JP2007181871A - Automatic arc welding system and method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an automatic arc welding system and method capable of reducing the number of necessary sensors. <P>SOLUTION: The welding system 101, in which a welding torch is moved along the extending direction of a groove 13 so as to follow a welding torch to a horizontal direction and a vertical direction of the groove, while by protoruding a wire 18 from the top end of the welding torch 9 and melting the wire 18 with the arc, thus beads are formed in the groove, is provided with: an imaging apparatus 4 for imaging images of the wire and the arc; an image processing means 6 for extracting the protrusion length of the wire by image-processing an imaged video; and a control means 6 for vertically following the groove by controlling the extracted protrusion length of the wire so as to be kept constant. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to an automatic arc welding system and method, and more particularly, to controlling welding conditions based on an image of a welded portion.

  In recent years, an automatic arc welding method has been proposed in order to improve the work efficiency of arc welding (see, for example, Patent Document 1). By the way, in order to make the laminated bead height uniform over the entire joint line regardless of the groove width variation in the multi-layer groove welding, it is necessary to know the increase / decrease distribution of the required bead cross-sectional area in each pass. For this purpose, information on the groove width at the wire tip height and the current stacking height is required. Therefore, in this automatic arc welding method, a laser slit light sensor is used to set various conditions for the next pass in order to grasp the shape of the bead immediately after lamination, a welded portion direct-view camera is used for groove left-right copying, and By using an arc sensor for groove up-and-down copying, information on the groove width at the wire tip height and the current stack height is obtained. Thereby, automation is achieved.

In addition, an automatic arc welding method has been proposed in which features of the shape of the arc high-brightness part are extracted from the image of the welded part obtained by a camera attached in front of the welded part, and groove left-right tracing is performed based on the extracted feature. (See Patent Document 2 and Patent Document 3).
JP 2001-340966 A JP 2004-042067 A JP 2005-193277 A

  However, the automatic arc welding method of Patent Document 1 has a problem that a large number of sensors are required.

  Further, Patent Document 2 and Patent Document 3 do not disclose a sensor that is used for purposes other than groove left / right copying.

  The present invention has been made to solve the above-described problems, and an object thereof is to provide an automatic arc dynamic welding system and method capable of reducing the number of necessary sensors.

  In order to solve the above-mentioned problems, an automatic arc welding system according to the present invention causes a wire to protrude from the tip of a welding torch and melts it by arc so that the welding torch follows the groove from side to side and groove from top to bottom. In the welding system that moves along the extending direction of the groove and thereby forms a bead in the groove, the imaging device that images the wire and the arc, Image processing means for extracting the protruding length of the wire, and control means for performing the groove up-and-down copying by controlling the extracted protruding length of the wire to be a constant length. With such a configuration, the number of necessary sensors can be reduced by performing groove left / right copying control based on image processing of a video imaged by the imaging device.

  The imaging apparatus may perform the imaging from the front in the moving direction along the extending direction of the groove of the welding torch and from above. With such a configuration, a video suitable for image processing can be obtained.

  The control means may control the welding speed based on the current height of the tip position of the wire and the current data for following the groove left and right, which is strongly related to the bead surface height of the front layer. . With such a configuration, the laminated bead height can be suitably controlled.

  The image processing means may extract the feature of the shape of the high-intensity part of the arc in the image processing, and the control means may perform the groove left-right tracing based on the extracted feature of the shape. With such a configuration, the number of necessary sensors can be reduced.

  Further, the automatic arc welding method according to the present invention extends the groove by causing the wire to protrude from the tip of the welding torch and melting the welding torch while following the groove to the left and right and the groove up and down. In a welding system and method for moving along a current direction and thereby forming a bead in the groove, a step of imaging the wire and the arc, and a projection length of the wire by image processing the captured image And a step of performing the groove up-and-down copying by controlling so that the protruding length of the extracted wire is maintained at a constant length. With such a configuration, the number of necessary sensors can be reduced by performing groove left / right copying control based on image processing of a captured image.

  You may perform the said imaging from the front and the upper direction in the moving direction along the extension direction of the said groove | channel of the said welding torch. With such a configuration, a video suitable for image processing can be obtained.

  The welding speed may be controlled based on the current height of the tip position of the wire and the current data for the groove left and right tracing, which is strongly related to the bead surface height of the front layer. With such a configuration, the laminated bead height can be suitably controlled.

  In the image processing, a feature of the shape of the high-intensity part of the arc may be extracted, and the groove left-right copying may be performed based on the extracted feature of the shape. With such a configuration, the number of necessary sensors can be reduced.

  The present invention has the configuration as described above, and has an effect of providing an automatic arc welding system and method capable of reducing the number of necessary sensors.

Hereinafter, preferred embodiments of the present invention will be described with reference to the drawings.
(Embodiment)
FIG. 1 is a schematic diagram showing a configuration of an automatic arc welding system according to an embodiment of the present invention. FIG. 2 is a perspective view showing the arrangement position of the CCD camera. FIG. 3 is a perspective view schematically showing a welded portion.

  As shown in FIGS. 1 to 3, the automatic arc welding system 101 includes a welding device 1, a robot 2, and an automatic arc welding control device 3.

  In the present embodiment, a pair of workpieces 12 are butt welded. And the groove | channel 13 is formed in the location which should form the joint. The pair of workpieces 12 are arranged with a predetermined root gap RG, and each groove surface is formed to have a predetermined bevel angle. And the plate-shaped backing metal 16 is arrange | positioned so that it may span between the lower surfaces of the butt | matching edge part of a pair of to-be-welded object 12. FIG. A groove 13 is formed so as to be surrounded by the backing metal 16 and the groove surfaces of the pair of workpieces 12.

  The welding apparatus 1 has a welding torch 9 and a welding power source 11. The welding torch 9 is composed of a nozzle that supplies an inert gas to the welded portion. From the tip of the welding torch 9, a wire 10 made of a welding material is fed out at a predetermined speed by a feeding mechanism (not shown). A current is supplied to the wire 10 from a welding power source 11.

