JP2019063841A - Automatic welding machine and automatic welding method - Google Patents

Abstract

To provide an automatic welding machine which, even if there is a tack welding bead, identifies the tack weld bead by processing an image captured by a CCD camera, and can perform copy welding to a weldment target line.SOLUTION: A tack welding bead judgement part 141 of an automatic welding machine comprises at least one of the followings: a density value judgement part D which compares a density value ρ of a groove edge region image with a prescribed density threshold, and judges existence or absence of a tack welding bead K1 to K4; an edge angle judgement part A which compares an edge angle θ, which is formed by a tangential line to a border line of an area in which a density is changed in the groove edge region image and a line orthogonal to a groove edge OE52, OE53, with a prescribed edge angle threshold, and judges existence or absence of the tack welding bead K1 to K4; and a standard deviation value judgement part Σ which compares a standard deviation value σ of a detection segment position along the border line with a prescribed standard deviation threshold, and judges existence or absence of the tank welding bead K1 to K4.SELECTED DRAWING: Figure 4

Description

  The present invention relates to an automatic welding machine and an automatic welding method for welding a joint having two beveled edges on which tack welding has been performed.

  If there is a tack weld bead, the weld is affected by the tack weld bead, so it was not easy to weld along the weld line. In this regard, automatic welders have been developed for welding joints having grooved edges on which tack welding has been performed (see, for example, Patent Document 1). This conventional automatic welding machine arranges a total of four welding line detection sensors, two each on the left and right along the welding line, to detect a tack welding bead and to eliminate the influence of the tack welding bead.

Japanese Patent Application Laid-Open No. 9-234565

  The above-mentioned conventional automatic welding machine is provided with a large number of welding line detection sensors (laser type photoelectric sensors), and there is a problem that the automatic welding machine becomes expensive. In addition, it is necessary to process the signals from a large number of detection sensors, and the signal level from the detection sensors becomes low in the welding line, in the tacking weld bead and in the scratched area, so the copying control processing becomes complicated. I will. There is also a problem that the reliability of the automatic welding machine is lowered when the copying control process becomes complicated.

  The present invention has been made in view of the above problems, and an object thereof is to provide an automatic welding machine and an automatic welding method in which a copying control process is simplified.

  The present inventor has completed the present invention in consideration of using an imaging device for welding target line detection in order to simplify the copying control process.

  An automatic welding machine according to the present invention made to solve the problem is a welding robot having a welding torch, and a tack welding which is supported by the welding robot and in which two groove edges are butt-welded and tack-welded in flight. An image processing unit for setting a welding target line by processing an imaging unit configured to capture a grooved edge region image ahead of a portion to be welded of a member to be welded having a bead, and the grooved edge region image captured by the imaging unit And a robot control unit configured to control the welding robot such that the welding torch follows the welding target line set by the image processing apparatus, the image processing apparatus including a tacking weld bead determination unit; A deviation calculating unit for calculating a deviation between the welding target line and a welding scheduled point to be welded by the welding torch; and the temporary welding bead determination unit determines the density value ρ of the groove edge area image A density value determination unit that determines the presence or absence of the tacked weld bead by comparing with a fixed density threshold value, and an approximate straight line and a reference line passing through a plurality of detection segments along a boundary where the density changes in the groove edge area image An edge angle determination unit that determines the presence or absence of the temporary welding bead by comparing an edge angle θ formed with the line with a predetermined edge angle threshold, and intersects the welding target line of a plurality of detection segment positions along the boundary line At least one of a standard deviation value determination unit that determines the presence or absence of the tacking weld bead by comparing the standard deviation value σ of the direction with a predetermined standard deviation threshold value, and the robot control unit determines the tackiness weld bead If it is determined that there is no tacking weld bead in the part and the shift is calculated in the shift calculating section, correction control is performed on the welding torch so as to reduce the shift, and the tacking weld bead is determined in the tacking weld bead determination section. Existence Once, characterized in that it does not perform the correction control.

  The automatic welding method according to the present invention made to solve the problem is an imaging unit supported by a welding robot having a welding torch, in which two groove edges are butt-welded and tack-welded with tack-welding in a fly-over manner. An imaging step of imaging a grooved edge area image ahead of a portion to be welded of the welding target member and a welding target line setting of a welding target line by processing the grooved edge area image imaged in the imaging step And a robot control step of controlling the welding robot such that the welding torch follows the welding target line set in the welding target line setting step, the welding target line setting step includes the tack welding And a deviation calculation step of calculating a deviation between the welding target line and a planned welding point to be welded by the welding torch. In the contact bead determination step, the concentration value determination step of determining the presence or absence of the temporary attachment weld bead by comparing the concentration value ρ of the groove edge region image with a predetermined concentration threshold, the concentration in the groove edge region image An edge angle determination step of determining the presence or absence of the temporary welding bead by comparing an edge angle θ formed by an approximate straight line passing through a plurality of detection segments along the boundary line where the changes and the reference line with a predetermined edge angle threshold value; At least a standard deviation value determining step of determining the presence or absence of the temporary welding bead by comparing a standard deviation value σ in a direction intersecting with the welding target line of a plurality of detection segment positions along the boundary with a predetermined standard deviation threshold In the robot control process, when it is determined that the temporary welding bead is not present in the temporary welding bead determination process and the deviation is calculated in the deviation calculation process, the welding torch is determined to have a small deviation. And correction control to so that, if it is determined that tack welding bead Yes in the tack welding bead determination step, characterized in that it does not perform the correction control.

