JPH07299565A - Visual sensor for welding and welding control method - Google Patents

Abstract

PURPOSE:To obtain a clear picture so as to execute high precision weld line copying control by arranging a CCD camera, interference filter, pin hole and applying air or gas pressure from the inside and outside of pin hole in the visual sensor for welding. CONSTITUTION:In front of a camera 1, a lens 3 protected through a first casing 2a, an interference filter 4 to transmit a wave length having 0.6-0.8mum center wave length with <=0.1mum width range and a protective glass 5 are successively arranged. In front of the protective glass 5, a pin hole 8 protected through a second casing 2b and pierced with <=2PHImm hole is arranged. Removal of a sputter 7 and fume 12 in the visual sensor is executed so that the air introduced from an air introducing pipe 10 applies internal pressure to a hood 9 and is discharged from the hole pierced to the pin hole 8, the intrusion of sputter 7 and fume 12 from this hole is prevented.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a visual sensor for welding and a welding control method, and particularly to remote monitoring of welding conditions for GMA welding, or image processing based on the obtained image to perform automatic welding and unmanned welding. The present invention relates to a welding visual sensor used as a purpose.

[0002]

2. Description of the Related Art Conventionally, as a sensor for monitoring the welding condition, there are a method using an infrared camera and a method combining a CCD camera and an interference filter. In addition, as a method of copying the groove line by image processing, it is assumed that the wire insertion position is from the position of the depression formed in the molten pool or arc shape or the upper part of the arc to the center of the molten pool or the center of the arc. The groove line was traced by using the deviation of G as a control amount.

[0003]

With the infrared camera method, the shape of the molten pool could be obtained, but other conditions inside the groove could not be recognized by the infrared camera, and humans could I couldn't get enough information from it to monitor. In addition, the intensity of welding light is strong only with a CCD camera (with a lens, of course), and it is impossible to obtain an image capable of monitoring the welding situation due to halation or blooming, and a person is welding. A visual sensor has been developed and put into practical use, in which an interference filter is placed in front of a CCD camera as a light-shielding filter used at times and a more vivid image is displayed.

Further, in the GMA welding, spatter and fume are often generated, and it is necessary to frequently replace glass or the like to protect the lens, which is an obstacle to automation. Furthermore, although these sensors are used as a groove line copy, as a control method, the amount of deviation of the wire insertion position from the center of the molten pool or the center of the arc is calculated, and control is performed to eliminate this deviation. . However, in this case, it was not possible to provide high-quality welding due to the time delay and the bead shape that was placed before.

The present invention has been made in consideration of such circumstances, and it is possible to achieve a long service life in a bad welding environment and obtain a clear image and perform highly accurate welding line copying control. Another object of the present invention is to provide a visual sensor for welding and a welding control method capable of judging the welding condition from the shape of the molten pool and feeding back from this shape to the welding condition to perform higher quality welding.

[0006]

SUMMARY OF THE INVENTION A first invention of the present application is a welding visual sensor used for remotely monitoring an electrode consumable welding or automatically controlling a monitor image thereof by a pixel processing device or the like. , CCD camera, 0.6 ~
An interference filter having a center wavelength of 0.8 μm and a wavelength in the range of 0.1 μm or less is provided, and a pinhole provided in front of the interference filter and having a hole of 2 φmm or less is provided. The visual sensor for welding is characterized in that pressure is applied by air or gas from both inside and outside.

According to a second aspect of the present invention, the welding visual sensor and an image processing device connected to the sensor are used to detect the wire insertion position, the inclination of the front bead shape, and the moving speed of the torch shaft in the axial direction across the groove line. The welding control method is characterized in that fuzzy inference is performed using the direction as a parameter, the moving speed and moving direction of the next torch axis or the torch axis position is determined, and the groove line copying is performed.

