智能优化算法的精密无人智能光电系统
CN109187637B (zh) * 2018-07-31 2020-05-19 中国科学院自动化研究所 基于红外热像仪的工件缺陷测量方法和系统
EP3611496B1 (fr) * 2018-08-16 2025-10-15 Eiffage Energie Systemes - Clemessy Procédé de contrôle d'un cordon de jointure d'une pièce industrielle et système pour mettre en oeuvre ledit procédé
CN109741311B (zh) * 2018-12-28 2022-09-02 西南交通大学 带伪边缘的铝合金熔焊背面熔宽检测方法
CN109559314B (zh) * 2019-01-18 2023-04-07 西南交通大学 一种基于机器视觉的电磁悬浮球系统及其图像处理方法
CN109604767B (zh) * 2019-01-30 2021-01-26 浙江师范大学 用于厚板窄间隙gma机器人的智能焊接控制系统及方法
CN109848524B (zh) * 2019-02-02 2020-12-01 浙江师范大学 窄间隙熔化极气体保护立焊焊缝跟踪方法
US11176656B2 (en) * 2019-02-28 2021-11-16 Fei Company Artificial intelligence-enabled preparation end-pointing
CN111805062A (zh) * 2019-04-12 2020-10-23 住友重机械工业株式会社 焊接观察装置及焊接系统
CN110091104A (zh) * 2019-06-18 2019-08-06 江苏国电新能源装备有限公司 一种焊接装置及焊接方法
CN111121639B (zh) * 2019-12-18 2021-07-27 同济大学 一种刚柔一体的建筑物狭窄空间的裂缝检测系统
CN113000988A (zh) * 2019-12-22 2021-06-22 深圳市新盛机器人有限公司 一种视觉焊接跟踪传感系统处理方法
CN110866920B (zh) * 2020-01-20 2020-05-08 长沙琢冶信息技术有限公司 镀铜焊丝的缺陷检测方法、装置、设备及存储介质
CN111398748A (zh) * 2020-03-12 2020-07-10 山西耀源电力工程有限公司 一种冲击电弧图像边缘检测方法及装置
JP7412237B2 (ja) * 2020-03-23 2024-01-12 株式会社東芝 検査装置及び溶接装置
CN111754490B (zh) * 2020-06-28 2023-08-08 湖南红太阳光电科技有限公司 一种基于视觉的石墨舟倒片检测方法、装置及系统
CN112171022B (zh) * 2020-09-10 2022-03-08 中车长春轨道客车股份有限公司 一种基于同步电参数与视觉的摆动焊接优化方法
CN112179245B (zh) * 2020-09-30 2022-05-13 中国十七冶集团有限公司 一种电渣压力焊焊包高度检测装置及使用方法
CN112489010A (zh) * 2020-11-27 2021-03-12 桂林电子科技大学 一种焊缝快速识别方法、装置及存储介质
CN113523508A (zh) * 2021-08-02 2021-10-22 江苏徐工工程机械研究院有限公司 环焊缝窄间隙mag焊接过程模拟、熔池信息采集装置及方法
CN115761252A (zh) * 2021-08-27 2023-03-07 中国石油天然气集团有限公司 一种提高边缘采样精度的焊缝图像预处理方法
CN113936152B (zh) * 2021-09-26 2024-04-12 武汉纺织大学 一种基于全局—局部双摄图像的丝束匹配定位方法
CN113850802B (zh) * 2021-10-19 2024-06-14 内蒙古工业大学 基于matlab的焊接电弧特征谱线图像光谱数据自动提取方法
