CN103521890B - Two-sided pair of arc vertical position welding near infrared vision sensing and penetration control device and method - Google Patents

Two-sided pair of arc vertical position welding near infrared vision sensing and penetration control device and method Download PDF

Info

Publication number
CN103521890B
CN103521890B CN201310477515.6A CN201310477515A CN103521890B CN 103521890 B CN103521890 B CN 103521890B CN 201310477515 A CN201310477515 A CN 201310477515A CN 103521890 B CN103521890 B CN 103521890B
Authority
CN
China
Prior art keywords
welding
penetration
molten bath
welding gun
arc
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.)
Active
Application number
CN201310477515.6A
Other languages
Chinese (zh)
Other versions
CN103521890A (en
Inventor
王晓宇
李明杰
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.)
Individual
Original Assignee
Individual
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 Individual filed Critical Individual
Priority to CN201310477515.6A priority Critical patent/CN103521890B/en
Publication of CN103521890A publication Critical patent/CN103521890A/en
Application granted granted Critical
Publication of CN103521890B publication Critical patent/CN103521890B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B21—MECHANICAL METAL-WORKING WITHOUT ESSENTIALLY REMOVING MATERIAL; PUNCHING METAL
    • B21C—MANUFACTURE OF METAL SHEETS, WIRE, RODS, TUBES OR PROFILES, OTHERWISE THAN BY ROLLING; AUXILIARY OPERATIONS USED IN CONNECTION WITH METAL-WORKING WITHOUT ESSENTIALLY REMOVING MATERIAL
    • B21C37/00—Manufacture of metal sheets, bars, wire, tubes or like semi-manufactured products, not otherwise provided for; Manufacture of tubes of special shape
    • B21C37/06—Manufacture of metal sheets, bars, wire, tubes or like semi-manufactured products, not otherwise provided for; Manufacture of tubes of special shape of tubes or metal hoses; Combined procedures for making tubes, e.g. for making multi-wall tubes
    • B21C37/12—Making tubes or metal hoses with helically arranged seams
    • B21C37/122—Making tubes or metal hoses with helically arranged seams with welded or soldered seams
    • 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
    • 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
    • B23K2101/00—Articles made by soldering, welding or cutting
    • B23K2101/04—Tubular or hollow articles
    • B23K2101/06—Tubes

Landscapes

  • Engineering & Computer Science (AREA)
  • Mechanical Engineering (AREA)
  • Physics & Mathematics (AREA)
  • Plasma & Fusion (AREA)
  • Arc Welding In General (AREA)
  • Butt Welding And Welding Of Specific Article (AREA)

Abstract

The invention discloses a kind of two-sided pair of arc vertical position welding near infrared vision sensing and penetration control device and method, be specially adapted to the online penetration control of the double sides double arc termination process such as aluminium alloy, high-strength steel, stainless steel, titanium alloy spiral pipe; The online penetration of the field joint vertical position welding processes such as boats and ships, petroleum storage tank, nuclear power container that are also applicable to controls.Real time double face of the present invention detects vertical position welding molten bath visual pattern, by image procossing and penetration characteristic parameter extraction, obtain the information of two-sided pair of arc vertical position welding root penetration process, set up the Controlling model of root penetration, and automatically adjust the technological parameters such as welding current, torch swinging width, speed of welding, control the whole penetration of the root of weld, ensure that double sides double arc connects quality.