  The robot 2 includes a 6-axis articulated arm 7 and a robot controller 8 that controls the operation of the arm 7. A welding torch 9 is attached to the hand portion 7 a of the arm 7. The robot controller 8 controls the arm 7 to move the welding torch 9 so that the tip 10 a of the wire 10 moves in the groove 13. As shown in FIGS. 2 and 3, the tip of the wire 10 moves in the extending direction of the groove 13 while weaving in the groove 13 in the width direction (left-right direction) 19 of the groove 13. The moving direction along the extending direction of the groove 13 of the tip of the wire 10 (and hence the welding torch 9) is hereinafter referred to as a welding direction 18. In conjunction with this movement, a welding current is supplied to the wire 10 by the welding power source 11 described above, and an arc 14 is formed between the tip 10 a of the wire 10 and the bead 17 of the front layer formed in the groove 13. And the tip of the wire 10 melts to form the molten metal pond 15. The molten metal pond 15 is cooled and solidified to form a bead 17.

  The automatic arc welding control device 3 performs image processing on the CCD camera 4, a CCU (camera control unit) 5 that converts and outputs a video image captured by the CCD camera 4 to a predetermined image signal, and an image signal from the CCU 5. It has a personal computer 6 that extracts feature data and outputs a required command or the like to the robot controller 8 based on the extracted feature data.

  As the CCD camera 4, for example, a single-lens monochrome camera is used. Here, the imaging device is a CCD camera, but another imaging device such as a CMOS image sensor may be used. The CCD camera 4 is disposed in front of the welding torch 9 in the welding direction 18 and above the welded portion so that the welded portion can be imaged therefrom (so that the welded portion is in the field of view). Here, as shown in FIG. 3, the welded portion refers to a region composed of a welding torch 9, a wire 10, an arc 14, and a molten metal pond 15. In the present embodiment, as shown in FIG. 2, the CCD camera 4 is attached to the hand portion 7a of the arm 7 so as to face rearward and obliquely downward. That is, the CCD camera 4 is disposed so as to maintain a certain distance from the welded portion. Of course, the CCD camera 4 may be moved so as to maintain a certain distance from the welding torch 9 by another moving mechanism.

  As is well known, the personal computer 6 is controlled by a calculation unit 31 such as a CPU, a storage unit 32 such as an internal memory, an input unit 33 such as a keyboard for inputting data to the calculation unit 31, and the calculation unit 31. And a display unit 34 for performing a required display. A predetermined control program is stored in the storage unit 32. The calculation unit 31 reads out and executes this control program, thereby performing welded portion image processing, groove left / right scanning control, groove vertical scanning control, welding condition control, and the like. In the groove left / right scanning control, groove vertical scanning control, and welding condition control, the calculation unit 31 performs these controls by inputting necessary commands and the like to the robot controller 8. That is, the arithmetic unit 31 controls the operation of the entire automatic arc welding system 101 in cooperation with the robot controller 8. Further, the calculation unit 31 performs a required display on the display unit 34.

  Next, the operation (automatic arc welding method) of the automatic arc welding system 101 configured as described above will be described. Here, the general operation is omitted, and a description will be given of the constant bead height control that characterizes the present invention. This control is performed by cooperative control between the calculation unit 31 of the personal computer 6 and the robot controller 8.

  First, an outline of the stack bead height constant control will be described.

  FIG. 4 is a flowchart showing an outline of the stack bead height constant control.

  1 to 4, when the automatic arc welding system is started, welding conditions are set (step S1). Specifically, the robot 2 is taught the start and end points of welding by teaching. Next, necessary control parameters such as a groove angle and a route gap are input to the calculation unit 31. Next, welding conditions such as weaving conditions are set. Next, an appropriate value or the like of the major axis inclination angle θ of the arc high-brightness portion described later is set. These inputs and settings are performed by inputting required data to the calculation unit 31 via the input unit 33 of the personal computer 6 and storing it in the storage unit 32.

  Next, welding is started by turning on a welding start switch (not shown) (step S2).

  Next, the welded part is imaged (step S3). Specifically, the welded part is imaged as follows.

  FIG. 11 is an image (hereinafter referred to as a welded portion image) showing a welded portion image captured by the CCD camera 4, wherein (a) is a welded portion image before image processing, and (b) is obtained by image processing. It is the welding part image which displayed the extracted welding torch front-end | tip position and wire front-end | tip position.

  The CCD camera 4 captures an image of the welded part taken from the front and above as shown in FIG. The CCU 5 converts this video into a predetermined image signal and inputs it to the personal computer 6. The personal computer 6 temporarily stores an image (welded part image) of a predetermined number of frames (here, 3 frames) of the image signal in the storage unit 32. This welded portion image is a digital image having a predetermined number of pixels and a predetermined number of gradations. In the welded portion image, as shown in FIG. 11A, the molten pool is slightly brightly reflected in the center of the dark background of the screen, and the arc is very bright in the molten pool (hereinafter also referred to as an arc high luminance portion). ) Then, the tip of the wire appears black so as to cut out a part of the upper portion of the bright arc high-luminance portion, and the tip of the arc-like torch is shown slightly brighter at the upper end of the wire.

  Next, the calculation unit 31 of the personal computer 6 reads out the welded part image (more precisely, image data of the welded part image) from the storage unit 32 and performs image processing (step S4).

  Next, on the basis of the data extracted by the image processing, the groove left / right scanning control (horizontal direction, step S5) of the wire 10 and the groove vertical scanning control (vertical direction, step S6) of the wire 10 are performed in parallel. Is called.

  Next, welding condition (welding speed) control is performed (step S7).

  And until welding is complete | finished, step S3-step S7 are repeated, and when welding is complete | finished (step S8), lamination | stacking bead height constant control will be complete | finished.

  Next, image processing (step S4), groove left / right scanning control (step S5), groove vertical scanning control (step S6), and welding condition control (step S7) will be described in detail.

  First, image processing for groove left / right scanning control and groove left / right scanning control (hereinafter referred to as groove left / right scanning control based on image processing) will be described. Note that the groove left / right copying control based on this image processing is well known, and therefore only the minimum necessary description will be given here. For details, refer to JP 2004-042067 A or JP 2005-193277 A.

  The groove left / right scanning control based on image processing differs depending on whether the groove bottom width is large or the groove bottom width is small.

  First, groove left / right tracing control based on image processing when the groove bottom width is large will be described.