  By processing the grooved edge area image captured by the imaging unit, it is possible to reliably determine the presence or absence of the tack weld bead and accurately align the welding torch with the welding target line.

BRIEF DESCRIPTION OF THE DRAWINGS It is a schematic block diagram of the automatic welding machine which concerns on Embodiment 1 of this invention. It is an upper surface view of the to-be-welded member 50 of FIG. 1, and is a figure explaining determination of a tack welding bead. It is a flowchart of operation | movement of the automatic welding machine which concerns on Embodiment 1. FIG. FIG. 6 is a partially enlarged view of a beveled edge area image captured by a CCD camera 13; This is a sensing area SE2 schematically showing how to detect the boundary of a figure in normal image processing. It is a figure which shows typically sensing area SE1 which image-processed the grooved edge area | region image and erase | eliminated grooved edge OE52 and OE53. It is a figure which shows typically sensing area SE2 which image-processed the grooved edge area | region image and erase | eliminated grooved edge OE52 and OE53. It is a figure which shows typically a mode that the presence or absence of a tack welding bead is determined by the standard deviation value of detection segment S in sensing area SE1 without a tack welding bead. It is a figure which shows typically a mode that determination of the presence or absence of a tack welding bead is carried out by the standard deviation value of detection segment S in sensing area SE2 with a tack welding bead. In sensing area SE1 without a tack weld bead, it is shown schematically how the presence or absence of a tack weld bead is determined by an angle formed by an approximate straight line of a detected segment S along a boundary where concentration changes and the reference line Lb. FIG. In sensing area SE2 with a tack weld bead, it is shown schematically how the presence or absence of a tack weld bead is judged by the angle formed by the approximate straight line of the detection segment S along the boundary where the concentration changes and the reference line Lb. FIG. It is a figure which shows the operation screen of the automatic welding machine which concerns on Embodiment 4. FIG. It is a figure which shows the to-be-welded test piece of verification experiment 1. FIG. It is a graph of the result of verification experiment 1, and shows welding line following accuracy when not operating a tack welding bead judging part. It is a graph which shows welding-line follow-up accuracy when a tacking welding bead determination part is operated by the result of verification experiment 1, and a tacking area | region is determined by three, a density value, an edge angle, and a standard deviation. It is a figure which shows the to-be-welded test piece of verification experiment 2. FIG. It is a figure which shows the method of verification experiment 2. FIG. It is a screen to input each threshold of concentration, standard deviation and edge angle to make it possible to judge the presence or absence of tacked weld bead in verification experiment 2, and displays the current value of concentration value ρ, standard deviation value σ and edge angle θ It is also a screen to It is a result of verification experiment 2, and is a graph which shows welding target line follow-up accuracy when only a concentration value judgment part is operated. It is a graph of the result of verification experiment 2, and shows welding line following accuracy when only a standard deviation value judgment part is operated. It is a graph which shows the welding target line follow-up precision at the time of operating only an edge angle judgment part by the result of verification experiment 2. FIG. It is a graph which shows the welding target line follow-up precision at the time of making a density | concentration value determination part or a standard deviation value determination part operate by the result of verification experiment 3. FIG. It is a graph which shows the welding target line follow-up precision at the time of making a density | concentration value determination part or an edge angle determination part operate by the result of verification experiment 4. FIG. It is a graph which shows the welding target line follow-up precision at the time of operating a standard deviation value judging part or an edge angle judging part by the result of verification experiment 5. It is a figure which shows the to-be-welded test piece of verification experiment 6. FIG. It is a graph which shows the welding target line tracking accuracy by the result of verification experiment 6.

Hereinafter, embodiments of the present invention will be described in detail with reference to the attached drawings.
(Embodiment 1)
As shown in FIGS. 1, 2 and 4, the automatic welding machine according to the present embodiment is supported by the welding robot 11 having the welding torch 12 and the welding robot 11, and two groove edges OE52 and OE53 are abutted and fly off. An imaging unit 13 for imaging a grooved edge area image ahead of the welding portion of the welded member 50 having the tacked weld beads K1 to K3 tacked and welded to each other, and the grooved edge area image imaged by the imaging unit 13 Image processing device 14 for processing welding to set the welding target line WL, and a robot control unit 15 for controlling the welding robot 11 so that the welding torch 12 follows the welding target line WL set by the image processing device 14, The processing device 14 includes a temporary welding bead determination unit 141 and a deviation calculation unit 142 for calculating the deviation between the welding target line WL and the welding scheduled point P welded by the welding torch 12. . When the temporary welding bead determination unit 141 determines that the temporary welding bead is not present and the deviation calculating unit 142 calculates the deviation, the robot control unit 15 performs correction control on the welding torch 12 so that the deviation becomes small. If it is determined by the tack welding bead determination unit 141 that the tack welding bead is present, the correction control is not performed. Further, the temporary welding bead determination unit 141 of the present embodiment compares the density value ρ of the grooved edge area image captured by the imaging unit 13 with a predetermined density threshold to determine the presence or absence of the temporary welding beads K1 to K3. And a density value determination unit D (not shown).