Description

Two-sided pair of arc vertical position welding near infrared vision sensing and penetration control device and method
Technical field
The present invention relates to a kind of arc-welding penetration control method, particularly a kind of two-sided pair of arc vertical position welding near infrared vision sensing and penetration control device and method.
Background technology
The welding production of the helix tube such as aluminium alloy, low-alloy high-strength steel, form helix tube through helix tube production line through manufacture technics such as uncoiling, smoothing, trimming, shaping, compression, welding, full circle by band, welding procedure is the key technology that helix tube manufactures, welding quality particularly " penetration " quality directly determines the mechanical property of helix tube welding point, determines the bearing capacity of helix tube.
Current steel helix tube generally adopts submerged arc welding technique, relies on welding by both sides to ensure penetration, does not adopt visual sensing, also do not adopt the penetration control technology based near infrared vision sensing; The aluminum alloy spiral pipe domestic only Jiangsu prosperous electrical equipment of gold and Institutes Of Technology Of Nanjing have carried out the research that aluminum alloy spiral pipe double sides double arc connects technique, and there is not been reported for aluminum alloy spiral pipe two-sided pair of arc penetration control technology.Naval vessel hull housing segments and complete cross section are built and are closed up welding, execute-in-place adopts vertical position welding technique, Harbin Institute of Technology and Jinxi Shipbuilding Co., Ltd have studied two-sided double T IG+MIG solder technology for ship body folding, but do not propose not study the two-sided pair of arc visual sensing penetration control method based near infrared vision sensing and technology.
Double sides double arc connects two-sided, twin arc and synchronously heats simultaneously, form a common molten bath, welding process influence factor is numerous, complex process, thermal weld stress is not enough or heating location is forbidden, groove root can produce incomplete fusion, cause non-penetration, affect joint carrying, thermal weld stress is excessive, groove root is burnt, cause welding line joint burn-through defect, weld unsuccessfully, welding process detects weld penetration information, on-line tuning technological parameter ensures two-sided pair of arc penetration quality, is the important technical method of helix tube, naval vessel body vertical position welding.But there is no associated description in prior art.
Summary of the invention
The object of the present invention is to provide a kind of double sides double arc near infrared vision sensing and penetration control method.
The technical solution realizing the object of the invention is: a kind of two-sided pair of arc vertical position welding near infrared vision sensing and penetration control device, comprise near-infrared vision inner sensor, near-infrared vision outer sensor, shielded signal line, image pick-up card, computer, inside weld PTIG head, interior welding gun, outside weld PTIG head, outer welding gun, leveling apparatus, bead cutter, three-point bending forming machine;
Bead cutter is positioned at the front of leveling apparatus, and three-point bending forming machine is positioned at the front of bead cutter, and strip digs limit after leveling apparatus smoothing in bead cutter, is finally processed into helix tube through three-point bending forming machine;
Interior welding gun is fixed on inside weld PTIG head, outer welding gun is fixed on outside weld PTIG head, interior welding gun is positioned at the inner side of helix tube to be welded, outer welding gun is positioned at the outside of helix tube to be welded, seam center to be welded all aimed at by interior welding gun, outer welding gun, and interior welding gun, outer welding gun are coaxially arranged, the axis of the two overlaps with the normal of helix tube contact to be welded, interior welding gun is arranged interior welding gun tungsten electrode, outer welding gun is arranged outer welding gun tungsten electrode, the dead in line of interior welding gun tungsten electrode and outer welding gun tungsten electrode; Interior welding gun is connected near-infrared vision inner sensor, the axis of near-infrared vision inner sensor and the axis of interior welding gun are 35 ° ~ 45 °, outer welding gun is connected near-infrared vision outer sensor, the axis of near-infrared vision outer sensor and the axis of outer welding gun are 35 ° ~ 45 °, and the axis of near-infrared vision inner sensor and near-infrared vision outer sensor is all through center, molten bath; Near-infrared vision inner sensor is arranged inner sensor camera lens, near-infrared vision inner sensor overlaps with inner sensor lens axis, near-infrared vision outer sensor is arranged outer sensor camera lens, the dead in line of near-infrared vision outer sensor and outer sensor camera lens, inner sensor camera lens and outer sensor camera lens are all apart from weld pool surface 3-6cm;
Near-infrared vision inner sensor, near-infrared vision outer sensor are all connected with image pick-up card by shielded signal line, and by the Signal transmissions that detects to image pick-up card, this image pick-up card is connected with computer simultaneously.
Based on a welding quality control method for above-mentioned two-sided pair of arc vertical position welding near infrared vision sensing and penetration control device, comprise the following steps:
Step 1, near-infrared vision inner sensor and near-infrared vision outer sensor real time bidirectional continuous sensing molten bath visual pattern, and by image transmitting to image pick-up card;
Molten bath visual pattern is transferred to computer by step 2, image pick-up card, carry out molten bath visual pattern subregion contour curve and characteristic parameter extraction, after windowing, medium filtering, edge corrosion micronization processes, obtain the penetration contour images of molten bath visual pattern outline and central area, molten bath aperture;
