JP2000024777A - Groove shape detecting device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To automatically detect the position of the welding bead in a multi- layer welding by transmitting each information on the detection starting command, the detected algorithm number, the welding torch angle and the welding torch position shift amount to an image processing unit by a welding control device, and detecting a groove shape in the welding pass based on the information stored in a detection condition table by the image processing unit. SOLUTION: A groove shape detection device comprising a welding torch 1, a light flooding means 6 and a two-dimensional light receiving means 10 is integratedly arranged with a welding torch position driving mechanism 18 movable in the vertical and horizontal direction orthogonal to the welding line. The b6-b8 out of b1 to b9 denote immediately preceding welding layer bead, and b9 denotes an immediately preceding welding bead. The groove shape detection device is arranged forward of the advancing direction of the welding torch 1, and a bead boundary between b9 and b7 is detected to effect the delay copying control. Thus, a groove 5 is detected with excellent accuracy, and the copy welding is effected with excellent accuracy even when the welding position deviation due to the working error, the assembly error or the thermal deformation is generated.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a process for capturing a reflection image from a groove surface by a camera when irradiating a slit-like light beam on the groove surface of a welding target and performing image processing on the image to obtain a position of a groove position of a portion to be welded. The present invention relates to an apparatus that detects a shape or the like and performs automatic welding using the detected information.

[0002]

2. Description of the Related Art As a conventional shape detecting apparatus of this type, a shape detector based on a light section method using a slit light beam is known, for example, from Japanese Patent Application Laid-Open No. Hei 5-296734. According to the present invention, a groove surface is irradiated with slit light, a light-section image at this time is captured, and image processing is performed to detect a groove dimension.

In addition, the same light cutting method is used, but the reflected light image obtained by oscillating the laser light from the laser oscillator and irradiating the groove surface is taken and subjected to calculation processing to calculate the distance to the work. A method is known from JP-A-2-276907.

Japanese Patent Application Laid-Open No. 3-47680 discloses two types of information, that is, information on the inside shape of a groove by a laser displacement sensor disposed in front of a welding torch, information on the left and right ends of a groove by an ITV camera, and information on the position of a welding wire tip. The scanning control of the welding torch position is performed based on the type of detection information. This reference proposes a method of indirectly detecting a relative position between a welding bead intersection and a welding wire tip position from two types of information obtained by using two types of detecting means.

[0005]

In multi-pass welding, since the number of passes of a weld bead is large due to multi-pass multi-pass welding, a welded work is deformed due to thermal distortion or the like due to welding. For this reason, in order to form a high-quality weld bead, it is necessary to accurately measure the position to be welded even when the welded work is deformed, and to accurately follow the welding torch for welding.

Normally, in order to obtain a high quality bead in multi-layer welding, the position of the tip of the welding wire is accurately determined at a target position determined on the basis of the groove position immediately before the welding bead is formed for each welding pass. It is necessary to follow the welding.

On the other hand, a conventional optical detection method (Japanese Unexamined Patent Publication No.
JP-A-96734, JP-A-2-276907 and JP-A-3-4
In Japanese Patent No. 7680), an optical shape detecting means is disposed integrally with a welding torch to detect the shape and position of a groove before welding. In an actual welding operation, welding is performed at an optimum angle of the position of the welding torch with respect to the groove. A change in the position of the welding torch also changes the detection position of the optical detection means, and as a result, the resulting light-section image also changes.
In multi-layer welding, not only the groove cross-sectional shape at the time of initial layer welding and at the time of bead lamination is different, but also the groove position detection point for performing welding profiling control is different for each welding pass.

For example, when performing one-layer multi-pass welding, a light-cut image obtained by welding on the right side or left side of the welding progress direction or by welding a boundary position between beads and a position to be detected changes. I do. Further, there is a case where welding is performed by shifting a target position of a welding torch position by a predetermined value with respect to an actual groove position detected to secure welding quality.
If the light-section image changes for each of these welding passes, a method of predicting in advance the light-section image obtained by optical means is not disclosed in the above-mentioned known example. The conventional method has a problem that image processing cannot be stably and accurately performed on a groove whose shape pattern changes.