  The exact welding target line WL is a middle line of the groove edge OE52, OE53 as shown in FIG. FIG. 4 schematically shows an enlarged view, and in fact, the gap between the cross-hatched grooved edge OE52 and OE53 is narrow, and the welding target line WL substantially coincides with the grooved edge OE52, OE53. .

  The laser device 16 is connected to the welding torch 12 by an optical waveguide 17. A condensing lens system is incorporated in the welding torch 12. In the case of the arc welding torch, for example, the welding torch 12 is connected to the arc welding power supply 16 by the electric wire 17. The imaging unit 13 is, for example, a CCD camera, and the image processing apparatus 14 includes a CPU.

  As shown in FIG. 2, two members of the metal plates 52 and 53 are butted and welded to each other in three places. In FIG. 2, K1 to K3 are temporary welding beads, and Wb indicates a welding bead welded by the welding torch 12 of the automatic welding machine according to the present embodiment.

  FIG. 4 is a partially enlarged view of a grooved edge area image captured by the CCD camera 13. For convenience of explanation, the surface of the metal plates 52 and 53 has a high density value (whiteish), so a one-dot hatching is applied. The surface of the tack weld bead K1 is shaded (solid line) because its density value is lower than that of the surfaces of the metal plates 52 and 53 (greyish). The gap between the grooved edge OE 52 of the metal plate 52 and the grooved edge OE 53 of the metal plate 53 hardly reflects light, so the density becomes low (blackish). Therefore, the cross hatching of a solid line is given to this gap. In FIG. 4, rectangular areas SE1 and SE2 are sensing areas, and process image data in this rectangular area. As an example, the outer diameter of the temporary welding bead K1 to K3 is on the order of 3 mm, the sensing areas SE1 and SE2 have a size of about 3 mm × 3 mm, and the gap between the groove edge OE52 and OE53 is on the order of 0.1 mm. is there. Further, one side intersecting in the rectangular sensing areas SE1 and SE2 intersects with the welding target line WL, and the other side is along the welding target line WL.

  As shown in FIG. 4, in the sensing area SE1 in which there is no tack welding bead K1 to K3, the line connecting the detection segments s1 and s2 along the boundary line (groove edge OE52) at which the concentration changes is the groove edge OE52. Since it becomes parallel, the line which ties s1 and s2 can be made into welding target line WL. However, as shown in FIG. 5, in the sensing area SE2 in which the tack weld bead K1 is present, the detection segment sn, sn + 1 follows the groove edge OE52, but the detection segment sn + 2, sn + 3, sn + 4 ... comes along the contour of the tack welding bead K1. Then, the contour of the tack weld bead K1 is erroneously detected as a welding target line, and welding is performed so as to follow the erroneously detected welding target line. So, the automatic welding machine of this embodiment was made not to set a welding target line in a sensing area with a tack welding bead. That is, the automatic welding machine of the present embodiment determines the presence or absence of the tacking weld bead in the sensing area, and controls the welding torch 12 so as to be along the preset welding scheduled line, when the tacking weld bead is present. Do. The code | symbol P of FIG. 4 is a welding planned point welded by the welding torch 12.

  The density value determination unit D of the present embodiment compares the predetermined density threshold value with the average density value aveave of the sensing areas SE1, SE2, ... of the groove edge area image, and the presence or absence of the tack weld bead K1 to K3. Is determined as follows. First, when the interval between the groove edge OE 52 and the OE 53 is smaller than a predetermined value, the image processing device 14 performs image processing so that the groove edges OE 52 and OE 53 can not be visually recognized. That is, the image processing device 14 performs processing so as to be displayed as black as the distance between the groove edge OE52 and the OE53 increases and as to be displayed as lighter as the distance from the groove edge OE52 to the OE53 decreases. Then, if the interval between the groove edge OE52 and the OE53 is equal to or less than the reference value, the welding target line WL is not displayed. Therefore, the density value determination unit D is configured to display only the tack welding bead K1 so that the welding target line WL is not displayed by the image processing, and determines the presence or absence of the tack welding bead K1.

  FIG. 6 shows a sensing area SE1 in which the concentration value determination unit D prevents the welding target line WL from being displayed by the input of a predetermined concentration reference value. FIG. 7 shows a sensing area SE2 in which the density value determination unit D performs image processing so that the welding target line WL is not displayed by the input of a predetermined concentration reference value. Since the welding target line WL is not displayed in this manner, the presence or absence of the tack welding bead K1 can be reliably determined.