Step 3, the molten bath visual pattern outline extracted according to step 2 and the penetration contour images of central area, molten bath aperture, obtain the geometric parameter of the penetration contour feature of molten bath outline and central area, molten bath aperture, specifically comprise pool width B, molten bath length L, penetration hole diameter d, molten bath area S, rear towing angle β, penetration aperture circularity Δ vwith the mean value s of aperture area v, maximum diameter d m, and d mthe maximum pool width B of homonymy m, long, the minimum diameter d in molten bath i, and d ihomonymy pool width B i;
Step 4, set up two-sided pair of arc penetration judgment models and burn-through judgment models; Described two-sided pair of arc penetration judgment models is: bilateral penetration aperture maximum diameter d mbe greater than 3mm, homonymy pool width B mbe greater than 0.75h and be less than 1.2h, wherein h is thickness of slab, and unit is mm, homonymy pool width B m0.6 is greater than with the ratio of molten bath length L;
Double sides double arc wears judgment models: bilateral penetration aperture minimum diameter d ibe greater than 5mm, homonymy pool width B mbe greater than 1.1h, homonymy pool width B m0.9 is greater than with the ratio of molten bath length L.
Two-sided pair of arc penetration judgment models in step 5, foundation step 4 carries out molten bath situation judgement: if burn-through, then improve arc voltage and wire feed rate, welding gun flashes back and again welds and return step 1; If non-burn-through, judge whether through welding according to the penetration model in step 4, if through welding, the welding current that the fuzzy control method according to step 6 obtains, the optimum value of torch swinging width, adjustment twin arc welding connects electric current, torch swinging width; If lack of penetration but at least one side there is penetration aperture, i.e. d>0, and at least one side molten bath area is greater than critical value S k, then the welding current degree obtained according to the fuzzy control method of step 6, the optimum value of speed of welding, adjustment twin arc welding connects electric current, speed of welding; If there is not penetration aperture, namely d be 0 or molten bath area S be less than critical value S k, then reduce speed of welding and again weld and return step 1;
Step 6, adopt Fuzzy-neural-network control method to calculate the optimum value of welding current in two-sided pair of arc penetration weld termination process, torch swinging width, speed of welding in real time, adjust and return step 1 when accordingly welding current, swing width, speed of welding being put into effect according to the molten bath estimate of situation in step 5.
Two-sided pair of arc penetration welding condition Fuzzy-neural-network control method is: with the pool width B, the molten bath length L that obtain in step 3, penetration hole diameter d, molten bath area S, rear towing angle β, penetration aperture circularity Δ vwith the mean value s of aperture area vas the input variable of fuzzy neural network, set up five layers of fuzzy neural network, one to three layers of network realize the fuzzy reasoning of fuzzy control, and be input layer, obfuscation layer Sum fanction layer respectively, four layers realize de-fuzzy, and five layers is output layer; Wherein, in obfuscation layer, adopt triangular function to carry out obfuscation to input variable, transform to respective domain scope, obtain fuzzy variable and the fuzzy variable degree of membership of each input variable; In rules layer, the degree of membership that obfuscation obtains is multiplied between two, the fuzzy inference rule of Erecting and improving; In de-fuzzy layer, adopt weight average de-fuzzy method, the fringe of inference conclusion variable is converted into the amount of the determination state corresponding to network input value; Formula is utilized at output layer: calculate and export determine state welding current, torch swinging width, speed of welding optimum value, wherein be the weights of the 5th layer network, O (4)be the output of the 4th layer network, y ibe the output of the 5th layer network, i is the neuronic number of layer 5.
According to the molten bath estimate of situation in step 5, if through welding, the output layer parameter of two-sided pair of arc penetration welding condition Fuzzy-neural-network is the optimum value of welding current, torch swinging width; If lack of penetration but at least one side there is penetration aperture, i.e. d>0, at least one side molten bath area is greater than critical value S k, then the output layer parameter of two-sided pair of arc penetration welding condition Fuzzy-neural-network is the optimum value of welding current, speed of welding.
The present invention compared with prior art, its remarkable advantage is: (1) a kind of double sides double arc near infrared vision sensing of the present invention and penetration control method, is specially adapted to the online penetration of the double sides double arc termination process such as aluminium alloy, high-strength steel, stainless steel, titanium alloy spiral pipe and controls; (2) the present invention is also applicable to the online penetration control of the field joint vertical position welding processes such as boats and ships, petroleum storage tank, nuclear power container; (3) real time double face of the present invention detects vertical position welding molten bath visual pattern, obtain the information of two-sided pair of arc vertical position welding root penetration process, set up the Controlling model of root penetration, and automatically adjust the technological parameter such as welding current, arc voltage, control the whole penetration of the root of weld, ensure that double sides double arc connects quality.
Below in conjunction with accompanying drawing, the present invention is described in further detail.
Accompanying drawing explanation
Fig. 1 is two-sided pair of arc vertical position welding near infrared vision sensing of the present invention and penetration control device schematic diagram.
Fig. 2 is visual pattern subregion, molten bath of the present invention contour images extracting method flow chart.
Fig. 3 is profile of the present invention and penetration aperture image aspects characteristic geometric dimension parameter extracting method flow chart.
Fig. 4 is welding condition Fuzzy Neural-network Control structure chart of the present invention.