As a means for solving the above problem, a method is used in which a welding operator visually judges a boundary position of a weld bead by visually observing a groove surface and performs welding manually, but requires skill or is quantitative. There is a practical problem that measurement cannot be performed with high accuracy. In addition, a method in which the welding from the first layer welding to the final pass welding is planned by a computer in advance and the welding is controlled based on the plan is often adopted. There is a problem that there is a limit to the welding work that can perform multi-layer welding by this method, for example, it is necessary to set in advance, or to accurately predict thermal deformation due to welding in advance.

[0010]

SUMMARY OF THE INVENTION It is an object of the present invention to advantageously solve the above-mentioned problems and to provide a method for automatically detecting a weld bead position in multi-pass welding.

In order to achieve the above object, a welding position detecting method of the present invention irradiates a slit-like light from a light projecting means to a surface of an object to be welded, and captures reflected light obtained at this time by a light receiving means. Then, an optical sensor that processes a captured image to detect a position shift, detection information that is detected by an image processing device that processes a groove image obtained from the optical sensor, and control of a welding head and a welding torch. In a control method for performing multi-pass welding based on calculation information processed by a welding control device capable of performing output control of welding conditions, the welding control device sends a detection start command to the image processing device,
The image processing apparatus transmits each information of a detection algorithm number, a welding torch angle, and a welding torch position shift amount which are previously determined between the welding control device and the image processing device, and the image processing device stores the information in a detection condition table, and stores the information in the detection condition table. The groove shape is detected based on the above information.

[0012]

DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the present invention will be described below with reference to FIGS.

FIG. 1 is a schematic diagram showing an embodiment of a welding position detecting device and an automatic welding device using the same according to the present invention. In the figure, 1 is a welding torch, 2 is a welding wire, 3 and 4 are members to be welded, 5 is a welding groove (hereinafter, a groove including a welding bead is referred to as a groove), and 6 is a slit-shaped light beam 7. The light emitting means 10 for irradiating is a light receiving means.
The light receiving means 10 includes an interference filter 8 and a two-dimensional light receiving camera 9 such as an ITV. The light projecting means 6 adjusts the angle and position of the light projecting means 6 so that the groove shape of the welded members 3 and 4 irradiates the slit light 7 to the groove 5 of the V-shaped butt. The light cutting line Q obtained at that time is observed (imaged) by the light receiving means 10 from above the groove. Reference numeral 16 denotes a control circuit of the light projecting means 6. 17 is a two-dimensional light receiving camera 9
And outputs an analog video (image) signal to the outside.

Welding torch 1, light emitting means 6, and light receiving means 10
Are arranged integrally (not shown), and can move freely above the groove 5 by driving the welding torch position control mechanism 18. In the figure, reference numeral 20 denotes an image processing apparatus comprising the following parts. Reference numeral 21 denotes an image input unit that A / D converts an analog image signal output from the control circuit 17 into a digital amount and outputs multivalued image data, 22 denotes a multivalued image storage unit that stores multivalued image data, 23 Is a multivalued image storage unit 22
A binarization processing unit that extracts only a bright part in the image from the multivalued image data stored in the binarization processing unit; a binary image storage unit that stores a binary image obtained by the binarization processing unit 23;
Is an external device control unit, 26 is an external detection condition storage unit that stores detection condition information of a detection algorithm number, a welding torch angle, and a welding torch position shift value transmitted from the overall control device 13, 27 is a binarization, Or a storage unit for various parameters necessary for detecting a groove position, etc., 28
Is a line element data storage unit for storing line element data calculated based on the light-section image, 29 is an arithmetic processing unit for binarizing, extracting line elements, or finding a welding bead position or the like from the line element data, 30 is It is a main control section that controls the above sections in an integrated manner. 11, a welding power source; and 12, a welding torch position control device that drives and controls the welding torch position control mechanism 18. 1
3 is an image processing device 20, a welding torch position control device 1
2, and an overall control device for controlling the welding power supply 11 in a comprehensive manner.

FIG. 2 is a flowchart showing an embodiment of a groove shape detecting method according to the present invention. Hereinafter, the contents of each processing step in FIG. 2 will be described with reference to FIGS. 3 to 11.

First, in step S1, the image processing device 2
The groove light cutting image obtained from the light receiving means control circuit 17 is input by the image input unit 21 of 0 and stored in the multi-value image storage unit 22. FIG. 3 is a schematic diagram illustrating an example of a method of laminating a multi-layer weld at a butt groove. In this example, the first layer (initial layer) is a case of one-pass welding, the second and third layers are two-pass welding, and the fourth layer is a case of three-pass welding. FIG. 4 shows each detection condition information of a detection algorithm number, a welding torch angle, and a welding torch position shift amount (ΔY, ΔZ) transmitted when the overall control device 13 gives a detection command to the image processing device 20. Things.