  For example, assuming that white and black densities are divided into 256, white and white are 255, and true black is 0, as described above, the surfaces of the metal plates 52 and 53 are close to white and the density value mm is large. The surface of the tack weld bead K1 is near gray and the concentration value k k is smaller than the concentration value m m of the surface of the metal plates 52, 53. That is, ρ k <m m. Therefore, assuming that the average concentration value of the entire sensing area SE2 having the tack weld bead K1 of FIG. 7 is aveave, ρk ≦ ρave <ρm. Therefore, assuming that predetermined density thresholds are Dk and mm (> ρk), the density value determination unit D sets at least one of 平均 k ≦ ρave and セ ン シ ン グ ave <の m as the average density value aveave of the sensing area of the beveled edge area image. When satisfied, it can be determined that there is a tack weld bead. By doing this, it is possible to stably determine the presence or absence of the tack weld bead.

  As a result of experiment with welded members having various grooved edges and tacked weld beads, when the black and white density of grooved edge image is divided into 256, white is 255 and black is 0, ρ k <20.0 , Mm> 115.0. Therefore, the concentration value determination unit D may determine that there is a tack weld bead when the average concentration value aveave is 20 ≦ ρave and / or 又 は ave <115.

  Next, the operation of the automatic welding machine of the present embodiment will be described with reference to the flowchart of FIG. First, the temporarily welded workpiece 50 is placed on the table 20, and the welding robot 11 is taught so that the welding torch 12 follows a planned welding line connecting the welding start point Ws and the welding end point We, and welding is performed. Start. When welding starts, in step 100, the CCD camera 13 acquires a beveled edge area image (an image of an area including the beveled edges OE 52 and OE 53). Next, at step 110, the image processing device 14 acquires groove edge OE 52, OE 53 (-welding target line WL) coordinates. Next, in step 120, the tack welding bead determination unit 141 determines the presence or absence of the tack welding beads K1 to K3. When it is determined in step 120 that the tacking weld bead is not present, the deviation calculation unit 142 is to perform welding processing at the welding edge 12 with the groove edge OE 52, OE 53 (~ welding target line WL) coordinate position and the welding torch 12 in step 130 The deviation (difference) from P is calculated and determined. If it is determined in step 130 that the difference is large, the robot control unit 15 performs correction control so as to reduce the difference in step 150, and welding is continued in step 160. If it is determined in step 120 that the tacked weld bead is present, the robot control unit 15 stops copying to the welding target line WL in step 140, proceeds to step 160, and welds connecting the welding start point Ws and the welding end point We It is welded to follow the schedule line.

Second Embodiment
In the automatic welding machine of this embodiment, the temporary welding bead determination unit 141 of the automatic welding machine of Embodiment 1 includes a standard deviation value determination unit 部 (not shown).

  FIG. 8 shows the state of standard deviation determination in the sensing area SE1. Since there is no tack welding bead K1 in the sensing area SE1, the detection segments s1, s2,... Sn follow the boundary (groove edge OE 52) at which the concentration changes. Therefore, the standard deviation value σse1 is small when the x coordinate (detection segment position) of each detection segment s1, s2, ... sn is x1, x2, ... xn.

  On the other hand, in the case of the sensing area SE2 shown in FIG. 9, the detection segments sn, sn + 1, sn + 2 are along the boundary (the groove edge OE 52) where the concentration changes, and the detection segments sn + 3, sn + 4, sn +5 is along the boundary line of the tack welding bead K1. Therefore, when the x-coordinates of the detection segments sn, sn + 1, sn + 2, sn + 3, sn + 4, and sn + 5 are xn, xn + 1, xn + 2, ... xn + 5 The standard deviation value σse2 becomes larger, and σse2> σse1.

  When the standard deviation value σ is less than a predetermined standard deviation threshold value σl, the standard deviation value determination unit Σ of this embodiment determines that there is no tacked weld bead.

  As a result of experiment with welded members having various groove edges and tack welding beads, it was found that the presence or absence of tack welding beads can be accurately determined when σ 1 = 50. Therefore, the standard deviation value determination unit Σ of the automatic welding machine in the modified aspect determines the presence or absence of the tacked weld bead as σ 1 = 50.

(Embodiment 3)
In the automatic welding machine of this embodiment, the temporary welding bead determination unit 141 of the automatic welding machine of Embodiment 1 includes an edge angle determination unit A (not shown).

  FIG. 10 shows the state of edge angle determination in the sensing area SE1. Since there is no tack welding bead K1 in the sensing area SE1, the detection segments s1, s2, s3, s4 are along the boundary line (groove edge OE 52) where the concentration changes. Therefore, the edge angle θ formed by the approximate straight line SL1 connecting s1 and s4 and the reference line Lb is close to 90 °.

  On the other hand, in the case of the sensing area SE2 shown in FIG. 11, the edge angle θ formed by the reference line Lb and the approximate straight line SL2 connecting the detection segments sn and sn + 1 is larger than 90 °. So, in this embodiment, 90 degrees is made into a predetermined | prescribed edge angle threshold value.