Fig. 5 is two-sided pair of arc vertical position welding near infrared vision sensing of the present invention and penetration control method flow chart.
Detailed description of the invention
Below in conjunction with accompanying drawing, the present invention is described in further detail.
Composition graphs 1, strip 17 is first smoothed at leveling apparatus 14, limit is dug in bead cutter 15, three-point bending forming machine 16 is rolled up bending and becomes helix tube 1 to be welded, interior welding gun 9 is fixed on inside weld PTIG head 8, outer welding gun 11 is fixed on outside weld PTIG head 10, interior welding gun 9 is positioned at the inner side of helix tube 1 to be welded, outer welding gun is positioned at the outside of helix tube 1 to be welded, interior welding gun 9, seam center to be welded all aimed at by outer welding gun 11, and interior welding gun 9, outer welding gun 11 is coaxially arranged, the axis of the two overlaps with the normal of helix tube 1 contact to be welded, interior welding gun 9 arranges interior welding gun tungsten electrode 12, outer welding gun 11 arranges outer welding gun tungsten electrode 13, the dead in line of interior welding gun tungsten electrode 12 and outer welding gun tungsten electrode 13, interior welding gun 9 is connected near-infrared vision inner sensor 2, the axis of near-infrared vision inner sensor 2 and the axis of interior welding gun 9 are 35 ° ~ 45 °, outer welding gun 11 is connected near-infrared vision outer sensor 3, the axis of near-infrared vision outer sensor 3 and the axis of outer welding gun 11 are 35 ° ~ 45 °, and the axis of near-infrared vision inner sensor 2 and near-infrared vision outer sensor 3 is all through center, molten bath 7, near-infrared vision inner sensor 2 is arranged inner sensor camera lens 2-1, near-infrared vision inner sensor 2 and inner sensor camera lens 2-1 dead in line, near-infrared vision outer sensor 3 is arranged outer sensor camera lens 3-1, the dead in line of near-infrared vision outer sensor 3 and outer sensor camera lens 3-1, inner sensor camera lens 2-1 and outer sensor camera lens 3-1 is all apart from the surperficial 3-6cm in molten bath 7.In welding process, near-infrared vision inner sensor 2 and near-infrared vision outer sensor 3 are from helix tube 1 real time bidirectional continuous sensing acquisition molten bath 7 visual pattern, during welding, interior welding gun 9, outer welding gun 11 move upward relative to aluminum alloy spiral pipe 1, form vertical-up welding, near-infrared vision inner sensor 2, near-infrared vision outer sensor 3 servo follow-up tracing aim at molten bath 7.The two-way molten bath visual pattern that near-infrared vision inner sensor 2 and near-infrared vision outer sensor 3 detect is sent to computer 6 by shielded signal line 4 through image pick-up card 5.
Composition graphs 2, the two-way molten bath visual pattern being sent to computer 6 obtains the penetration contour images of molten bath visual pattern outline and central area, molten bath aperture after extracting through windowing, medium filtering, region segmentation, edge corrosion refinement, edge thinning and left and right contours extract, penetration district BORDER PROCESSING, penetration district boundary profile;
Composition graphs 3, carries out to molten bath penetration district center aperture area, penetration district each point tangent line, penetration district circularity penetration hole diameter d the geometric parameter calculating the penetration contour feature obtaining central area, molten bath aperture; Carry out calculating the geometric parameter obtaining molten bath outline feature to molten bath elemental area S, molten bath afterbody each point tangent line, rear towing angle β.Specifically comprise two-way pool width B, molten bath length L, penetration hole diameter d, molten bath area S, rear towing angle β, penetration aperture circularity Δ vwith the mean value s of aperture area v, maximum diameter d m, and d mthe maximum pool width B of homonymy m, minimum diameter d i, and d ihomonymy pool width B i;
Composition graphs 4, the Fuzzy-neural-network structure that the present invention adopts is: adopt five layers of fuzzy neural network, one to three layers of network realize the fuzzy reasoning of fuzzy control, two-layerly realizes de-fuzzy afterwards.Below every one deck of neutral net and input-output mappings relation are further explained.Following various in represent a jth neuronic output of the i-th layer network.
1st layer: input layer.Network input layer is pool width B, molten bath length L, penetration hole diameter d, molten bath area S, rear towing angle β, penetration aperture circularity Δ vwith the mean value s of aperture area v, use x respectively i(i=1,2 ... 7) represent, totally 7 nodes.This layer network neuron exports wherein x irepresent i-th input.
2nd layer: obfuscation layer.Obfuscation is carried out to input.For each input have 5 fuzzy sets S, SE, M, LE, L}={ " little ", " less ", " in ", " larger ", " greatly " corresponding with it.Fuzzy set membership function adopts triangular function this layer network neuron exports: represent the jth fuzzy set that i-th input is corresponding, for the membership function of corresponding fuzzy set, i=1,2 ... 7, j=1,2 ... 5.
3rd layer: rules layer.Degree of membership obfuscation obtained is multiplied between two, and its output represents the intensity of fuzzy rule.This layer network neuron exports O ij ( 3 ) = O 1 i ( 2 ) O 2 j ( 2 ) , ( i , j = 1,2 , . . . 5 ) .
4th layer: de-fuzzy layer.Adopt weight average de-fuzzy method here.This layer network neuron exports wherein be the weights of the 4th layer network, its meaning is the output membership function center of each control law.
5th layer: output layer.This layer network neuron exports wherein be the weights of the 5th layer network, represent welding current, torch swinging width, welding gun speed of welding respectively.