  FIG. 5 is a schematic diagram showing a change in the shape of the arc high-intensity part accompanying weaving when the bottom width of the groove is large, and (a) shows the shape of the arc high-intensity part when the wire is positioned at the center of the weaving. FIG. 5B is a diagram showing the shape of the arc high luminance portion when the wire is positioned at the right end of the weaving, and FIG. 5C is a diagram showing the shape of the arc high luminance portion when the wire is substantially in contact with the groove surface. FIG. FIG. 6 is a graph showing a change in the inclination angle of the major axis of the arc high brightness portion with respect to the distance between the electrode and the groove surface when the groove bottom width is large.

  When the arc high-intensity part 14 extracted from the welded part image is observed, the shape changes as shown in FIGS. In other words, when the wire 10 is positioned at the center of the weaving as viewed from the CCD camera 4, the arc high brightness portion 14 forms an ellipse having a major axis L in the substantially horizontal direction, as shown in FIG. is doing. When the wire 10 is positioned at the right end of the weaving as viewed from the CCD camera 4, the welding arc is constrained by the groove surface and the right side is narrowed. Therefore, as shown in FIG. It becomes a slightly thin ellipse, and its long axis L is directed to the upper right. Furthermore, when the wire 10 approaches the groove surface and comes into almost contact, as shown in FIG. 5 (c), the shape of the arc high brightness portion 14 becomes thinner, and the long axis L further rises to the right. The inclination angle increases.

  Such a change in the shape of the arc high-brightness portion 14 is determined by the horizontal distance between the wire and the groove surface and the bevel angle of the groove surface, and it is known that there is almost reproducibility.

  A measurement example of a change in the inclination angle θ (see FIG. 3) of the major axis L of the arc high brightness portion 14 with respect to a horizontal distance between the wire and the groove surface (hereinafter referred to as a wire-groove surface distance) will be described. .

  FIG. 6 is a graph showing a measurement example of a change in the inclination angle θ of the major axis L of the arc high-luminance part with respect to the distance between the wire and the groove surface. FIG. 6 is a plot of measured values obtained for a pair of workpieces 12 arranged so that the groove surface has a bevel angle of 30 degrees.

  In FIG. 6, as the distance between the wire and the groove surface decreases, the inclination angle θ of the major axis L of the arc-high luminance portion 14 (hereinafter simply referred to as the inclination angle of the major axis) decreases substantially monotonously. In general, an appropriate weaving end point is a position where the wire 10 is about 2 mm away from the groove surface. Therefore, according to this measurement example, if the inclination angle θ of the major axis at the end point of the weaving is approximately 30 degrees, it can be considered that proper weaving welding has been performed. It is necessary to adjust the weaving so that the inclination angle θ approaches 30 degrees. This adjustment is performed when the calculation unit 31 of the personal computer 6 inputs to the robot controller 8 a weaving end point position correction amount corresponding to a deviation from an appropriate value (here, 30 degrees) of the inclination angle θ of the long axis. .

  The groove left / right tracing control based on the image processing in the present embodiment is performed using the relationship between the wire-groove surface distance and the major axis inclination angle θ.

  Next, groove left / right tracing control based on this image processing will be described in detail with reference to FIGS. 1 to 3 and FIG.

  First, when it is confirmed that the weaving has reached the right end, the calculation unit 31 of the personal computer 6 takes in the welded part image from the storage unit 32 and calculates the major axis inclination angle θ by image processing. Specifically, the acquired welded portion image is binarized using a predetermined threshold value, and the shape of the arc high luminance portion 14 is extracted. Since the arc high luminance part 14 has extremely high luminance as compared with other regions such as the molten pool 15, the threshold value can be easily set, and this can be extracted easily.

  Next, these noise components are removed by deleting a high luminance region having only an area equal to or less than a predetermined threshold.

  Next, the center-of-gravity coordinates G are calculated for the extracted arc high luminance part 14. The center-of-gravity coordinates G (Xg, Yg) can be easily calculated by calculating the average value of the x-coordinate xi and the y-coordinate yi for each pixel in the arc high luminance part 14.

  Further, using the coordinates (xi, yi) of each pixel of the arc high luminance portion 14, the angle θ of the symmetry axis that minimizes the moment of inertia of the graphic of the arc high luminance portion 14 is calculated by the following equation.

This equation obtains θ such that dM / dθ = 0 with respect to the equation M representing the moment around the straight line y = tan θ × (x−Xg) + Yg.
θ = −0.5 atan [2Σ (xi−Xg) (yi−Yg) / {Σ (xi−Xg) ² −Σ (yi−Yg) ² }]
It is. However, −90 ° ≦ θ ≦ + 90 °.

  The angle θ is along the longest axis of this figure. Here, an axis passing through the center of gravity G and having an angle θ is defined as a major axis of the arc high-brightness unit 14, and an axis passing through the center of gravity G and perpendicular to the major axis is defined as a minor axis. Here, the xy coordinate system is such that the x coordinate is taken in the horizontal direction and the y coordinate is taken in the vertically downward direction. The angle θ is measured counterclockwise with respect to the positive x axis.

  Next, the calculation unit 31 compares the inclination angle θ of the major axis L obtained in this way with an appropriate value, and calculates a position correction amount ΔR at the right end from the deviation.

  Next, when it is confirmed that the weaving has reached the left end, the arithmetic unit 31 calculates the inclination angle θ of the major axis in the same manner as described above, and calculates the position correction amount ΔL at the left end.

  Next, when it is confirmed that the weaving has reached the center position, the calculation unit 31 calculates a position correction amount at the right end (hereinafter referred to as a right correction amount) ΔR and a position correction amount at the left end (hereinafter referred to as a left correction amount) ΔL. A correction amount for correcting the deviation of the central position of the weaving by combining is calculated. Specifically, the calculation unit 31 sets the half of the difference between the right correction amount ΔR and the left correction amount ΔL as the correction amount ΔXc at the weaving center position, and calculates the sum of the right correction amount ΔR and the left correction amount ΔL as the weaving width. The correction amount is ΔW. Then, the weaving center position correction amount ΔXc and the weaving width correction amount ΔW are input to the robot controller 8. Then, the robot controller 8 adjusts the weaving operation of the wire 10 according to the correction amount. As a result, the weaving operation is corrected. In this way, the groove left / right tracing control of the wire 10 when the bottom width of the groove is large is performed.