[0008]

According to the present invention, by adopting the interference filter in the near infrared region, the arc light is reduced and the molten pool can be seen by approaching the light emitting region of the molten pool more, and it is not visible in the infrared region. It becomes possible to obtain more information in the groove such as the bead state and arc shape in front of, and by placing a pinhole in front of it and applying pressure with air or gas before and after that, the path of spatter and fume The lens and interference filter can now be used without damaging them for a long time.

Further, the "welding line copying" includes the shape of the front bead in the item to be judged, so that if the front bead is high on the left, for example, the welding torch is moved slightly to the right, and if the right bead is high, the welding torch is moved to the left. , The finished bead can be piled up flatter, and it is possible to operate by adapting to the shape of the front bead. Further, by using the torch moving speed and its direction as input parameters, for example, if the deviation is slightly right when the welding torch is moving left, the welding torch is stopped.
If the offset is slightly to the left, move the welding torch to the right at a slightly faster rate. This is fuzzy inferred to determine the next torch movement speed and direction. This enables more accurate torch positioning.

[0010]

DESCRIPTION OF THE PREFERRED EMBODIMENTS A visual sensor for GMA according to an embodiment of the present invention will be described below with reference to FIG. Reference numeral 1 in the figure is a CCD (Charge Coupled Device) camera. In front of the camera 1, the lens 3, the interference filter 4, and the protective glass 5 which are protected by the first housing 2a.
Are provided in sequence. Here, the protective glass 5 is
This is installed to protect the interference filter 4 or the lens 3 considering that the spatter 7 has just entered through the pinholes described later due to insufficient pressure of the air 6a for applying the internal pressure. Either of the lens 3 and the interference filter 4 may be arranged back and forth.

In front of the protective glass 5, there is provided a pinhole 8 spaced apart from the protective glass 5 and protected by the second housing 2b. Here, the second housing 2b
The end portion of is fitted into the outer peripheral portion of the front end of the first housing 2a. A third casing 2c is fitted on the outer peripheral portion of the second casing 2b. A half-split hood 9 is attached to the tip of the third housing 2c. The second
The casing 2b and the third casing 2c are provided with an air introduction pipe 10 for introducing air into a region surrounded by the protective glass 5, the pinhole 8 and the second casing 2c. The third housing 2
c is provided with an air introduction pipe 11 for introducing air into a region surrounded by the second casing 2b, the third casing 2c and the hood 9 and blowing the air from the front of the pinhole 8. There is. In the figure, reference numeral 12 indicates fume and reference numeral 6b indicates air.

In the GMA visual sensor having such a structure, the spatter 7 and the fumes 12 are removed by the air 6a introduced from the air introduction pipe 10 being applied with the internal pressure of the hood 9 and discharged from the hole formed in the pinhole 8. The spatter 7 and fume 12 are prevented from entering through this hole. In addition to this, air 6b is removed from the welded portion in order to remove the adhesion of the pinhole 8 to the entire periphery of the hole and to prevent the air 6a from the pinhole 8 from destroying the gas atmosphere of the welded portion. It is introduced from the introduction pipe 12. The air 6b is the air introduction pipe 12
The air 6b is introduced from above and blows out the air 6b on the upper side, and the trajectory of the air 6a is also changed upward by this momentum. The fume 12 also has air 6 in front of the camera.
Since it is excluded by a and 6b, an open view can be obtained.

As described above, the visual sensor for GMA according to the above-described embodiment has the CCD camera 1 and the lens 3 provided in front of the camera 1 and protected by the first casing 2a.
The interference filter 4 and the protective glass 5, and the protective glass 5
The pinhole 8 provided in front of the second housing 2b and protected by the second housing 2b, the air introducing pipe 10 for introducing the air 6a from the inside of the pinhole 8, and the air outside the tip of the second housing 2b. It has a configuration including an air introduction pipe 11 for introducing 6b. Therefore, by using the GMA visual sensor according to the above-described embodiment, it is possible to obtain a clear image having environmental resistance without being affected by the spatter 7 and the fumes 12.