CN116060812B (zh) * 2021-10-29 2024-10-18 中联重科股份有限公司 轴套组件的焊接方法及焊接装置
CN113941763B (zh) * 2021-11-22 2023-07-25 江苏科技大学 一种采用粗焊丝的摇动/旋转电弧熔化极焊接方法
US12406522B2 (en) * 2021-12-10 2025-09-02 International Business Machines Corporation Filtered images for 3D body tracking
CN114406407A (zh) * 2022-01-12 2022-04-29 江苏科技大学 用于窄间隙脉冲电弧焊接的plc控制系统及控制方法
CN114393286B (zh) * 2022-01-19 2023-05-26 湘潭大学 一种用于机械手的扭转摆动电弧的焊缝偏差提取方法
CN115070172B (zh) * 2022-04-28 2025-09-02 江苏科技大学 一种采用摇动电弧的自由成形打底焊方法
CN115131236B (zh) * 2022-06-23 2025-02-14 上海应用技术大学 一种自适应弧光区域修复与增强方法
CN115255565B (zh) * 2022-07-07 2024-11-29 江苏科技大学 全局模式识别的窄间隙焊接坡口边缘视觉传感检测方法及应用
CN114871532B (zh) * 2022-07-11 2022-10-11 四川空分设备(集团)有限责任公司 一种换热管与管板手工深孔焊方法
CN115178837B (zh) * 2022-08-24 2024-07-23 中国核工业第五建设有限公司 焊缝偏差量实时检测方法、系统和焊缝偏差纠正方法
CN115415649A (zh) * 2022-09-06 2022-12-02 上海工程技术大学 基于长波滤光的gmaw熔滴熔池图像识别方法和设备
CN116038095B (zh) * 2023-02-21 2025-11-14 大连交通大学 一种基于图像识别的焊缝跟踪测温装置及方法
CN116100124B (zh) * 2023-03-17 2025-04-11 湘潭大学 一种基于电弧声复合激光视觉的焊缝跟踪方法及系统
CN116944722A (zh) * 2023-07-13 2023-10-27 哈尔滨理工大学 一种基于嵌入式Linux平台的焊缝跟踪系统构建方法
CN117047231B (zh) * 2023-08-17 2025-03-18 江苏科技大学 一种窄间隙焊接视觉传感焊缝跟踪方法
CN117324725A (zh) * 2023-10-08 2024-01-02 兰州理工大学 一种基于视觉的窄间隙焊缝跟踪控制方法
CN117252883B (zh) * 2023-11-20 2024-04-02 宁德时代新能源科技股份有限公司 焊接缺陷检测方法、装置和存储介质
CN117308801B (zh) * 2023-11-29 2024-02-13 克恩-里伯斯(太仓)有限公司 一种检测涡卷弹簧最小间隙的方法
CN117381261B (zh) * 2023-12-13 2024-03-08 德阳市华建机械设备有限公司 一种自动化焊接机的故障识别装置及方法
CN117444458B (zh) * 2023-12-21 2024-03-19 深圳杩蚁科技有限公司 一种用于焊接机器人的自动控制方法
CN117517998B (zh) * 2024-01-04 2024-04-02 深圳达温技术服务有限公司 一种电池焊接不良快速自动筛查方法
CN118196106B (zh) * 2024-05-20 2024-07-23 浙江泰嘉和电器有限公司 一种断路器导电系统图像检测方法、设备及介质
CN120696657B (zh) * 2025-06-18 2026-01-27 中国水利水电第十二工程局有限公司 1000MPa级水电钢焊接缺陷检测与返修方法
CN120876454B (zh) * 2025-09-24 2025-12-09 陕西蒲白矿山支护有限公司 一种注浆锚管焊端头焊接质量检测方法