  Next, groove left / right tracing control based on image processing when the groove bottom width is small will be described.

  FIG. 7 is a perspective view schematically showing a welded portion when the groove bottom width is small. As shown in FIG. 7, when the bottom width of the groove is small, that is, when the initial layer of the downward V groove or the groove angle is small, the arc 14 is simultaneously constrained by the groove surfaces on both sides, and welding is performed. In the partial image, the arc high-intensity portion 14 appears in a lump shape, and does not appear in a pseudo ellipse that can determine the long axis and accurately calculate the inclination. Therefore, the measurement of the weaving position depending on the inclination angle θ of the long axis is inferior in reliability. Therefore, in such a case, the weaving position is measured using a ratio P (hereinafter, referred to as a long / short axis length ratio) P between the major axis length and the minor axis length of the arc high luminance portion 14.

  FIG. 8 is a schematic diagram showing a change in the shape of the arc high luminance part accompanying weaving when the bottom width of the groove is small. FIG. 8A shows the shape of the arc high luminance part when the wire reaches the left end of the weaving. The figure which shows, (b) is a figure which shows the shape of an arc high-intensity part when a wire reaches | attains a weaving center position, (c) is a figure which shows the shape of an arc high-intensity part when a wire reaches | attains the weaving right end. is there. FIG. 9 is a graph showing a change in the long / short axis length ratio P of the arc high luminance portion with respect to the distance from the groove center.

  When the bottom width of the groove is small, the shape changes as shown in FIGS. 8A to 8C when the arc high luminance portion 14 extracted from the welded portion image is observed. That is, when the wire 10 approaches the left groove surface, as shown in FIG. 8 (a), the arc high luminance portion 14 does not spread to the left side, and thus has a slightly flat shape that rises to the left. When the wire reaches the weaving center position, as shown in FIG. 8 (b), the arc high luminance portion 14 becomes a block shape close to a triangle, and the length of the major axis and the length of the minor axis become substantially equal. Then, as the shape changes from the left shoulder rise to the right shoulder rise, the major axis and the minor axis are switched. When the wire 10 approaches the groove surface on the right side, as shown in FIG. 8C, the arc high brightness portion 14 has a right shoulder substantially symmetrical to the shape of the arc high brightness portion 14 in FIG. Form a rising shape.

  The calculation unit 31 calculates the long / short axis length ratio P as follows.

7 and 8, the calculation unit 31 calculates the center of gravity G of the arc high-brightness unit 14 by the calculation using the equation M representing the moment of inertia and the inclination of the long axis passing through the center of gravity G as described above. The angle θ is calculated. The minor axis inclination angle σ is calculated by adding ± 90 ° to the major axis inclination angle θ. Next, using the coordinates (Xg, Yg) of the center of gravity G, the inclination angle θ of the major axis, the inclination angle σ of the minor axis, and the real parameter r, an equation of the major axis and the minor axis is established. For example, the equation for the major axis is
x = Xg + r cos θ
y = Yg−rsinθ
It becomes. The absolute value of the real number parameter r represents the distance between the point on the straight line and the center of gravity G.

  Using this equation, the intersection of the arc high brightness portion 14 with the major axis and minor axis is obtained, and the major axis length a and minor axis length b are calculated, respectively. Then, the long / short axis length ratio P = b / a is calculated using the long axis length a and the short axis length b.

  The long / short axis length ratio P thus obtained is a relatively stable function with respect to the position of the wire, and in a downward V groove having the same bevel angle, it is symmetric with respect to the center of the groove. ing.

  A measurement example of the change in the long / short axis length ratio P with respect to the distance from the center of the groove will be described.

  FIG. 10 is a graph showing a measurement example of a change in the long / short axis length ratio P with respect to the distance from the center of the groove. FIG. 10 shows the center of the groove when weaving welding is performed on a pair of objects to be measured arranged so as to form a groove having a bevel angle of 30 ° on both sides and a root gap of 0 mm. The measured value of the long / short axis length ratio P with respect to the distance from is plotted. As can be seen from FIG. 10, the long / short axis length ratio P has a maximum value of 1 when the wire is located at the center of the groove, and decreases at substantially the same rate on the left and right as the wire moves away from the center.

  In the present embodiment, the calculation unit 31 compares the measured value of the long / short axis length ratio P at the weaving end with an appropriate value and performs correction so as to eliminate the deviation. The calculation unit 31 calculates the right correction amount ΔR and the left correction amount ΔL from the deviation of the long / short axis length ratio P from the appropriate value at the weaving end, respectively, and weaves the half of the difference between the right correction amount ΔR and the left correction amount ΔL. The correction amount ΔXc at the center position is set, and the sum of the right correction amount ΔR and the left correction amount ΔL is set as a weaving width correction amount ΔW. Whether the wire 10 is in the left region or the right region of the groove 13 is determined by whether the inclination angle θ of the major axis is positive or negative. Then, the weaving center position correction amount ΔXc and the weaving width correction amount ΔW are input to the robot controller 8. Then, the robot controller 8 adjusts the weaving operation of the wire 10 according to these correction amounts. As a result, the weaving operation is corrected. In this way, groove left / right copying control is performed when the groove bottom width is small.

  In the present embodiment, for example, a threshold (for example, 4 mm) is provided for the taught weaving width, and when the width is equal to or smaller than the threshold, the groove bottom width is small (the long / short axis length ratio P is defined as the feature amount). Execute groove left / right scanning control, and if the threshold is exceeded, switch to groove left / right scanning control when the bottom width of the groove is large (characterized by the major axis inclination angle θ). Is done. It should be noted that in the above-described double groove left / right scanning control, a procedure is described in which it is determined that the calculation unit 31 has reached the weaving end point and image processing is started. However, if the fact that the major axis tilt angle θ and the major / minor axis length ratio P are minimized or maximized at the weaving end point is used, the calculation unit 31 can always perform image processing to evaluate the time series change of these feature values. Can specify the weaving end point arrival time. In practice, the calculation unit 31 performs groove left / right copying control using the value of the feature amount at the time of arrival of the weaving end point specified in the above manner.