Composition graphs 5, carries out molten bath situation judgement according to two-sided pair of arc penetration judgment models and burn-through judgment models, if burn-through, then improve arc voltage and wire feed rate, welding gun flashes back and again welds and return step 1; If non-burn-through but through welding, the welding current that the fuzzy control method according to step 6 obtains, the optimum value of torch swinging width, adjustment twin arc welding connects electric current, torch swinging width continues weld and return step 1; If lack of penetration but at least one side there is penetration aperture, i.e. d>0, at least one side molten bath area is greater than critical value S k, then the welding current obtained according to the fuzzy control method of step 6, the optimum value of speed of welding, adjustment twin arc welding connects electric current, speed of welding continues weld and return step 1; If there is not penetration aperture, namely d be 0 or molten bath area S be less than critical value S k, then reduce speed of welding and again weld and return step 1.
Below in conjunction with embodiment, further detailed description is done to the present invention:
For the continuous welding of the above aluminum alloy spiral pipe 1 of 200mm diameter, thickness is that the strip 17 of 8mm is first smoothed at leveling apparatus 14, limit is dug in bead cutter 15, three-point bending forming machine 16 is rolled up bending and becomes helix tube 1 to be welded, interior welding gun 9 is fixed on inside weld PTIG head 8, outer welding gun 11 is fixed on outside weld PTIG head 10, interior welding gun 9 is positioned at the inner side of helix tube 1 to be welded, outer welding gun is positioned at the outside of helix tube 1 to be welded, interior welding gun 9, seam center to be welded all aimed at by outer welding gun 11, and interior welding gun 9, outer welding gun 11 is coaxially arranged, the axis of the two overlaps with the normal of helix tube 1 contact to be welded, interior welding gun 9 arranges interior welding gun tungsten electrode 12, outer welding gun 11 arranges outer welding gun tungsten electrode 13, the dead in line of interior welding gun tungsten electrode 12 and outer welding gun tungsten electrode 13, interior welding gun 9 is connected near-infrared vision inner sensor 2, the axis of near-infrared vision inner sensor 2 and the axis of interior welding gun 9 are 45 °, outer welding gun 11 is connected near-infrared vision outer sensor 3, the axis of near-infrared vision outer sensor 3 and the axis of outer welding gun 11 are 45 °, and the axis of near-infrared vision inner sensor 2 and near-infrared vision outer sensor 3 is all through center, molten bath 7, near-infrared vision inner sensor 2 is arranged inner sensor camera lens 2-1, near-infrared vision inner sensor 2 and inner sensor camera lens 2-1 dead in line, near-infrared vision outer sensor 3 is arranged outer sensor camera lens 3-1, the dead in line of near-infrared vision outer sensor 3 and outer sensor camera lens 3-1, inner sensor camera lens 2-1 and outer sensor camera lens 3-1 is all apart from the surperficial 5cm in molten bath 7, in welding process, near-infrared vision inner sensor 2 and near-infrared vision outer sensor 3 are from helix tube 1 real time bidirectional continuous sensing acquisition molten bath 7 visual pattern, during welding, interior welding gun 9, outer welding gun 11 move upward relative to aluminum alloy spiral pipe 1, form vertical-up welding, near-infrared vision inner sensor 2, near-infrared vision outer sensor 3 servo follow-up tracing aim at molten bath 7.Interior welding gun welding current is set to 130A, outer welding gun welding current is set to 220A, speed of welding is 0.24m/min, torch swinging width is set to 8mm and 9mm respectively, start welding, meanwhile, the two-way molten bath visual pattern that near-infrared vision inner sensor 2 and near-infrared vision outer sensor 3 detect is sent to computer 6 by shielded signal line 4 through image pick-up card 5.
The crater image information being sent to computer 6 obtains the penetration contour images of molten bath visual pattern outline and central area, molten bath aperture after extracting through windowing, medium filtering, region segmentation, edge corrosion refinement, edge thinning and left and right contours extract, penetration district BORDER PROCESSING, penetration district boundary profile; By calculating molten bath penetration district center aperture area, penetration district each point tangent line, penetration district circularity penetration hole diameter, molten bath elemental area, molten bath afterbody each point tangent line, rear towing angle β, obtain the geometric parameter of the penetration contour feature in central area, molten bath aperture and molten bath: two-way pool width B is respectively 8.5mm and 9.5mm, molten bath length L is 13mm, penetration hole diameter d is respectively 3.5mm and 3.2mm, molten bath area S is respectively 56.74mm 2and 70.88mm 2, rear towing angle β is 75 °, penetration aperture circularity Δ is 1, the mean value s of aperture area vbe respectively 9.62mm 2and 8.04mm 2, maximum diameter d mfor 3.5mm and d mwith the maximum pool width B of side direction mfor 9.5mm, minimum diameter d ifor 3.2mm and d ibe 8.5mm with side direction pool width B.
Molten bath situation judgement is carried out according to two-sided pair of arc penetration judgment models and burn-through judgment models, result is lack of penetration and penetration aperture all appears in two sides, adopt fuzzy Neural Network Control Method to adjust twin arc welding in real time and connect that electric current is respectively 135A, 220A, speed of welding is 0.24m/min, ensure two-sided pair of arc vertical position welding penetration quality.Continue Real-time Collection double sides double arc crater image information, judge molten bath situation, adopt fuzzy Neural Network Control Method to calculate the optimum value of two-way welding electric current, torch swinging width, speed of welding in real time, real-time adjustment two-way welding electric current, torch swinging width, speed of welding, ensure two-sided pair of arc vertical position welding penetration quality.
As from the foregoing, real time double face of the present invention detects vertical position welding molten bath visual pattern, by image procossing and penetration characteristic parameter extraction, obtain the information of two-sided pair of arc vertical position welding root penetration process, set up the Controlling model of root penetration, and adopt fuzzy-neural network method to adjust the technological parameters such as welding current, torch swinging width, speed of welding in real time, the whole penetration of the root of weld can be controlled, ensure that double sides double arc connects quality.