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH09295146A (ja) * 1996-04-30 1997-11-18 Mitsubishi Heavy Ind Ltd ウィービング溶接における溶接線倣い方法
JP2002361424A (ja) * 2001-06-01 2002-12-18 Hitachi Ltd パルスアーク溶接装置
CN102814574A (zh) * 2012-09-06 2012-12-12 江苏科技大学 基于红外视觉传感的窄间隙焊接监控及焊缝偏差检测方法
CN103464869A (zh) * 2013-08-26 2013-12-25 江苏科技大学 基于视觉传感的旋转电弧窄间隙mag焊焊缝偏差识别装置和方法
CN104551347A (zh) * 2014-12-30 2015-04-29 江苏科技大学 窄间隙焊缝偏差的红外视觉传感检测方法及装置

Family Cites Families (14)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US3532807A (en) * 1967-10-05 1970-10-06 Webb James E Automatic closed circuit television arc guidance control
GB1443452A (en) * 1973-06-18 1976-07-21 British Steel Corp Monitoring of welding
US4021840A (en) * 1975-01-24 1977-05-03 General Dynamics Corporation Seam tracking welding system
US4183055A (en) * 1978-08-24 1980-01-08 The United States Of America As Represented By The United States Department Of Energy Beam/seam alignment control for electron beam welding
JPS58132370A (ja) * 1982-01-29 1983-08-06 Mitsubishi Heavy Ind Ltd 溶接ア−ク位置の制御方法
US4532408A (en) * 1982-10-22 1985-07-30 The Ohio State University Apparatus and method for viewing molten pools in arc welding
US4599506A (en) * 1984-12-05 1986-07-08 Westinghouse Electric Corp. Visual observation of welding operation
US5292195A (en) * 1992-09-09 1994-03-08 Martin Marietta Corporation Thermographic evaluation technique
GB9707954D0 (en) * 1997-04-19 1997-06-11 Integrated Sensor Systems Limi An improved remote operator viewing system for arc welding
JP3408749B2 (ja) * 1998-09-07 2003-05-19 三菱重工業株式会社 自動溶接装置
US20070075048A1 (en) * 2005-09-30 2007-04-05 Nachi-Fujikoshi Corp. Welding teaching point correction system and calibration method
WO2009003702A1 (de) * 2007-07-04 2009-01-08 Thermosensorik Gmbh Verfahren zur automatischen inspektion einer schweissnaht mittels wärmefluss-thermograhpie
US8144193B2 (en) * 2009-02-09 2012-03-27 Recognition Robotics, Inc. Work piece tracking system and method
US8532402B2 (en) * 2011-06-22 2013-09-10 The Boeing Company Image registration

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH09295146A (ja) * 1996-04-30 1997-11-18 Mitsubishi Heavy Ind Ltd ウィービング溶接における溶接線倣い方法
JP2002361424A (ja) * 2001-06-01 2002-12-18 Hitachi Ltd パルスアーク溶接装置
CN102814574A (zh) * 2012-09-06 2012-12-12 江苏科技大学 基于红外视觉传感的窄间隙焊接监控及焊缝偏差检测方法
CN103464869A (zh) * 2013-08-26 2013-12-25 江苏科技大学 基于视觉传感的旋转电弧窄间隙mag焊焊缝偏差识别装置和方法
CN104551347A (zh) * 2014-12-30 2015-04-29 江苏科技大学 窄间隙焊缝偏差的红外视觉传感检测方法及装置

Cited By (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110009603A (zh) * 2019-03-14 2019-07-12 广州供电局有限公司 高压电缆绝缘检测方法及高压电缆维护方法
CN110009603B (zh) * 2019-03-14 2021-01-29 广东电网有限责任公司广州供电局 高压电缆绝缘检测方法及高压电缆维护方法
CN110942457A (zh) * 2019-11-30 2020-03-31 天津大学 基于数字图像处理技术的太阳能电池板缺陷检测方法
CN110942457B (zh) * 2019-11-30 2023-12-08 天津大学 基于数字图像处理技术的太阳能电池板缺陷检测方法
CN112270335A (zh) * 2020-09-04 2021-01-26 网络通信与安全紫金山实验室 一种搭接接头焊接质量缺陷预测的方法、系统及计算机可读存储介质
CN113837991A (zh) * 2021-06-18 2021-12-24 腾讯云计算（北京）有限责任公司 图像处理方法、装置、设备及存储介质
CN113649673A (zh) * 2021-09-23 2021-11-16 塔里木大学 一种用于镇压滚筒焊接的焊缝跟踪方法
CN113983959A (zh) * 2021-11-29 2022-01-28 兰州交通大学 不同v形坡口坡口面角下tig电弧形态辅助测量装置
CN116385434A (zh) * 2023-06-02 2023-07-04 同济检测(济宁)有限公司 一种用于预制梁裂缝的智能检测方法
CN116385434B (zh) * 2023-06-02 2023-08-08 同济检测(济宁)有限公司 一种用于预制梁裂缝的智能检测方法
CN117781912A (zh) * 2023-12-28 2024-03-29 宝鸡市东阳机械制造有限公司 一种螺旋齿轴的对齿测量方法及机床调整计算软件