  Next, image processing for groove vertical scanning control and groove vertical scanning control will be described.

  1 to 3, when a welding power source 11 having a constant voltage characteristic, which is generally used, is used, welding is performed by moving the welding torch 9 parallel to the surface of the front layer bead 17 by groove up-and-down copying. Since the change in the arc length La is small at the center, the apparent wire protrusion length (hereinafter simply referred to as wire protrusion length) Lw is constant. Therefore, the groove vertical scanning control can be regarded as performing a constant wire protrusion length control. When the groove height control is performed, the wire protrusion length is maintained constant, thereby maintaining a stable welding current, and as a result, welding quality is ensured.

  Therefore, as a specific method of groove up-and-down scanning control, a method in which the welding current is directly taken into the personal computer 6 and the height of the welding torch 9 is corrected so as to maintain an appropriate current using this as an index is easily conceived. In this embodiment, the method of correcting the height of the welding torch 9 (and hence the height of the tip 10a of the wire 10) so that the wire protrusion length is constant is adopted. This is because the CCD camera 4 as the welded part image sensor used for the groove lateral scanning control described above is also used for the groove vertical scanning control as it is, and a welding system that does not use any sensor other than the welded part image sensor is constructed. It is to do.

  The wire protrusion length is essentially the protrusion length of the wire from the tip of the contact tip for supplying power to the wire. However, when performing image processing, the tip of the contact tip is not necessarily seen in the welded part image, and here, from the tip of the gas nozzle constituting the tip of the torch (hereinafter simply referred to as the tip of the torch) to the tip of the wire. The distance is defined as the wire protrusion length. Therefore, in the image processing for groove vertical scanning control, finally, the wire tip position and the welding torch tip position indicated by x in FIG. 11B are extracted.

  First, the wire tip position extraction process will be described in detail.

  12A and 12B are images showing a process of image processing for extracting the wire tip position from the welded portion image, wherein FIG. 12A is an image showing the arc high-intensity portion and its peripheral region cut out after binarization, and FIG. An image obtained by closing the image of (a), (c) is an image obtained by extracting a cutout portion (hereinafter simply referred to as a cutout portion) of the arc high luminance portion of the image of (b), and (d) is an image of (c). It is an image which shows the area | region gravity center (wire tip position) of the notch part in FIG.

  The computing unit 31 of the personal computer 6 (hereinafter simply referred to as the computing unit 31) binarizes the image of the welded portion shown in FIG. 11A, and then arc high brightness from the binarized image. 12 and its peripheral region are cut out to obtain an image as shown in FIG. 12A, which is stored in a buffer (hereinafter referred to as buffer A) of the storage unit 32.

  Next, the calculation unit 31 performs a close process on the image stored in the buffer A to obtain an image as shown in FIG. 12B, and stores this in another buffer (hereinafter referred to as buffer B) of the storage unit 32. To do. Here, the close process refers to a process of removing a minute hole inside the white area formed by the arc high-luminance part and a notch portion of the area.

  Next, the calculation unit 31 subtracts the image stored in the buffer A from the image stored in the buffer B to extract the cutout portion of the arc high-luminance portion as shown in FIG. save. The subtraction of the two images is performed by subtracting the gradation values of the pixels corresponding to each other in both images.

  Next, the calculation unit 31 calculates the area center of gravity of the notch in FIG. 12C, and sets this as the wire tip position as shown in FIG.

  Next, the extraction process of the welding torch tip position will be described in detail.

  FIG. 13 is an image showing a process of image processing for extracting a welding torch tip position from a welded part image, wherein (a) is a welded part image showing a region to be cut out as a welding torch search region, and (b) is (a) ) Is an image showing a welding torch search area cut out from the image, (c) is an image obtained by improving the image quality of the image of (b), and (d) is an image showing the position of the tip of the welding torch.

  First, as shown in FIG. 13A, the calculation unit 31 cuts out a region (welding torch search region) where a welding torch is to be searched from the welded portion image. This is to narrow down the area to prevent erroneous detection and to reduce the load of image processing. This cutout is cut out with reference to the tip position of the wire. In this cut-out welding torch search region image, as shown in FIG. 13 (b), a slightly bright molten pool located in the lower half of the weld torch and a groove surface darker than the molten pool located on both sides of the molten pool. And a black wire extending vertically in the center of the molten pool, and a slightly bright arc-shaped welding torch tip positioned above the molten pool.

  Next, the calculation unit 31 improves the image quality of the image in the welding torch search area to obtain an image as shown in FIG. In this image quality improvement, so-called “γ correction” and so-called “histogram adjustment” are performed to increase the contrast of the tip of the welding torch, and so-called “histogram adjustment” is performed to remove noise while leaving the shading region edge portion clear. "Median filter" was applied. Thereby, in the image after image quality improvement shown in FIG. 13C, the region corresponding to the tip of the welding torch is whitened and becomes clear.

  Next, the calculation unit 31 extracts the welding torch tip position located on the same vertical line as the wire tip position as the welding torch tip position in the image of FIG. As shown in FIG. 13D, this extraction is performed by scanning the same vertical line as the wire tip position from the upper side of the image, and using the density gradient and Laplacian distribution information, Do it like detecting.

  Next, the entire groove vertical scanning control of the wire 10 based on such image processing will be described with reference to FIG.

  FIG. 10 is a flowchart showing groove vertical scanning control of the wire 10 based on image processing. 10, steps S11 to S17 correspond to step S4 (image processing) in the flowchart of FIG. 4, and steps S18 to S23 correspond to step S6 (groove up-and-down scanning control) in the flowchart of FIG.

  First, the calculation unit 31 takes in a welded part image from the storage unit 33 (step S11). Here, a welded part image for three frames is captured.

  Subsequently, the calculating part 31 extracts the feature-value of a welding part image (step S12). This feature value extraction indicates the acquisition of the arc high-luminance part major axis inclination angle θ or the major / minor axis length ratio P, which is necessary for groove left-right tracing, and weaving by examining the time-series changes of these feature quantities The end arrival time is identified and the weaving phase is grasped.