  In the present embodiment, the reference line Lb is a line (X axis) orthogonal to the groove edges OE 52 and OE 53, but a side parallel to the X axis of the rectangular sensing area may be used as the reference line.

  The edge angle determination unit A of the present embodiment determines that there is a tack weld bead when θ is greater than 90 °, and determines that there is no tack weld bead when θ is 90 °.

  Assuming that predetermined edge angle thresholds are θs and θl (> θs), the edge angle determination unit A determines that there is no tack weld bead when the edge angle θ of the grooved edge area image is θs or more and / or θl or less. It may be determined. By doing this, it is possible to stably determine the presence or absence of the tack weld bead.

  As a result of experiment with welded members having various groove edges and tack welding beads, it is possible that the edge angle judgment unit A can surely judge the presence or absence of the tack welding bead when θs = 75 ° and θl = 105 °. all right. Therefore, the edge angle determination unit A may determine that no tacking weld bead is present when θ is 75 ° or more and / or 105 ° or less.

(Embodiment 4)
In the automatic welding machine according to the present embodiment, the temporary welding bead determination unit 141 of the automatic welding machine according to the first embodiment includes the density value determination unit D, the standard deviation value determination unit Σ, and the edge angle determination unit A. It is.

  FIG. 12 shows an operation screen of the automatic welding machine according to the present embodiment. Since the button surrounded by the dotted circle is selected in FIG. 12, step 140 (when it is determined that the tack weld bead is present even if any one of the concentration value, the standard deviation value, or the edge angle is present). (See FIG. 3). In the example of FIG. 12, all of the concentration value, standard deviation value, and edge angle are determined to be no tack welding bead (since there is no display in the tack identification determination of the state on the operation screen), step 130 ( Go to Figure 3).

  In the automatic welding machine according to the present embodiment, it is possible to identify the tacked weld bead in various situations by freely combining the concentration value, the standard deviation value, and the edge angle by “AND” or “OR”, and the welding robot It leads to the stabilization of welding construction by For example, if identification of a tack welding bead is performed by any one of concentration value, standard deviation value, and edge angle, identification time can be shortened, but accuracy and reliability can not be improved. On the other hand, if identification of a tack weld bead is performed by three values of concentration value, standard deviation value, and edge angle, the identification time will be long, but the probability of false identification will be low and the accuracy and reliability of identification will be high. Depending on the type of welding work, there may be cases where identification of the tack weld bead may be performed by any one of concentration value, standard deviation value, and edge angle, but there are three cases of concentration value, standard deviation value, and edge angle. It may be necessary to identify the tack weld bead. Welding can be stably performed by freely combining the concentration value, the standard deviation value, and the edge angle with “AND” or “OR” according to the type of welding work.

Embodiment 5
The automatic welding method of the present embodiment is a tack welding bead K1 in which two grooved edges OE 52 and OE 53 are butt-welded and fly-tack welding is performed by a CCD camera 13 supported by a welding robot 11 having a welding torch 12. An imaging process P1 for imaging the grooved edge area image of the welding target member 50 having ~ K3 and a welding target which is an intermediate line between the grooved edges OE52 and OE53 by processing the grooved edge area image imaged in the imaging process P1 Welding target line setting process P2 for setting the line WL, and robot control process P3 for controlling the welding robot 11 so that the welding torch 12 follows the welding target line WL set in the welding target line setting process P2, The target line setting process P2 is a welding process using a temporary welding bead determination process P4 for determining the presence or absence of the temporary welding beads K1 to K3 and the welding target line WL and the welding torch 12 The temporary welding bead determination step P4 compares the density value ρ of the grooved edge area image captured in the imaging step P1 with a predetermined density threshold value. Then, a concentration value determination step P41 for determining the presence or absence of the tacking weld bead K1 to K3 and a plurality of detection segments s1 and s2 along the boundary line where the concentration changes in the grooved edge area image captured in the imaging step P1. · · Edge angle determination step P42 which determines the presence or absence of the tack weld bead K1 to K3 by comparing the edge angle θ formed by the approximate straight line passing through sn and the reference line Lb with a predetermined edge angle threshold, along the boundary · · · · · Standard deviation of the detection segments s 1, s 2 · · · · · · · · · Standard deviation in the direction intersecting with the welding target line WL in the sn position is compared with a predetermined standard deviation threshold value Value format The robot control process P3 includes at least one process P43, and the robot control process P3 shifts the welding torch 12 when it is determined that the temporary weld bead is absent in the temporary weld bead determination process P4 and the shift is calculated in the shift operation process P5. The correction control is performed so as to be small, and when it is determined that the temporary welding bead is present in the temporary welding bead determination process P4, the correction control is not performed.

  In the automatic welding method of the modified embodiment, when the robot control process P3 does not perform the correction control, the welding torch 12 is controlled to follow a preset welding schedule line. In the automatic welding method of another modification, when the robot control process P3 does not perform the correction control, the welding torch 12 is controlled to follow the extension line of the welding target line WL immediately before.