Claims (5)

1. two-sided pair of arc vertical position welding near infrared vision sensing and penetration control device, it is characterized in that, comprise near-infrared vision inner sensor (2), near-infrared vision outer sensor (3), shielded signal line (4), image pick-up card (5), computer (6), inside weld PTIG head (8), interior welding gun (9), outside weld PTIG head (10), outer welding gun (11), leveling apparatus (14), bead cutter (15), three-point bending forming machine (16);
Bead cutter (15) is positioned at the front of leveling apparatus (14), three-point bending forming machine (16) is positioned at the front of bead cutter (15), strip (17) digs limit after leveling apparatus (14) smoothing in bead cutter (15), is finally processed into helix tube (1) through three-point bending forming machine (16);
Interior welding gun (9) is fixed on inside weld PTIG head (8), outer welding gun (11) is fixed on outside weld PTIG head (10), interior welding gun (9) is positioned at the inner side of helix tube (1) to be welded, outer welding gun is positioned at the outside of helix tube (1) to be welded, interior welding gun (9), seam center to be welded all aimed at by outer welding gun (11), and interior welding gun (9), outer welding gun (11) is coaxially arranged, the axis of the two overlaps with the normal of helix tube (1) contact to be welded, interior welding gun (9) is arranged interior welding gun tungsten electrode (12), outer welding gun (11) is arranged outer welding gun tungsten electrode (13), the dead in line of interior welding gun tungsten electrode (12) and outer welding gun tungsten electrode (13), interior welding gun (9) is connected near-infrared vision inner sensor (2), the axis of near-infrared vision inner sensor (2) and the axis of interior welding gun (9) are 35 ° ~ 45 °, outer welding gun (11) is connected near-infrared vision outer sensor (3), the axis of near-infrared vision outer sensor (3) and the axis of outer welding gun (11) are 35 ° ~ 45 °, and the axis of near-infrared vision inner sensor (2) and near-infrared vision outer sensor (3) is all through molten bath (7) center, near-infrared vision inner sensor (2) is arranged inner sensor camera lens (2-1), near-infrared vision inner sensor (2) and inner sensor camera lens (2-1) dead in line, near-infrared vision outer sensor (3) is arranged outer sensor camera lens (3-1), the dead in line of near-infrared vision outer sensor (3) and outer sensor camera lens (3-1), inner sensor camera lens (2-1) and outer sensor camera lens (3-1) are all apart from molten bath (7) surperficial 3-6cm,
Near-infrared vision inner sensor (2), near-infrared vision outer sensor (3) are all connected with image pick-up card (5) by shielded signal line (4), by the Signal transmissions that detects to image pick-up card (5), this image pick-up card (5) is connected with computer (6) simultaneously.
2., based on a welding quality control method for two-sided pair of arc vertical position welding near infrared vision sensing and penetration control device described in claim 1, it is characterized in that, comprise the following steps:
Step 1, near-infrared vision inner sensor (2) and near-infrared vision outer sensor (3) real time bidirectional continuous sensing molten bath visual pattern, and by image transmitting to image pick-up card (5);
Molten bath visual pattern is transferred to computer by step 2, image pick-up card (5), carry out molten bath visual pattern subregion contour curve and characteristic parameter extraction, after windowing, medium filtering, edge corrosion micronization processes, obtain the penetration contour images of molten bath visual pattern outline and central area, molten bath aperture;
Step 3, the molten bath visual pattern outline extracted according to step 2 and the penetration contour images of central area, molten bath aperture, obtain the geometric parameter of the penetration contour feature of molten bath outline and central area, molten bath aperture, specifically comprise pool width B, molten bath length L, penetration hole diameter d, molten bath area S, rear towing angle β, penetration aperture circularity Δ vwith the mean value s of aperture area v, maximum diameter d m, and d mthe maximum pool width B of homonymy m, long, the minimum diameter d in molten bath i, and d ihomonymy pool width B i;
Step 4, set up double sides double arc and wear judgment models and two-sided pair of arc penetration judgment models;
Two-sided pair of arc penetration judgment models in step 5, foundation step 4 carries out molten bath situation judgement: if burn-through, then improve arc voltage and wire feed rate, welding gun flashes back and again welds and return step 1; If non-burn-through, judge whether through welding according to the penetration model in step 4, if through welding, the welding current that the fuzzy control method according to step 6 obtains, the optimum value of torch swinging width, adjustment twin arc welding connects electric current, torch swinging width; If lack of penetration but at least one side there is penetration aperture, i.e. d>0, and at least one side molten bath area is greater than critical value S k, then the welding current obtained according to the fuzzy control method of step 6, the optimum value of speed of welding, adjustment twin arc welding connects electric current, speed of welding; If there is not penetration aperture, namely d be 0 or molten bath area S be less than critical value S k, then reduce speed of welding and again weld and return step 1;
Step 6, adopt Fuzzy-neural-network control method to calculate the optimum value of welding current in two-sided pair of arc penetration weld termination process, torch swinging width, speed of welding in real time, accordingly electric current, torch swinging width, speed of welding are connect to twin arc welding according to the molten bath estimate of situation in step 5 and adjust when putting into effect and return step 1.
3. welding quality control method according to claim 2, is characterized in that, two-sided pair of arc penetration judgment models in step 4 is: bilateral penetration aperture maximum diameter d mbe greater than 3mm, homonymy pool width B mbe greater than 0.75h and be less than 1.2h, wherein h is thickness of slab, and unit is mm, homonymy pool width B m0.6 is greater than with the ratio of molten bath length L;
Double sides double arc wears judgment models: bilateral penetration aperture minimum diameter d ibe greater than 5mm, homonymy pool width B mbe greater than 1.1h, homonymy pool width B m0.9 is greater than with the ratio of molten bath length L.
4. welding quality control method according to claim 2, it is characterized in that, two-sided pair of arc penetration welding condition Fuzzy-neural-network control method in step 6 is: with the pool width B, the molten bath length L that obtain in step 3, penetration hole diameter d, molten bath area S, rear towing angle β, penetration aperture circularity Δ vwith the mean value s of aperture area vas the input variable of fuzzy neural network, set up five layers of fuzzy neural network, one to three layers of network realize the fuzzy reasoning of fuzzy control, and be input layer, obfuscation layer Sum fanction layer respectively, four layers realize de-fuzzy, and five layers is output layer; Wherein, in obfuscation layer, adopt triangular function to carry out obfuscation to input variable, transform to respective domain scope, obtain fuzzy variable and the fuzzy variable degree of membership of each input variable; In rules layer, the degree of membership that obfuscation obtains is multiplied between two, the fuzzy inference rule of Erecting and improving; In de-fuzzy layer, adopt weight average de-fuzzy method, the fringe of inference conclusion variable is converted into the amount of the determination state corresponding to network input value; Formula is utilized at output layer:
y i = Σ i = 1 2 ω i ( 5 ) O ( 4 )
Calculate and export determine state welding current, torch swinging width, speed of welding optimum value, wherein be the weights of the 5th layer network, O (4)be the output of the 4th layer network, y ibe the output of the 5th layer network, i is the neuronic number of layer 5.
5. welding quality control method according to claim 4, it is characterized in that, according to the molten bath estimate of situation in step 5, if through welding, the output layer parameter of two-sided pair of arc penetration welding condition Fuzzy-neural-network is the optimum value of welding current, torch swinging width; If lack of penetration but at least one side there is penetration aperture, i.e. d>0, at least one side molten bath area is greater than critical value S k, then the output layer parameter of two-sided pair of arc penetration welding condition Fuzzy-neural-network is the optimum value of welding current, speed of welding.
CN201310477515.6A 2013-10-12 2013-10-12 Two-sided pair of arc vertical position welding near infrared vision sensing and penetration control device and method Active CN103521890B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201310477515.6A CN103521890B (en) 2013-10-12 2013-10-12 Two-sided pair of arc vertical position welding near infrared vision sensing and penetration control device and method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201310477515.6A CN103521890B (en) 2013-10-12 2013-10-12 Two-sided pair of arc vertical position welding near infrared vision sensing and penetration control device and method