Next, an example of performing the groove copying using the GMA visual sensor 13 of FIG. 1 will be described with reference to FIGS. First, the configuration will be described. An image processing device 21 for processing an image situation and a TV monitor 22 for checking an image are sequentially connected to the GMA visual sensor 13. Here, the insertion position TC of the welding wire 24, the contour of the arc 25, and the contour of the molten pool 26, which emerge from the lower portion of the welding torch 23, are obtained by the processing by the image processing device 21. Reference numeral 27 in the drawing denotes a torch left-right scanning shaft, to which a control device 28 connected to the image processing device 21 and a welding power source 29 connected to the image processing device 21 are attached.
Are connected in sequence. Reference numeral 30 in the drawings is a welding member, reference numeral 31 is a groove surface, reference numeral 32 is a bead before welding, and reference numeral 33 is a bead after welding.

Next, the operation will be described. The GMA
The visual sensor for use 13 is arranged in the welding advancing direction, and images the welding situation as shown in FIG. This image information is stored in the image processing device 20.
Then, the insertion position TC of the welding wire 24, the contour of the arc 25, and the contour of the molten pool 26, which have come out from the lower portion of the welding torch 23, are obtained. For the contour from the molten pool 26, the X-coordinate values of the contour on the left side are averaged to obtain the left end position BL of the molten pool. Similarly,
The right end BR of the molten pool 26 is calculated. The difference between the center position of the left and right ends and the wire insertion position TC is the positional deviation Δ of the welding torch 23.
Calculate as X. In addition, some points at the tip of the molten pool 26 are calculated by the least square method or the Hough transform method to obtain a straight line expression LK.
Is calculated, and the inclination angle Δθ of the straight line LK is calculated using this as the inclination of the bead. Fuzzy inference is performed by using the calculated positional deviation amount ΔX of the welding torch 23, the bead inclination angle Δθ, and the groove-tracing speed V including the previously controlled direction of the torch left-right scanning shaft 27 as fuzzy inference, and the torch left-right scanning shaft 27 The groove copying speed V'including the direction is calculated, and the torch left and right copying shaft 27 is moved via the control device 28. The TV monitor 22 can be used for checking the image, but it is not always necessary. In addition, it is possible to control the welding power source 29 and other peripheral devices by feedback-controlling the welding conditions from the shape of the arc 25 and various information of the weld pool 26 (for example, tip rounding information, left and right corner shapes) obtained from this image. Become.
The fuzzy control can be controlled by setting control rules as shown in "Table 1", "Table 2", and "Table 3" below.

[0016]

[Table 1]

[0017]

[Table 2]

[0018]

[Table 3] In addition, in each of the above Tables 1 to 3, each label is as follows. NL; Negative Large NM; Negative Medium ZR; Zero PS; Positive Small PM; Positive Medium PL; Positive Large Like this, rather large, very large.
Types are given meanings, and rules are set with this. That is, the rules of 1 to 7 are controls for the deviation of the welding torch (electrode) 1, and in the example of 1, 1F torch position deviation = "NL" THEN torch speed (left and right copying axis 9) = "9L" → "torch position" If there is a large deviation to the left, move the torch speed to the right at a fairly high speed. ”In the case of 4, 1F Torch position deviation =" ZR "THEN Torch speed =
“ZR” → “If there is no torch position deviation, set the torch speed to 0.” Under the rules 3 to 16, the welding torch deviation and the bead inclination are controlled, and the 11th example is 1F torch position deviation = “ ZR "AND bead slope =
"NM" THEN Torch speed = "PS" → "If there is almost no torch position deviation and the bead inclination is negative (left side is high), move the welding torch to the right slowly.

In the example of 16, 1F torch position deviation = “PM” AND bead inclination =
"PM" THEN Torch speed = "NM" → "If the torch position is shifted to the right and the bead inclination is positive (right side is high), move the welding torch to the left at a high speed.

Further, in the rule examples of 17 to 22, the control is performed for the speed including the displacement of the welding torch and the moving direction of the previously moved welding torch. In the 19th example, 1F torch position displacement = “ZR "AND Previous torch speed =" NM "THEN Torch speed =" PS "→" If there is almost no torch position deviation and the previous torch speed is moving to the left, move the welding torch slowly to the right. .