The welding torch position shift amount is a movement amount of a target position of the tip of the welding torch with respect to an actual groove position detected by the groove shape detecting device. The image processing apparatus stores such information in the external detection condition storage. The overall control device 13 manages the entire construction of the multi-layer welding, and instructs the image processing device 20 as to which detection algorithm number which has been negotiated in advance with the image processing device to detect the groove shape. give.

FIG. 5 shows a two-dimensional light receiving camera 9 at the time of first layer welding.
3 shows an example of a grooved light section image captured by the multi-value image storage unit 22. In the figure, black and white are reversed, that is, a bright (high luminance) portion is expressed as black, and a dark (low luminance) portion is expressed as white, contrary to a normal image. Ql and Q
R is a primary reflection image obtained when a laser beam is applied to the groove surface, and Q1, Q2, Q3, and Q4 are unnecessary secondary reflection images (light irradiated to the groove surface is irradiated again to another groove surface). Image that is taken and imaged). QL, QR, Q1, Q2, Q3
The brightness (brightness) of each of the reflected images of Q4 and Q4
It is assumed that L, QR, Q1 and Q2 are sufficiently bright, and Q3 and Q4 are dark. Hereinafter, the processing contents, methods, results, and the like in the other processing steps will be described with reference to the light section image shown in FIG. 4 and an example in which this image is processed.

In step S 2, the groove light section image stored in the multi-level image storage section 22 is binarized by the binary image processing section 23, and this data is stored in the binary image storage section 24. FIG. 6 is an example of an image obtained by binarizing the image of FIG. 5, and is an example in which only QL, QR, Q1 and Q2 images are extracted. In the present invention, as a method of binarizing a multi-valued image, a fixed threshold value method or another method may be used.

In step S3, the binary image data is used to determine both edge portions of the light cutting line in the vertical direction of the screen of the binary image data, that is, the boundary position between the light cutting line portion and the background (a portion where no image exists). Then, the center position of the boundary on both sides is calculated. While sequentially moving this to the right in the horizontal direction of the screen, calculation is performed for each vertical line to obtain line data (referred to as a line image) indicating the center position.
Next, a linearization process is performed on the line image. In this linearization process, for example, a line image such as a bead surface including a curve as described later is divided into a plurality of linear line element groups,
Label (number) each divided line element,
A quantity representing a feature of each labeled line element (hereinafter, referred to as a feature quantity) is calculated. FIG. 7 schematically shows an example of a result of performing the linearization process on the binary image of FIG. 6 by the process of step S3. In the illustrated example, a total of four line elements are obtained.

In step S4, an evaluation / correction process is performed on the line element of FIG. 6 obtained in step S3. Part 1
First, processing for deleting unnecessary line elements is performed. As shown in the example of FIG. 1, images 1 and 2 (QL and QR in FIG. 6) of the light section line from the groove in FIG. 7 are obtained as substantially continuous line images. Speaking on the screen, the light cutting line is continuously generated in the horizontal direction of the screen. An image generated out of the precondition is unnecessary noise and is deleted. Here, the unnecessary line elements are, for example, the line elements 3 and 4 in FIG.
Here, it is a line element (secondary reflection image) other than the primary reflection image when the laser beam is irradiated on the groove surface. In this process, unnecessary line elements are deleted, and the line element numbers and their line element data tables (not shown) are rearranged.
FIG. 8 schematically shows the result obtained by this processing. In the illustrated example, only a total of two line elements are obtained.

In the next step S5, the image processing device 20
The detection algorithm number, the welding torch angle, and the welding torch position shift amount stored in the external detection condition storage unit 26 are acquired. The type of algorithm to be used to detect the groove shape is determined based on the detection algorithm number. The detection processing programs corresponding to the respective detection algorithm numbers are incorporated in the image processing apparatus in advance.
The welding torch angle and the welding torch position shift amount are auxiliary data when executing the detection processing.

The detection algorithm numbers in the above examples are for the detection of the groove shape of the first layer. In this case, as an example, the corner of the groove is the detection point for welding scanning. Therefore, the processing in the next step S6 is
The intersection coordinates C0 (Uc0, Vc0) of the straight lines are detected from the line elements 1 and 2 obtained by the processes up to 4.