  Next, the calculation unit 31 determines whether or not the weaving of the wire 10 has reached the center section by comparing the elapsed time from the weaving end arrival time with the weaving quarter cycle (step S13). If the center section has not been reached, the process returns to step S11, and steps S11 to S13 are repeated until the center section is reached. Here, the central section of the weaving is a section having a length corresponding to three frames of the image signal input from the CCU 5, and is a section centered on the central position of the weaving.

  When the weaving reaches the center section, the calculation unit 31 first extracts the wire tip position (step S14), and then extracts the welding torch tip position (step S15).

  Next, the calculation unit 31 calculates the distance between the wire tip position and the welding torch tip position, and sets this as the wire protrusion length Lw (see FIG. 3) (step S16).

  Next, the computing unit 31 determines whether or not the weaving is a section corresponding to the last frame of the sections for three frames (step S17). If it is not the section corresponding to the last frame, the process returns to step S11, and steps S11 to S17 are repeated until the section is reached. And if it becomes the area corresponding to the last frame, the calculating part 31 will progress to step S18.

  In step S <b> 18, the calculation unit 31 averages the wire protrusion length Lw for three frames and sets this as the latest measured value. Thereby, the variation in the measured value of the wire protrusion length is reduced.

  Next, the calculation unit 31 calculates a moving average for the past three control periods for the wire protrusion length, and sets this as a measured value of the wire protrusion length (step S19). This is because the wire protrusion length does not change abruptly in normal welding, and thus the measurement is stabilized by taking the moving average in this way. The control cycle is a cycle in which the welding torch 9 passes through the weaving center position, and thus is a 1/2 weaving cycle.

  Next, the calculation unit 31 compares the measured value of the wire protrusion length with the set appropriate value and calculates the difference (deviation) (step S20). Here, since the measured value of the wire protrusion length is extracted as the length on the welded part image, it is strongly influenced by the setting of the welding torch 9 and the CCD camera 4. Therefore, in this embodiment, in order to improve versatility, the measured value of the wire protrusion length obtained by the above-described image processing at a stable time after starting welding without setting an appropriate value in advance is calculated by the calculation unit. 31 is automatically set as an appropriate value. Alternatively, the welding test of the same joint may be performed in advance, the measured value of the wire protrusion length at that time may be sampled, and this value may be set in advance as an appropriate value.

  Next, the calculation unit 31 calculates a height correction amount (hereinafter referred to as a height correction amount) of the wire tip position based on this difference (step S21). Specifically, since this difference is obtained as a length on the welded part image, the calculation unit 31 converts this into an actual length, and multiplies the converted length by an appropriate gain to obtain a height correction amount. obtain.

  Next, the calculation unit 31 inputs this height correction amount to the robot controller 8. Then, the robot controller 8 adjusts the height position of the wire 10 according to the height correction amount. Thereby, the height position of the wire 10 is corrected (step S22).

  Next, the calculation unit 31 determines whether or not the welding is finished (step S23). And when welding is not complete | finished, it returns to step S11, repeats each above-mentioned step, and when welding is complete | finished, this control is complete | finished.

  In this way, groove vertical scanning control of the wire 10 is performed.

  Next, a function confirmation test of the groove vertical scanning control will be described.

  In order to evaluate the effectiveness of the groove up-and-down scanning control, the inventors of the present invention have the condition that in the teaching of the welding operation point, the welding start point is correct and the welding end point is given a shift amount in the groove height direction. A welding test was performed below.

  The appropriate value of the wire protrusion length was set by sampling the measured value of the wire protrusion length during the number of weaving cycles after the start of welding, assuming that the state immediately after the start of welding is an appropriate state.

  FIG. 14 is a schematic diagram showing a welding location in a function confirmation test of groove vertical scanning control. FIG. 15 is a schematic diagram showing the teaching contents in the function confirmation test of groove vertical scanning control.

  As shown in FIG. 14, in this function confirmation test, a base material having a plate thickness of 21 mm was used as the DUT 12. The root gap was 9 mm and the groove angle was 60 °.

  Further, as shown in FIG. 15, the amount of shift in the height direction at the welding end teaching point 42 with respect to the welding start teaching point 41 is set to 10.9 mm corresponding to 5% of the welding length 218 mm.

  FIG. 16 is a diagram showing the result of the function confirmation test when the shift amount is given to the upper side, where (a) is a graph showing the height correction amount with respect to the welding distance, and (b) is the height direction shift amount. It is a schematic diagram. FIG. 17 is a diagram showing the result of the function confirmation test when the shift amount is given to the lower side, where (a) is a graph showing the height correction amount with respect to the welding distance, and (b) is the height direction shift amount. It is a schematic diagram shown.

  In FIG. 16A and FIG. 17A, the horizontal axis represents the welding distance (unit: mm), and the vertical axis represents the height correction amount (unit: mm). The height correction amount is an integrated correction amount of the height of the welding torch 9 that is input to the robot controller 8. As a result of the function test, when the shift amount is given on the upper side with respect to the shift correction amount of 10.9 mm, the correction as an integrated amount of 11.5 mm is linear when the shift amount is given on the lower side. It was found that the groove up-and-down scanning control functioned effectively by keeping the wire protrusion length constant.

  Next, welding condition control will be described.

  When the groove cross-sectional area, which should be originally uniform, changes due to a taper gap or the like, there arises a problem that the laminated bead height is not uniform across the entire joint line. In order to eliminate welding monitoring, it is important to maintain the weld quality by keeping the laminated bead height uniform. In order to perform constant control of the laminated bead height over the entire joint line, it is first necessary to know the change in the groove cross-sectional area for each cross section along the weld line. For this reason, information on the groove width at the wire tip height and the current laminated bead height is required.

  On the other hand, in the present embodiment, these two pieces of information can be indirectly acquired as follows using groove left / right scanning control and groove vertical scanning control.

  First, by means of groove left / right tracing control using image processing, it is possible to maintain an appropriate weaving width according to the groove width variation, and it is possible to estimate the groove width at the wire tip height based on the value of this weaving width.

  Secondly, by controlling the wire protrusion length by groove up-and-down scanning control using image processing, the wire tip traces the front layer bead, so that the height information of the front layer bead can be obtained.