Publications (2)

Publication Number Publication Date
CN103521890A CN103521890A (en) 2014-01-22
CN103521890B true CN103521890B (en) 2015-09-30

Family

ID=49924437

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201310477515.6A Active CN103521890B (en) 2013-10-12 2013-10-12 Two-sided pair of arc vertical position welding near infrared vision sensing and penetration control device and method

Country Status (1)

Country Link
CN (1) CN103521890B (en)

Families Citing this family (23)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104874901A (en) * 2015-04-14 2015-09-02 招商局重工（江苏）有限公司 Double-side double-arc synchronous butt welding method of medium-thickness plate of sliding box of self-elevating drilling platform
CN104977305A (en) * 2015-06-29 2015-10-14 华中科技大学 Welding quality analysis device based on infrared vision and analysis method thereof
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
CN106001862B (en) * 2016-06-20 2018-09-18 江苏金鑫电器有限公司 A kind of aluminum alloy spiral pipe is flat to face upward composite welding process and its device
CN106624266B (en) * 2016-12-31 2018-08-07 东莞职业技术学院 A kind of weld seam deviation and penetration signal monitoring method for Automobile Welding
CN108941939B (en) * 2017-07-12 2020-04-07 温州大学激光与光电智能制造研究院 Closed-loop laser processing quality control method based on molten pool splash detection
CN107855687B (en) * 2017-10-17 2019-07-26 华中科技大学鄂州工业技术研究院 A kind of increasing material manufacturing fusion penetration on-line checking and control method and system
CN108856978B (en) * 2018-08-20 2020-08-07 南京理工大学 Angle joint penetration control method based on near-infrared binocular vision recognition
CN109128452A (en) * 2018-08-20 2019-01-04 南京理工大学 A kind of super-high strength steel arc-welding seam angle cognitive method and device
CN109409520B (en) * 2018-10-17 2021-10-29 深圳市微埃智能科技有限公司 Welding process parameter recommendation method and device based on transfer learning and robot
CN110238487B (en) * 2019-05-22 2020-09-22 北京航空航天大学 Medium plate aluminum alloy deep fusion welding system based on visual sensing and control method
CN111061231B (en) * 2019-11-29 2021-11-30 上海交通大学 Weld assembly gap and misalignment feed-forward molten pool monitoring system and penetration monitoring method
CN112122750B (en) * 2020-09-15 2022-04-29 天津大学 Circular seam butt welding method for seamless steel pipe
CN112561025B (en) * 2020-12-09 2022-10-14 安徽诚越电子科技有限公司 Method and device for prolonging service life of aluminum electrolytic capacitor
CN112705823A (en) * 2020-12-22 2021-04-27 吉林建筑大学 Welding method for assembled steel structure connecting node
CN112621006B (en) * 2020-12-30 2022-04-05 深圳艾灵网络有限公司 Welding monitoring method, system and related equipment
CN113028951A (en) * 2021-03-12 2021-06-25 烟台铁中宝钢铁加工有限公司 Detection sample plate, detection method and evaluation method of rack
CN113070618B (en) * 2021-05-19 2022-09-23 烟台大学 All-position back curtain type vertical welding device and welding method
CN113199184B (en) * 2021-07-05 2021-10-08 北京航空航天大学 Weld joint shape prediction method based on improved self-adaptive fuzzy neural network
CN113618204B (en) * 2021-07-29 2022-07-08 中建安装集团有限公司 K-TIG welding system and method
CN114273753B (en) * 2021-11-01 2023-01-13 天津大学 Method for monitoring filler wire GTAW-P transverse welding penetration degree
CN114951991A (en) * 2022-05-19 2022-08-30 哈尔滨焊接研究院有限公司 Ultrahigh-power scanning laser vertical welding method for aluminum alloy thick plate
CN116511652B (en) * 2023-06-30 2023-09-08 江苏永大化工设备有限公司 Gas shielded welding process parameter optimization system and method