Also Published As

Publication number Publication date
CN104551347B (zh) 2016-05-25
US20180015571A1 (en) 2018-01-18
US9889528B2 (en) 2018-02-13
CN104551347A (zh) 2015-04-29

Similar Documents

Publication Publication Date Title
CN104551347B (zh) 窄间隙焊缝偏差的红外视觉传感检测方法及装置
US10870162B2 (en) Adaptive control method and equipment of arc swing in narrow gap welding
CN103464869B (zh) 基于视觉传感的旋转电弧窄间隙mag焊焊缝偏差识别装置和方法
CN102814574B (zh) 基于红外视觉传感的窄间隙焊接监控及焊缝偏差检测方法
CN104493336B (zh) 基于视频分析的焊缝检测与跟踪方法
CN103506756B (zh) 基于熔池图像视觉传感的激光搭接焊间隙检测系统及方法
JP5338890B2 (ja) レーザ溶接の溶接位置検出装置および溶接位置検出方法
CN108620714B (zh) 基于gmaw熔池轮廓特征的焊接偏差检测系统及其检测方法
CN103537797A (zh) 基于等离子体图像激光搭接焊间隙检测方法及系统
CN106141382B (zh) 焊接坡口边缘位置视觉传感检测方法
CN118123194A (zh) 基于视觉厚板坡口tig仰焊根部熔透状态识别方法
CN104070264A (zh) 变坡口宽度的旋转电弧窄间隙mag焊自适应摆动装置和方法
CN118664025A (zh) 窄间隙焊缝偏差的被动视觉传感抗干扰检测方法及应用
CN115255565B (zh) 全局模式识别的窄间隙焊接坡口边缘视觉传感检测方法及应用
CN114292021A (zh) 石英光纤拉丝过程中预制棒实时对中调整系统及方法
Zhang et al. A study on seam tracking in robotic GMA welding process
JP2013025583A (ja) 車両監視装置
CN204789392U (zh) 一种视觉检测倒角面的光路系统
CN110852275A (zh) 一种球机跟踪过程中人脸检测再放大的跟踪方法
CN120782699A (zh) 焊接质量评估方法、工业相机及相关设备
CN115060648A (zh) 基于立体相机机器视觉的金属表面缺陷检测装置及方法
CN117047231B (zh) 一种窄间隙焊接视觉传感焊缝跟踪方法
CN119540203A (zh) 摇动电弧打底焊缝熔透的红外视觉传感检测方法及装置
CN118752037A (zh) 基于电弧视觉信息的窄间隙焊缝偏差检测方法及应用
CN105844268A (zh) 一种基于视频的引风机静叶监视方法及系统

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 15875086

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

WWE Wipo information: entry into national phase

Ref document number: 15541021

Country of ref document: US

122 Ep: pct application non-entry in european phase

Ref document number: 15875086

Country of ref document: EP

Kind code of ref document: A1

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 27/11/2017)

122 Ep: pct application non-entry in european phase

Ref document number: 15875086

Country of ref document: EP

Kind code of ref document: A1