  In order to adjust the bead cross-sectional area according to the amount of change obtained from these two pieces of information, a method of controlling the welding current (wire feeding speed) can be considered, but in this embodiment, the welding speed is controlled. The method was adopted. If this method is adopted, in addition to the above-mentioned groove lateral scanning control and groove vertical scanning control, three types of control including laminated bead constant control can be realized using only the welded part image sensor. Because it becomes. As a further effect, it is possible to omit pre-touch sensing that is generally used to grasp the reference height of the joint. This eliminates the need for complicated operations associated with touch sensing, saves touch sensing time, and makes it easy to control this stacked bead for not only a 6-axis articulated robot but also a simple orthogonal type automatic welder. Can be applied.

  Hereinafter, the welding condition control for the constant control of the laminated bead height will be specifically described.

  FIG. 18 is a schematic view showing the concept of constant control of the stacked bead height. FIG. 19 is a diagram showing a change in required bead cross-sectional area due to a change in groove width, where (a) shows a change in front bead height, and (b) shows a required bead cross-sectional area immediately after the start of welding. The figure shown, (c) is a figure which shows the required bead cross-sectional area at the time of welding progress.

  In FIG. 18, the code | symbol 51 shows a reference line (groove bottom face). Reference numeral 52 indicates the surface of the bead of the i-1 layer (front layer). Reference numeral 53 indicates the wire tip locus of the i layer (current layer). Reference numeral 54 denotes the surface of the i-layer (current layer) bead. In FIG. 19, reference numeral 54 ′ indicates the surface of the layered bead.

  In the present embodiment, the wire protrusion length of the wire 10 is controlled to be constant by the above-described groove vertical scanning control. As a result, the tip of the wire 10 traces the surface 52 of the front layer bead, and the wire tip locus 53 represents the surface height of the front layer bead over the entire joint line.

  Similarly, in the first layer welding, the wire tip locus 53 represents the height of the groove bottom surface 51 as a result of the wire protrusion length constant control. The computing unit 31 of the personal computer 6 uses this height as a reference for the joint.

  And the calculating part 31 monitors the change after the welding start about the front layer bead surface height from the reference line 51 for every pass after the 2nd layer.

  Two examples of a method for obtaining an appropriate welding speed for making the laminated bead height constant will be described below.

  First, the first method is a method in which a constant bead height is always maintained during welding as compared with the initial state immediately after the start of welding.

  As shown in FIG. 19A, as the deviation amount Δh is generated with the progress of welding, the bead height to be currently increased changes from h immediately after the start of welding to h ′ (= h + Δh).

  As for the weaving width, as shown in FIGS. 19 (b) and 19 (c), as a result of the above-described groove lateral scanning control, the appropriate weaving width is always obtained. The groove widths w and w ′ at the wire tip height at each time point are calculated using this weaving width.

  Here, assuming that the welding speed is controlled to control the welding speed of each pass, the welding speed immediately after the start of welding is S, and the welding speed (appropriate welding speed) at the time of welding progress is S ′. Then, during welding, the formula (1) is established from the relationship in which the amount of deposited metal per unit time is constant.

h · w · S = h ′ · w ′ · S ′ (1) Therefore, an appropriate welding speed for making the laminated bead height uniform is obtained as shown in Equation (2).
S ′ = h · w · S / (h ′ · w ′) = h · w · S / ((h + Δh) · w ′) Equation (2) Equation The calculation unit 31 uses these Equations (1) and (2). Is used to calculate the proper welding speed sequentially.

  Next, the second method is a method of maintaining a constant laminated bead height by monitoring state changes during welding at regular intervals and coping with these changes.

As shown in FIG. 20 (a), when the welding state is sequentially observed at a predetermined interval, the bead height that should be currently increased is welded as the amount of deviation in the height of the previous layer bead occurs as the welding progresses. h ₁ from h ₀ immediately after the start, ··· h _i, changes in _{··· h n-1, h n} . As for the width of the weaving, as shown in FIGS. 20 (b), 20 (c), and 20 (d), the result of the above-described groove lateral scanning control is always an appropriate weaving width. The groove width w _i at the wire tip height immediately after the start of welding and at each time of welding progress is calculated using this weaving width. Then, weaving width in the (n-1) th and the n th monitoring point is respectively determined as w _n-1, w _n. Here, as operating stepwise control of the welding speed for each appropriate interval, the welding speed in the (n-1) th and n-th monitoring point and S _n-1, S _n. Then, during welding, equation (3) is established from the relationship that the amount of deposited metal per unit time is constant.

h _n-1 · w _n-1 · S _n-1 = h _n · w _n · S _n (3) Therefore, the appropriate welding speed to make the laminated bead height uniform is obtained as shown in the following equation (4). It is done.

S _n = h _n-1 · w _n-1 · S _n-1 / (h _n · w _n ) (4) Equation The calculation unit 31 uses the equations (3) and (4) to determine the appropriateness sequentially. Calculate the welding speed.

  Then, the calculation unit 31 inputs a required command relating to the moving speed of the welding torch 9 in the welding direction 18 to the robot controller 8 in order to realize the appropriate welding speed calculated by any method. Thereby, a bead having a constant laminated bead height is formed.

  In this manner, the welding condition (welding speed) control (step S7 in the flowchart of FIG. 4) for controlling the laminated bead height is performed.

  As described above, according to the automatic arc welding system and method of the present embodiment, the laminated beat height constant control including groove left / right scanning control, groove vertical scanning control, and welding condition control is performed on the welded portion. Since it is performed using the image of the welded part imaged by the image sensor (CCD camera 4), the number of necessary sensors can be reduced.

  The automatic arc welding system of the present invention is useful as an arc welding system that performs groove welding of a multi-layer pile.

  The automatic arc welding method of the present invention is useful as an arc welding method for performing multi-layer groove welding.