(Verification experiment 1)
The followability of the welding torch to the welding target line WL when the temporary welding bead determination unit 141 determines the presence or absence of the temporary welding bead based on the concentration value ρ, the standard deviation value σ, and the edge angle θ without emitting a laser beam. I examined. That is, the tracking accuracy to the welding target line WL was compared in the case where the judgment function of the tack weld bead was exhibited using the butt joint material of the SPCC steel plate having a plate thickness of 2.6 mm and in the case where it was not exhibited.

  The test piece is shown in FIG. Two SPCC steel plates with a thickness of 2.6 mm, a length of 410 mm, and a width of 30 mm were used as butt joints, and the welding torch was run from the welding start point Ws of the test piece to the welding end point We. The tack welding beads K1, K2 and K3 are formed by laser spot welding and are round with a diameter of 3.2 mm. In addition, in this verification experiment, it experimented as the groove gap 0 with the butt joint which does not have a groove. The welding robot was taught to intentionally shift the planned welding line with respect to the welding target line WL so that the welding detachment point is 0 mm at welding start point Ws and 2 mm at welding completion point We. .

  FIG. 14 shows a change in the amount of deviation from the welding target line WL of the welding torch 12 when the temporary welding bead determination unit 141 is not functioned. A maximum deviation of 0.7 mm has occurred.

  FIG. 15 shows a change in the amount of deviation from the welding target line WL of the welding torch 12 when the temporary welding bead determination unit 141 is made to function. From this, it can be understood that the maximum value of the amount of deviation from the welding target line WL of the welding torch 12 when the tack welding bead determination unit 141 is made to function is 0.095 mm. The amount of deviation does not cause a problem in securing a good welding quality when performing laser welding.

(Verification experiment 2)
When laser welding is performed, and the temporary welding bead determination unit 141 determines the presence or absence of the temporary welding bead by any of the concentration value ρ, the standard deviation value σ, and the edge angle θ, the welding target wire WL to the welding bead is The followability was examined. That is, using the butt joint material of the SPCC steel plate having a plate thickness of 1.2 mm, the temporary welding bead determination unit 141 uses any one of the concentration value 、, the standard deviation value σ, and the edge angle θ The tracking accuracy of the welding torch to the welding target line WL when the presence or absence of the welding torch was determined was examined.

  The test specimen is shown in FIG. Two SPCC steel plates with a thickness of 1.2 mm, a length of 320 mm and a width of 30 mm were butt-welded and tack welded at three points at a pitch of 100 mm. In this experiment, the experiment was conducted by setting the groove gap to 0 with a butt joint having no groove.

  The laser power is 2 KW and the welding speed is 3 m / min. In the welding robot, the planned welding line is intentionally shifted with respect to the welding target line WL so that the welding start point Ws will have a Nerai displacement of 0 mm and the welding end point We will have a Nerai displacement of 1.5 mm. I was teaching. After welding, using a microscope, the position of the welding target line WL and the welding bead center position after welding were measured to evaluate the copying accuracy. After welding of seven 30 mm lengths at a pitch of 40 mm for a total welding length of 300 mm, the deviation δ between the position of the welding target line WL and the weld bead center position after welding was measured (see FIG. 17). .

  In order to determine the presence or absence of the tack weld bead, threshold values of density value 値, standard deviation value σ, and edge angle θ were input. The input screen is shown in FIG. When the concentration value ρ is greater than or equal to the concentration threshold 20 and less than or equal to 115, it is determined that the tacked weld bead is present. When the standard deviation value σ is equal to or greater than the standard deviation threshold 50, it is determined that the tacked weld bead is present. When the edge angle θ is equal to or greater than an edge angle threshold of 75 ° and 105 ° or less, it is determined that the tacked weld bead is not present. That is, when the edge angle θ is less than 75 ° and exceeds 105 °, it is determined that the tacked weld bead is present.

  FIG. 19 shows the amount of deviation from the welding target line WL of the weld bead when the presence or absence of the tack weld bead is determined by the concentration value 単 独 alone. The average displacement amount is 0.083 mm, and a scanning displacement of a maximum of +0.120 mm and a minimum of 0.02 mm is generated with respect to the coordinate position of the welding target line WL. The scanning deviation is +0.120 mm at the maximum, and the judgment of only the concentration value satisfies the scanning accuracy of ± 0.15 mm, which enables good welding by laser welding.

  FIG. 20 shows the amount of deviation from the welding target line WL of the weld bead when the presence or absence of the tack weld bead is determined with the standard deviation value σ alone. The average displacement amount is 0.087 mm, and a scanning displacement of a maximum of +0.130 mm and a minimum of 0.05 mm occurs with respect to the coordinate position of the welding target line WL. Even with the determination of only the standard deviation value, the scanning accuracy of ± 0.15 mm, which enables good welding by laser welding, is satisfied.