In the example of 22, 1F torch position shift = “PM” AND previous torch speed = “PM” THEN torch speed = “NM” → “the torch position is shifted to the right and the previous torch velocity is to the left. If it is moving, move the welding torch to the left at a high speed.

With these rule settings, "MIN-MAX
By using the "center of gravity method" or the "algebraic product-addition-center of gravity method", fuzzy inference is performed, the next torch speed is obtained, and this is controlled via the control device, so that an accurate groove tracking can be performed. The label setting of this rule can be changed according to the situation.

[0023]

As described above in detail, according to the present invention,
The life of the GMA welding can be extended under a bad environment, and a clear image can be obtained to perform highly accurate welding line tracing control. In addition, the welding condition can be judged from the shape of the molten pool, and feedback can be made from this shape to the welding conditions, which enables higher quality welding.

[Brief description of drawings]

FIG. 1 is an explanatory view of a GMA welding visual sensor according to an embodiment of the present invention.

FIG. 2 is an explanatory diagram in the case of performing groove tracking using the visual sensor for welding shown in FIG.

FIG. 3 is an explanatory view of a welding situation when the visual sensor for welding shown in FIG. 1 is used.

[Explanation of symbols]

1 ... CCD camera, 2a, 2b, 2c ... housing
3 ... Lens, 4 ... Interference filter, 5 ... Protective glass,
6a, 6b ... Air, 7 ... Sputter,
8 ... Pinhole, 9 ... Hood, 10, 11
… Air inlet tube, 12… Hume.

─────────────────────────────────────────────────── ───

[Procedure amendment]

[Submission date] October 25, 1994

[Procedure Amendment 1]

[Document name to be amended] Statement

[Correction target item name] 0001

[Correction method] Change

[Correction content]

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a welding visual sensor and a welding control method, and particularly to remote monitoring of welding conditions for GMA welding , or image processing based on the obtained image to perform automatic welding and unmanned welding. The present invention relates to a welding visual sensor used as a purpose.

[Procedure Amendment 2]

[Document name to be amended] Statement

[Name of item to be corrected] 0002

[Correction method] Change

[Correction content]

[0002]

2. Description of the Related Art Conventionally, as a sensor for monitoring the welding condition, there are a method using an infrared camera and a method combining a CCD camera and an interference filter. Also, as a method of following the groove line by image processing, a molten pool or arc shape is used .
The groove line was traced using the deviation from the center of the molten pool or the center of the arc as the control amount, assuming that the wire insertion position was from the position of the depression formed in the upper part of the groove .

[Procedure 3]

[Document name to be amended] Statement

[Name of item to be corrected] 0003

[Correction method] Change

[Correction content]

[0003]

With the infrared camera method, the shape of the molten pool could be obtained, but other conditions inside the groove could not be recognized by the infrared camera, and humans could I couldn't get enough information from it to monitor. In addition, the intensity of welding light is strong only with a CCD camera (with a lens, of course), and it is impossible to obtain an image capable of monitoring the welding situation due to halation or blooming, and a person is welding. Even if the light-shielding filter used sometimes is placed in front of the CCD camera, it is clear
I couldn't get the image. So CCD camera
A visual sensor that puts an interference filter in front and reproduces an image more clearly has been developed and put into practical use.

[Procedure amendment 4]

[Document name to be amended] Statement

[Correction target item name] 0004

[Correction method] Change

[Correction content]

Further, in GMA welding , spatter and fume are often generated, and it is necessary to frequently replace glass or the like to protect the lens, which is an obstacle to automation. Furthermore, although these sensors are used as a groove line copy, as a control method, the amount of deviation of the wire insertion position from the center of the molten pool or the center of the arc is calculated, and control is performed to eliminate this deviation. . However, in this case, it was not possible to provide high-quality welding due to the time delay and the bead shape that was placed before.