It is a mimetic diagram showing composition of an automatic arc welding system for carrying out an automatic arc welding method concerning an embodiment of the invention. It is a perspective view which shows the arrangement | positioning position of a CCD camera. It is a perspective view which shows a welding part typically. It is a flowchart which shows the outline | summary of laminated bead height constant control. It is a schematic diagram which shows the shape change of the arc high-intensity part accompanying weaving when the bottom width of a groove | channel is large, Comprising: (a) is a figure which shows the shape of an arc high-intensity part when a wire is located in a weaving center position. (B) is a figure which shows the shape of an arc high-intensity part when a wire is located in a weaving right end, (c) is a figure which shows the shape of an arc high-intensity part when a wire substantially contacts a groove surface. . It is a graph which shows the change of the inclination-angle of the arc high-intensity part major axis with respect to the distance between an electrode and a groove surface in case the groove bottom width is large. FIG. 7 is a perspective view schematically showing a welded portion when the groove bottom width is small. It is a schematic diagram showing the shape change of the arc high intensity part accompanying weaving when the bottom width of the groove is small, (a) is a diagram showing the shape of the arc high intensity part when the wire reaches the left end of the weaving, (B) is a figure which shows the shape of the arc high-intensity part when a wire reaches | attains the weaving center position, (c) is a figure which shows the shape of the arc high-intensity part when a wire reaches | attains the weaving right end. It is a graph which shows the change of the long-short axis length ratio P of the arc high-intensity part with respect to the distance from a groove center. It is a graph which shows the example of a measurement of the change of the major-short axis length ratio P with respect to the distance from the center of a groove | channel. It is the image which shows the image of the welding part imaged with the CCD camera, Comprising: (a) is a welding part image before image processing, (b) is the welding torch tip position and wire tip position extracted by image processing. It is the displayed welded part image. It is an image which shows the process of the image processing which extracts a wire tip position from a welding part image, Comprising: (a) is an image which shows the arc high-intensity part cut out after binarization, and its peripheral region, (b) is (a). (C) is an image obtained by extracting the cutout portion of the arc high-brightness portion of the image of (b), and (d) is an image showing the area center of gravity of the cutout portion in the image of (c). . It is an image which shows the process of the image processing which extracts a welding torch front-end | tip position from a welding part image, Comprising: (a) is a welding part image in which the area | region cut out as a welding torch search area | region was shown, (b) is an image of (a). (C) is an image obtained by improving the image quality of the image of (b), and (d) is an image showing the position of the tip of the welding torch. It is a schematic diagram which shows the welding location in the function confirmation test of groove vertical scanning control. It is a schematic diagram which shows the teaching content in the function confirmation test of groove vertical scanning control. It is a figure which shows the result of the function confirmation test in case shift amount is given to the upper side, Comprising: (a) is a graph which shows a height tracing locus etc. with respect to welding distance, (b) is a schematic diagram which shows height direction shift amount It is. It is a figure which shows the result of the function confirmation test at the time of giving shift amount to the lower side, (a) is a graph which shows height tracing locus etc. with respect to welding distance, (b) is a model which shows height direction shift amount FIG. It is a schematic diagram which shows the concept of laminated bead height constant control. It is a figure which shows the change of required bead cross-sectional area by the fluctuation | variation of the groove width in the 1st method which calculates | requires an appropriate welding speed, (a) is a figure which shows the change of front layer bead height, (b) is a welding start. The figure which shows the required bead cross-sectional area immediately after, (c) is a figure which shows the required bead cross-sectional area at the time of welding progress. It is a figure which shows the change of required bead cross-sectional area by the fluctuation | variation of the groove width in the 2nd method which calculates | requires an appropriate welding speed, (a) is a figure which shows the change of front layer bead height, (b) is a welding start. The figure which shows the required bead cross-sectional area immediately after, (c) is a figure which shows the required bead cross-sectional area in the (n-1) th monitoring point, (d) is the figure which shows the required bead cross-sectional area in the nth monitoring point. is there.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Welding apparatus 2 Robot 3 Automatic arc welding control apparatus 4 CCD camera 5 CCU
6 PC 7 Arm 7a Hand part 8 Robot controller 9 Welding torch 10 Wire 10a Wire tip 11 Welding power source 12 Work piece 13 Groove 14 Arc (arc high brightness part)
15 weld pool 16 backing metal 17 bead 18 welding direction 19 groove width direction 31 calculation unit 32 storage unit 33 input unit 34 display unit 41 welding start teaching point 42 welding end teaching point 51 Reference line (bottom surface of groove)
52 i-1 layer surface 53 wire tip trajectory 54 i layer surface 54 'scheduled lamination bead 101 automatic arc welding system a long axis length b short axis length G center of gravity L long axis RG root gap S, S' welding speed w, w ' Groove width θ Inclination angle

Claims (8)

While the wire is projected from the tip of the welding torch and melted by an arc, the welding torch is moved along the groove extending direction so as to follow the groove from side to side and from above and below the groove. In the welding system that forms the bead first,
An imaging device for imaging the wire and the arc;
Image processing means for performing image processing on the captured video and extracting the protruding length of the wire;
An automatic arc welding system comprising: control means for performing the groove vertical tracing by controlling the extracted protruding length of the wire to be constant.   2. The automatic arc welding system according to claim 1, wherein the imaging device performs the imaging from a front side and a top side in a moving direction along an extending direction of the groove of the welding torch.   The control means controls the welding speed based on the current height of the tip position of the wire and the current data for following the groove left and right, which is strongly related to the bead surface height of the front layer. 2. The automatic arc welding method according to 1. The image processing means extracts a feature of the shape of the high-intensity part of the arc in the image processing,
The automatic arc welding system according to claim 3, wherein the control unit performs the groove left-right copying based on the extracted shape characteristics. While the wire is projected from the tip of the welding torch and melted by an arc, the welding torch is moved along the groove extending direction so as to follow the groove from side to side and from above and below the groove. In the welding method of forming the bead first,
Imaging the wire and the arc;
Processing the imaged video to extract the protruding length of the wire;
Performing the groove up-and-down copying by controlling the extracted protruding length of the wire to be a constant length.   The automatic arc welding method according to claim 5, wherein the imaging is performed from the front and the top in a moving direction along the extending direction of the groove of the welding torch.   6. The automatic control according to claim 5, wherein the welding speed is controlled based on the current height of the tip position of the wire and the current data for scanning the groove laterally, which is strongly related to a bead surface height of a front layer. Arc welding method. Extracting the characteristics of the shape of the high-intensity part of the arc in the image processing;
The automatic arc welding method according to claim 7, wherein the groove left-right copying is performed based on the extracted feature of the shape.