The weld bead boundary position determined in step S6 is a detection result on the two-dimensional light receiving means, that is, on the camera screen. Detection position obtained on this screen
It is necessary to convert (U, V) from sensor constants (Y, Z) (in other words, work coordinate positions viewed from the sensor) from various constants such as the mounting position and angle of the optical system or the imaging magnification of the camera. There is. The processing in step S7 performs this coordinate conversion.

In the last step S8, it is evaluated whether the detected bead boundary position is appropriate. When the result of the evaluation is determined to be invalid (detecting an erroneous position or the like), error processing is performed, and the processing ends. Regarding the evaluation of the recognition result, for example, there is a method of making an error when the detected corner position has an abnormally large value, but the method is not particularly limited to this method.

The detection algorithm number 1 has been described above, that is, an example of detection at the first groove is described. Next, another detection algorithm number will be described.

FIG. 9 is a schematic diagram of a groove light section image captured by the two-dimensional light receiving camera 9 and stored in the multi-value image storage unit 22 at the time of the detection algorithm number 2, that is, the second pass welding of FIG. In this case, the intersection point L1 of the image QL of the slope on one side and the reflection image QB from the front pass bead surface is detected. The optical groove detecting means including the light projecting means and the light receiving means and the welding torch are arranged integrally. Therefore, it is possible to predict in advance the shape pattern of the light-cut image obtained when the detection algorithm number, the welding torch angle and the welding torch position shift amount shown in FIG. 4 are known, and the rotation and arrangement on the screen. It is. Therefore, it is possible to accurately perform the groove shape feature amount detection processing in the processing step S6 of FIG. 6 described above.

FIG. 10 is a schematic view showing a groove cutting light image captured by the two-dimensional light receiving camera 9 at the time of the third detection welding in FIG. In the figure, QB1 is an image of the bead surface after one-pass welding, and QB2 is an image of the bead surface after two-pass welding. In this case, the images QR and 1
The intersection R1 of the reflection image QB1 from the pass bead surface is detected.

FIG. 11 shows detection algorithm number 4
That is, a schematic diagram of a groove light section image captured by the two-dimensional light receiving camera 9 at the time of the seventh pass welding shown in FIG. In the figure, QB45 is the former layer 4,5.
QB6 is an image of the bead surface after two-pass welding. In this case, the front bead Q45
And the boundary portion C4 of the previous pass bead QB6.

FIG. 12 shows a case where the boundary between weld beads is detected, and the shift amount (Δ
An example of a light-section image obtained when Y) is given is shown. In the figure, the solid line QB is the light section image. Two locations, Cn1 and Cn2, are candidates for detecting a boundary point between a weld bead and a bead. It is not clear from this image which point to detect. However, based on the value of the shift amount (ΔY), the light-section image and the boundary position when there is no shift amount are the dotted line portion and the Cn point (the Cn point is a point slightly shifted from the center by ΔU) in FIG. Can be said in reverse. In other words,
From the value of the shift amount (ΔY), Cn1 may be selected and determined from two welding bead-bead boundary point detection candidates, and this point may be set as the groove position.

The above description shows a method for detecting a groove position based on a detection algorithm number. In the present invention,
The overall control device may transmit the current welding path instead of the detection algorithm number. In this case, in the image processing device, a correspondence table between the welding path and the detection algorithm number is created in advance, and the detection algorithm number is determined with reference to the correspondence table to perform the detection processing.

Also, in the present processing procedure, the case where the image in which the groove light section image extends in the horizontal direction of the screen is input is described. By performing the processing in reverse, the bead boundary position can be similarly detected.

As described above, the groove shape input method in the embodiment of the present invention has been described in connection with the case where the image is taken by the light projecting means for irradiating a slit-like light beam and the two-dimensional light receiving camera such as an ITV.
A method of scanning a laser beam to capture reflected light with a line sensor and detecting the groove shape based on the principle of triangulation can also detect the groove shape in the same manner as described above.
As described above, any method may be used as long as it is a method of inputting surface data of a groove shape based on the principle of triangulation, and the present invention is not particularly limited to a method of inputting a surface shape of a groove shape.