[Procedure Amendment 5]

[Document name to be amended] Statement

[Correction target item name] 0008

[Correction method] Change

[Correction content]

[0008]

SUMMARY OF] According to the present invention, both decreasing the arc light by employing interference filters in the near-infrared region, visible even molten pool by approaching the emission region of more molten pool and open invisible to the infrared region It becomes possible to obtain more information in the groove such as the state of the front bead in the tip and the arc shape.By placing a pinhole in front of it and applying pressure with air or gas before and after it, spatter and fumes can be obtained. You can change the path of, and you can use the lens and interference filter for a long time without damaging.

[Procedure correction 6]

[Document name to be amended] Statement

[Correction target item name] 0012

[Correction method] Change

[Correction content]

In the GMA visual sensor having such a structure, the spatter 7 and the fume 12 are removed by applying an internal pressure to the hood 9 by the air 6a introduced from the air introduction pipe 10 .
And is discharged from a hole formed in the pinhole 8 to prevent the spatter 7 and the fume 12 from entering through this hole. to this,
Further, in order to remove the adhesion of the entire surface of the pinhole 8 to the periphery of the hole and to prevent the air 6a from the pinhole 8 from destroying the gas atmosphere of the welded part, the air 6b is introduced into the air introduction pipe 11 in order to remove its trajectory from the welded part. Introduced from. Air 6b is air
The air 6b is introduced from the introduction pipe 11 , hits the hood 9, and jets air 6b upward. With this momentum, the trajectory of the air 6a can be changed upward. The fume 12 is also excluded from the front of the camera by the air 6a and 6b, so that an open field of view can be obtained.

[Procedure Amendment 7]

[Document name to be amended] Statement

[Name of item to be corrected] 0015

[Correction method] Change

[Correction content]

Next, the operation will be described. The GMA
The visual sensor for use 13 is arranged in the welding advancing direction, and images the welding situation as shown in FIG. This image information is stored in the image processing device 21.
Then, the insertion position TC of the welding wire 24, the contour of the arc 25, and the contour of the molten pool 26, which have come out from the lower portion of the welding torch 23, are obtained. For the contour from the molten pool 26, the X-coordinate values of the contour on the left side are averaged to obtain the left end position BL of the molten pool. Similarly,
The right end BR of the molten pool 26 is calculated. The difference between the center position of the left and right ends and the wire insertion position TC is the positional deviation Δ of the welding torch 23.
Calculate as X. In addition, some points at the tip of the molten pool 26 are calculated by the least square method or the Hough transform method to obtain a straight line expression LK.
Is calculated, and the inclination angle Δθ of the straight line LK is calculated using this as the inclination of the bead. Fuzzy inference is performed by using the calculated positional deviation amount ΔX of the welding torch 23, the bead inclination angle Δθ, and the groove-tracing speed V including the previously controlled direction of the torch left-right scanning shaft 27 as fuzzy inference, and the torch left-right scanning shaft 27 The groove copying speed V'including the direction is calculated, and the torch left and right copying shaft 27 is moved via the control device 28. The TV monitor 22 can be used for checking the image, but it is not always necessary. In addition, it is possible to control the welding power source 29 and other peripheral devices by feedback-controlling the welding conditions from the shape of the arc 25 and various information of the weld pool 26 (for example, tip rounding information, left and right corner shapes) obtained from this image. Become.
The fuzzy control can be controlled by setting control rules as shown in "Table 1", "Table 2", and "Table 3" below.