  FIG. 21 shows the amount of deviation from the welding target line WL of the weld bead when the presence or absence of the tack weld bead is determined by the edge angle θ alone. The average displacement amount is 0.080 mm, and a scanning displacement of a maximum of +0.11 mm and a minimum of 0.05 mm occurs with respect to the coordinate position of the welding target line WL. Even with the determination of only the edge angle, the scanning accuracy of ± 0.15 mm, which enables good welding by laser welding, is satisfied.

(Verification experiment 3)
The followability of the welding bead to the welding target line WL was examined when the tacking weld bead determination unit 141 determined the presence / absence of the tacking weld bead by “OR” of the concentration value ρ and the standard deviation value σ. . The test pieces, the experimental conditions, and the like are the same as those in Verification Experiment 2, and thus the description thereof is omitted.

  An experimental result is shown in FIG. The average shift amount is 0.052 mm, and a scan shift of +0.130 mm at the maximum and 0.06 mm at the minimum with respect to the coordinate position of the welding target line WL occurs. Even when the concentration value ρ and the standard deviation value σ are combined by “OR”, the scanning accuracy of ± 0.15 mm, which enables good welding by laser welding, is satisfied without being affected by the disturbance due to the temporary welding bead. .

(Verification experiment 4)
The tackiness of the weld bead to the weld target line WL was examined when the tack weld bead determination unit 141 determined the presence / absence of the tack weld bead by “OR” of the density value ρ and the edge angle θ. The test pieces, the experimental conditions, and the like are the same as those in Verification Experiment 2, and thus the description thereof is omitted.

  The experimental results are shown in FIG. The average displacement amount is 0.088 mm, and a scanning displacement of a maximum of +0.140 mm and a minimum of 0.04 mm occurs with respect to the coordinate position of the welding target line WL. Even when the concentration value ρ and the edge angle θ are combined by “OR”, the scanning accuracy of ± 0.15 mm, which enables good welding by laser welding, is satisfied without being influenced by the disturbance due to the temporary welding bead.

(Verification experiment 5)
The followability of the weld bead to the weld target line WL was examined when the tack weld bead determination unit 141 determined the presence or absence of the tack weld bead by “OR” of the standard deviation value σ and the edge angle θ. .

  An experimental result is shown in FIG. The average shift amount is 0.083 mm, and a scan shift of +0.120 mm at the maximum and 0.05 mm at the minimum with respect to the coordinate position of the welding target line WL occurs. Even when the standard deviation value σ and the edge angle θ are combined by “OR”, the scanning accuracy of ± 0.15 mm, which enables good welding by laser welding, is not affected by the disturbance due to the temporary welding bead. .

(Verification experiment 6)
The followability of the weld bead to the welding target line WL was examined when the tack weld bead determination unit 141 determined the presence or absence of the tack weld bead based on the concentration value ρ, the standard deviation value σ, and the edge angle θ.

  The test specimens are shown in FIG. Two SPCC steel plates having a thickness of 1.2 mm, a length of 610 mm, and a width of 30 mm were butt-welded and tack welded at four points at a pitch of 150 mm. In this experiment, the experiment was conducted by setting the groove gap to 0 with a butt joint having no groove.

  The laser power is 2 KW and the welding speed is 3 m / min. The welding robot was taught to be intentionally shifted with respect to the welding target line WL so that the welding removal point is 0 mm at the welding start point Ws and 2 mm at the welding completion point We. After welding, using a microscope, the position of the welding target line WL and the welding bead center position after welding were measured to evaluate the copying accuracy. After welding at 15 points and 30 mm at a pitch of 40 mm for a total welding length of 600 mm, the deviation δ between the position of the welding target line WL and the weld bead center position after welding was measured (see FIG. 17). .

  In order to determine the presence or absence of the tack weld bead, threshold values of density value ρ, standard deviation value σ, and edge angle θ were input (see FIG. 18). When the concentration value ρ is greater than or equal to the concentration threshold 20 and less than or equal to the concentration threshold 115, it is determined that the tacked weld bead is present. When the standard deviation value σ is equal to or greater than the standard deviation threshold 50, it is determined that the tacked weld bead is present. When the edge angle θ is equal to or greater than the edge angle threshold 75 ° and equal to or less than the edge angle threshold 105 °, it is determined that the tacked weld bead is not present. That is, when the edge angle θ is less than 75 ° and exceeds 105 °, it is determined that the tacked weld bead is present.

  When the tack welding bead determination unit 141 determines the presence or absence of the tack welding bead by combining the concentration value ρ, the standard deviation value σ, and the edge angle θ by “OR”, the welding target line WL to the welding bead is determined. The followability was examined.

  The experimental results are shown in FIG. The average shift amount is 0.050 mm, and a scan shift of +0.140 mm at the maximum and −0.035 mm at the minimum occurs with respect to the coordinate position of the welding target line WL. The scanning accuracy of ± 0.15 mm, which enables good welding by laser welding, without being affected by the disturbance caused by the tack welding bead, is satisfied.