Next, an automatic welding method to which the present invention is applied will be described. In FIG. 1, welding torch 1 and light emitting means 6
The groove shape detecting device of the present invention comprising the light receiving means 10 and the light receiving means 10 is disposed integrally with a drive control mechanism 18 which is movable in the vertical and horizontal directions perpendicular to the welding progress direction (welding line) and the welding line. I have. The present automatic welding device includes welding members 3 and 4
Are joined together by multi-layer welding. In the drawing, b1 to b9 are already welded beads. Among them, b6, b7 and b8 are pre-weld layer (one-layer three-pass) beads. b9 is the immediately preceding weld bead. The figure shows an example of a case where the boundary between the bead b9 and the bead b7 is automatically welded. The present automatic welding device arranges a groove shape detection device in front of the welding torch 1 in the traveling direction and detects a bead boundary position between the bead b9 and the bead b7 at the groove portion, thereby setting the welding torch position to the bead boundary position. Performs delayed copying control.

FIG. 13 shows a flowchart of the welding torch position follow-up control operation in a case where a groove shape (position) is detected by using a groove shape detecting means and an image processing device. In this operation, the overall control device 13 becomes a master station for managing the entire operation, and the welding power source 11, the welding torch position control device 12 and the image processing device 20 are in a slave station relationship. In other words, the welding power source 11, the welding torch position control device 12, and the image processing device 20 of the slave station execute a predetermined operation according to a command from the overall control device 13 as the master station.

After the start of the operation, in step F1, the general control device inquires the current position of the electrode to the welding torch position control device, and reports the position information (step F1).
7) Receive.

In step F2, the general control device issues a movement command for raising the welding torch to the electrode position control device, and raises the electrode (step F18).

In step F3, the general control device issues a movement command to the reference position at which welding is started to the welding torch position control device, and moves the electrode to the reference position (step F19).

In step F4, the general control device issues a command to the welding torch position control device to move the welding torch position to a predetermined height at which the arc can be started, and lowers the welding torch to the predetermined position (step F20).

At step F5, the general control device issues an arc start (ON) command to the welding power source, and the arc ON is started.
Is started (step F24).

In step F6, the overall control device issues a command to start traveling in the X direction to the welding torch position control device, and starts traveling in the X direction (step F2).
1). The traveling operation in the X-direction traveling operation is continued until a traveling stop command described later is issued.

In step F7, the general control device issues a detection command to the image processing device, and the image processing device executes the groove shape (position) detection in the above-described procedure (step F7).
26). At the time of issuing the detection command, a detection algorithm number previously determined between the welding control device and the image processing device,
Each information of the welding torch angle and the welding torch position shift value is transmitted. The image processing apparatus stores the information in the detection condition table, and starts detecting the groove shape based on the information in the detection condition table.

In step F8, the overall control device inquires the current position of the electrode position to the welding torch position control device, and receives a report of the position information (step F22).

In step F9, the general control device issues a command to report a groove shape (position) detection result to the image processing device, and the image processing device reports the result (step F2).
7) Receive.

In step F10, groove shape (position) detection result information is stored internally.

In step F11, the welding torch tracing position is calculated based on the groove shape (position) detection information. here,
The amount of position correction in the Y travel axis direction can be calculated using multiple groove shapes (positions).
The detection information may be obtained by averaging, and there is no particular limitation.

In step F12, the information calculated in step F11 is transferred to the welding torch position control device to perform control for correcting the position of the Y traveling bogie (step F23).

In step F13, it is determined whether or not the welding operation is at the end position based on the welding torch current position information, and the electrode position is still at the end position (it is determined and stored in advance in the overall control device). If not, the process returns to the processing step F7 to repeat the above operation. On the other hand, when it is determined that the electrode position has exceeded the end position, the process proceeds to processing step F14.

In step F14, the general control device issues an arc end (OFF) command to the welding power source, and the arc is turned on.
Is stopped (step F25).

In step F15, the general control device issues a command to stop traveling in the X direction to the electrode position control device, and stops traveling in the X direction.

In step F16, it is determined whether or not the welding of all the welding passes planned in advance has been completed. If it has been completed, a series of operations is completed, but if not, the process proceeds to step F2.

Although the welding end position in step F16 is determined and stored in advance in the overall control device, it may be determined by detecting the position with a limit switch or the like.

In the present invention, although not shown in FIG. 12, when the next welding pass is to be welded, the welding conditions in the past welding pass such as immediately before, that is, the welding torch scanning correction control data described above are used. You can also do it. Whether the welding of each welding pass is performed using the previous welding condition control data or the content of each processing step described above is to be incorporated in advance as a database in the overall control device.