[Procedure Amendment 8]

[Document name to be amended] Statement

[Correction target item name] 0018

[Correction method] Change

[Correction content]

[0018]

[Table 3] In addition, in each of the above Tables 1 to 3, each label is shown below.
On the street . NL; Negative Large NM; Negative Medium ZR; Zero PS; Positive Small PM; Positive Medium PL; Positive Large Like this, rather large, very large.
Types are given meanings, and rules are set with this. That is, the rules of 1 to 7 are control for the deviation of the welding torch (electrode) 1. In the example of 1, the IF torch position deviation = "NL" THEN torch speed (left and right copying axis 9) = "9L" → "torch position" If there is a large deviation to the left, move the torch speed to the right at a fairly high speed. ”In the case of 4, IF torch position deviation =" ZR "THEN Torch speed =
“ZR” → “If there is no torch position deviation, set the torch speed to 0.” Under the rules of 3 to 16, the welding torch deviation and the bead inclination are controlled, and the 11th example is IF torch position deviation = “ ZR "AND bead slope =
"NM" THEN Torch speed = "PS" → "If there is almost no torch position deviation and the bead inclination is negative (left side is high), move the welding torch to the right slowly.

[Procedure Amendment 9]

[Document name to be amended] Statement

[Correction target item name] 0019

[Correction method] Change

[Correction content]

In the example of 16, the IF torch position shift = “PM” AND bead inclination =
"PM" THEN torch speed = "NM" → "torch position has shifted to the right, the slope of the bead is plastic
If it is high (right side is high) , move the welding torch to the left at high speed.

[Procedure Amendment 10]

[Document name to be amended] Statement

[Correction target item name] 0020

[Correction method] Change

[Correction content]

Further, in the rule examples 17 to 22, the control is performed for the speed including the displacement of the welding torch and the moving direction of the previously moved welding torch. In the example of 19, the IF torch position displacement = “ZR "AND Previous torch speed =" NM "THEN Torch speed =" PS "→" If there is almost no torch position deviation and the previous torch speed is moving to the left, move the welding torch slowly to the right. .

[Procedure Amendment 11]

[Document name to be amended] Statement

[Correction target item name] 0021

[Correction method] Change

[Correction content]

In the example of 22, the IF torch position shift = “PM” AND the previous torch speed = “PM” THEN torch speed = “NM” → “The torch position is shifted to the right, and the previous torch speed is shifted to the left. If it is moving, move the welding torch to the left at a high speed.

[Procedure Amendment 12]

[Document name to be amended] Statement

[Name of item to be corrected] Brief description of the drawing

[Correction method] Change

[Correction content]

[Brief description of drawings]

FIG. 1 is an explanatory diagram of a GMA welding visual sensor according to an embodiment of the present invention.

FIG. 2 is an explanatory diagram in the case of performing groove tracking using the visual sensor for welding shown in FIG.

FIG. 3 is an explanatory view of a welding situation when the visual sensor for welding shown in FIG. 1 is used.

[Explanation of Codes] 1 ... CCD camera, 2a, 2b, 2c ... Case,
3 ... Lens, 4 ... Interference filter, 5 ... Protective glass,
6a, 6b ... Air, 7 ... Sputter,
8 ... Pinhole, 9 ... Hood, 10, 11
… Air inlet tube, 12… Hume.

[Procedure Amendment 13]

[Document name to be corrected] Drawing

[Name of item to be corrected] Figure 1

[Correction method] Change

[Correction content]

[Figure 1]

[Procedure Amendment 14]

[Document name to be corrected] Drawing

[Name of item to be corrected] Figure 3

[Correction method] Change

[Correction content]

[Figure 3]

Claims (2)

[Claims] 1. A visual sensor for welding, which is used when remotely monitoring the electrode consumable welding or automatically controlling the monitor image thereof by a pixel processing device, etc.
Camera with a center wavelength of 0.6 to 0.8 μm of 0.1
an interference filter that transmits wavelengths in the range of μm or less,
A visual sensor for welding, comprising: a pinhole provided in front of the interference filter, the hole having a diameter of 2 mm or less, and pressure is applied by air or gas from both inside and outside of the pinhole. 2. The welding visual sensor according to claim 1, and a wire insertion position by an image processing device connected to the sensor.
Fuzzy inference is performed by using the torch axis movement speed in the axial direction that crosses the front bead shape and the groove line and its direction as a parameter, and the moving speed and movement direction of the next torch axis or torch axis position is determined to determine the groove line. A welding control method characterized by performing copying.