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5990446A (en) * 1998-03-27 1999-11-23 University Of Kentucky Research Founadtion Method of arc welding using dual serial opposed torches
CN101062529A (en) * 2006-04-25 2007-10-31 南京理工大学 Weld seam appearance, fused bath and seam near-infrared vision integral sensing checkout gear
CN202701575U (en) * 2012-07-16 2013-01-30 中国化学工程第七建设有限公司 Double-faced dual-arc welding device of pipe root layer

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP4425690B2 (en) * 2004-04-28 2010-03-03 新日本製鐵株式会社 Method and apparatus for measuring outer peripheral length of spiral steel pipe, method for manufacturing spiral steel pipe and equipment therefor

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5990446A (en) * 1998-03-27 1999-11-23 University Of Kentucky Research Founadtion Method of arc welding using dual serial opposed torches
CN101062529A (en) * 2006-04-25 2007-10-31 南京理工大学 Weld seam appearance, fused bath and seam near-infrared vision integral sensing checkout gear
CN202701575U (en) * 2012-07-16 2013-01-30 中国化学工程第七建设有限公司 Double-faced dual-arc welding device of pipe root layer

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
刘鹤年.第11章 钢管.《建筑用钢》.2009,第417~418页. *
闫志鸿.P-GMAW熔池双面视觉传感与图像处理.《材料科学与工艺》.2007,第15卷(第5期),593-596. *

Also Published As

Publication number Publication date
CN103521890A (en) 2014-01-22

Similar Documents

Publication Publication Date Title
CN103521890B (en) Two-sided pair of arc vertical position welding near infrared vision sensing and penetration control device and method
Xu et al. Welding seam tracking in robotic gas metal arc welding
Xu et al. Application of sensing technology in intelligent robotic arc welding: A review
CN105478976B (en) Termination MICROBEAM PLASMA WELDING shaping control method based on Dynamic System Identification
CN108856978B (en) Angle joint penetration control method based on near-infrared binocular vision recognition
WO2022188482A1 (en) Active and passive vision combination-based welding deviation detection system and detection method
Zhang et al. Modeling of human welder response to 3D weld pool surface: Part I-principles
CN102950432B (en) Q245R vertical masonry joint submerged-arc welding natural gas vent pipe manufacturing method
CN105345235A (en) Cylindrical no-root full-penetration union melt welding method
CN104874901A (en) Double-side double-arc synchronous butt welding method of medium-thickness plate of sliding box of self-elevating drilling platform
CN105478975B (en) Termination MICROBEAM PLASMA WELDING shaping control method based on telecentricity visual sensing
Yu et al. Application of arc plasma spectral information in the monitor of Al–Mg alloy pulsed GTAW penetration status based on fuzzy logic system
CN109514043A (en) A kind of effective welding robot welding control system
CN115026483B (en) Boiler water wall pipe weld joint identification welding method
CN110193679A (en) A kind of appearance of weld control device and method
CN108098134A (en) A kind of new pattern laser vision weld joint tracking system and method
CN107610103A (en) A kind of steel pipe helical arc welded seam area extracting method based on x-ray imaging
CN107598401A (en) Plasma arc welding (PAW) electric current molten wide control system based on Fuzzy Reasoning Neural Network
CN108672893A (en) A kind of control method of asymmetry angle joint penetration form and fusion penetration
CN108152291A (en) End seam welding incomplete fusion shortcoming real-time detection method based on weld pool dynamics feature
CN109894779A (en) A kind of machine vision tracking system and method
Xiong et al. Feedback control of variable width in gas metal arc-based additive manufacturing
CN113551599A (en) Welding seam position deviation visual tracking method based on structured light guidance
CN102601512A (en) Automobile body spot welding splashing detection and control device and method thereof
JP7075311B2 (en) Welding control device, display control device, welding system, welding control method and program

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C14 Grant of patent or utility model
GR01 Patent grant
EE01 Entry into force of recordation of patent licensing contract

Application publication date: 20140122

Assignee: Jiangsu Jinxin Electrical Equipment Co., Ltd.

Assignor: Wang Xiaoyu|Li Mingjie

Contract record no.: 2017320000077

Denomination of invention: The two-sided pair of arc vertical position welding near infrared vision sensing and penetration control device and method

Granted publication date: 20150930

License type: Exclusive License

Record date: 20170313

EE01 Entry into force of recordation of patent licensing contract