The drive control mechanism 18 of the present invention is, for example, a self-propelled welding robot, an articulated welding robot that travels independently on a plane, or a robot that travels on a travel axis provided on the outer periphery of a pipe. is there.

The above is an example of an automatic welding apparatus in which the present invention is applied to the detection of the shape of a welded work. However, the present invention is not limited to this. Applicable to

[0056]

According to the object shape detecting method and the automatic welding apparatus of the present invention, the shape pattern of the light cutting image obtained from the information of the detection algorithm number, the welding torch angle and the welding torch position shift amount, and the rotation on the screen. Since the arrangement can be predicted in advance, it is possible to accurately detect the groove shape (position). As a result, profile welding can be performed with high accuracy even when a welding position shift caused by a processing error or an assembly error of a joint portion or a thermal deformation of a work occurs.
A high-quality weld bead can be obtained without depending on the skill of the operator. Also, since the groove position, the boundary position between the weld bead and the weld bead, or the boundary position between the weld bead and the groove can be detected, the welding torch position profiling welding can be performed on a wide variety of grooves including a weld bead such as multi-pass welding. Is possible. Further, since the profile welding operation can be automatically performed using the object shape detecting means and the drive control mechanism, unmanned welding can be performed.

[Brief description of the drawings]

FIG. 1 is a schematic configuration diagram showing an embodiment of a welding position detecting device of the present invention and an automatic welding device using the same.

FIG. 2 is a flowchart showing an embodiment of a welding torch position shift detecting method according to the present invention.

FIG. 3 is a schematic diagram of a lamination method of multi-layer welding at a butt groove.

FIG. 4 is a view showing detection condition information of the present invention.

FIG. 5 is a diagram showing an example of a groove light section image according to the present invention.

FIG. 6 is a view showing an example of a groove light section image according to the present invention.

FIG. 7 is a diagram showing an example of a groove light section image according to the present invention.

FIG. 8 is a diagram showing an example of a groove light section image according to the present invention.

FIG. 9 is a diagram showing an example of a groove light section image according to the present invention.

FIG. 10 is a diagram showing an example of a groove light section image according to the present invention.

FIG. 11 is a diagram showing an example of a groove light section image according to the present invention.

FIG. 12 is a diagram showing an example of a groove light section image according to the present invention.

FIG. 13 is a flowchart of welding torch position shift detection and welding torch copying control operation of the present invention.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Welding torch, 2 ... Welding wire, 3, 4 ... Member to be welded, 5 ... Groove, 6 ... Light emitting means, 7 ... Slit light, 8 ... Interference filter, 9 ... ITV camera, 10 ... 2D light receiving means , 11 ... welding power source, 12 ... welding torch position control device,
13: Overall control device, 16: Light emitting means control circuit, 17:
Light receiving means control circuit, 18: welding torch position drive mechanism, 2
0: Image processing device.

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Mitsuaki Haneda 502, Kunitachi-cho, Tsuchiura-shi, Ibaraki Pref.Mechanical Research Laboratories, Inc. Inside the Machinery Research Laboratory (72) Inventor Yu Aratani 502 Kandachi-cho, Tsuchiura-city, Ibaraki Pref. Inside the Machine Research Laboratories, Hitachi, Ltd. Terms (reference) 2F065 AA12 AA22 AA45 BB17 BB18 BB24 CC27 DD03 FF02 FF04 GG04 HH05 HH12 JJ03 JJ16 JJ26 QQ03 QQ04 QQ23 QQ24 3C029 BB01 BB10

Claims (1)

[Claims] An optical sensor comprising a welding head, an irradiating means for irradiating a slit-like light to the surface of a welding work, and a two-dimensional light receiving means for imaging a reflected image thereof, and a groove obtained by the optical sensor. It consists of an image processing device that processes images, and a welding control device that can control the welding head and output the welding conditions to the welding torch. The groove information detected by the image processing device and the arithmetic processing by the welding control device In the apparatus for controlling the multi-pass welding based on the calculation information to be performed, the welding control device sends a detection start command to the image processing device,
The image processing apparatus transmits each information of a detection algorithm number, a welding torch angle, and a welding torch position shift amount which are previously determined between the welding control device and the image processing device, and the image processing device stores the information in a detection condition table, and stores the information in the detection condition table. A groove shape detecting device for detecting a groove shape in a corresponding welding pass based on the information of the groove shape.