  The welding speed is 1 m / min (laser power: 1.2 KW), 5 m / min (laser power: 3 KW), 7 m / min (laser power: 3.5 KW), 10 m / min (laser power: 5 KW) Experiment was conducted.

  The average deviation at each welding speed is 0.041 mm at welding speed 1 m / min, 0.062 mm at welding speed 5 m / min, 0.082 mm at welding speed 7 m / min, and 0.045 mm at welding speed 10 m / min Met. It was found that the scanning accuracy of ± 0.15 mm, which enables good welding by laser welding under any welding conditions, is satisfied.

  The present invention is not limited to such embodiment and verification experiment, and square joints other than butt joints, lap joints, fillet joints, lap joints, T joints, flare joints, edge joints, etc. Applicable to The present invention can also be applied to overlaying during overlaying and brazing other than welding.

11 ··········································································································· The image processing apparatus 141 ........ with pre-welded bead determination unit 142 ........ deviation calculating section 15 .......... robot control unit 50 ..... ----- weld members D ........... density value determination unit sigma ........... standard deviation judgment unit A ········ ············································································································································································································································································ ... Density value θ ... ... ... ... Edge angle σ ... ... ... Standard deviation value

Claims (6)

A grooved edge ahead of a welding portion of a welding robot having a welding torch and a welding member supported by the welding robot and having a tack welding bead supported by the welding robot and having two beveled edges butt-welded and tack-welded An imaging unit configured to capture an area image; an image processing apparatus configured to set a welding target line by processing the grooved edge area image captured by the imaging unit; and the welding target line set by the image processing apparatus. A robot control unit that controls the welding robot such that a torch follows the robot;
The image processing apparatus includes a tacking weld bead determination unit, and a shift calculation unit that calculates a shift between the welding target line and a planned welding point to be welded by the welding torch.
The temporary welding bead determination unit compares the density value ρ of the groove edge region image with a predetermined concentration threshold to determine the presence or absence of the temporary welding bead; Edge angle determination that determines the presence or absence of the tack weld bead by comparing an edge angle θ formed by an approximate straight line passing through a plurality of detection segments along the boundary where the concentration changes and the reference line with a predetermined edge angle threshold And a standard deviation value determination to determine the presence or absence of the tackiness weld bead by comparing a standard deviation value σ in a direction intersecting the weld target line of a plurality of detection segment positions along the boundary with a predetermined standard deviation threshold value Equipped with at least one of the
The robot control unit corrects and controls the welding torch so as to reduce the displacement when the tacking weld bead determination unit determines that there is no tacking weld bead and the displacement calculation unit calculates a displacement. The automatic welding machine according to claim 1, wherein the correction control is not performed when it is determined that the temporary welding bead is present in the temporary welding bead determination unit.   Assuming that the predetermined concentration threshold is kk and mm (> ρk), the concentration value determination unit determines that the concentration value ρ of the groove edge region image satisfies at least one of kk ≦≦ and ρ <ρm. The automatic welding machine according to claim 1, wherein it is determined that the attached welding bead is present.   Assuming that the predetermined edge angle threshold is θs and θl (> θs), the edge angle determination unit determines that the tacking weld bead is absent when the edge angle θ is θs or more and / or θ1 or less. The automatic welding machine according to 1 or 2. In an imaging unit supported by a welding robot having a welding torch, a groove edge area ahead of a portion to be welded of a workpiece to be welded having a tack weld bead in which two groove edges are butted and tack-welded temporarily An imaging process for capturing an image;
A welding target line setting step of setting a welding target line by processing the grooved edge area image picked up in the imaging step;
A robot control step of controlling the welding robot such that the welding torch follows the welding target line set in the welding target line setting step;
The welding target line setting step is a temporary welding bead determination step of judging presence / absence of the temporary welding bead, and a deviation operation for calculating a deviation between the welding target line and a welding scheduled point to be welded by the welding torch. Process, and
The tacking weld bead determination step is a concentration value determination step of determining the presence or absence of the tacking weld bead by comparing the density value ρ of the grooved edge area image with a predetermined concentration threshold value; Edge angle determination that determines the presence or absence of the tack weld bead by comparing an edge angle θ formed by an approximate straight line passing through a plurality of detection segments along the boundary where the concentration changes and the reference line with a predetermined edge angle threshold Determining the presence or absence of the tack weld bead by comparing a standard deviation value σ of the plurality of detection segment positions along the boundary in the direction intersecting the weld target line with a predetermined standard deviation threshold value Comprising at least one of the steps
The robot control process corrects and controls the welding torch so as to reduce the deviation when it is determined that there is no tacking weld bead in the tacking weld bead determination step and the deviation is calculated in the deviation calculation step. The automatic welding method characterized in that the correction control is not performed when it is determined that the temporary welding bead is present in the temporary welding bead determination step.   5. The automatic welding method according to claim 4, wherein the robot control step controls the welding robot such that the welding torch follows a predetermined welding scheduled line when the correction control is not performed.   The automatic welding method according to claim 4, wherein the robot control step controls the welding robot such that the welding torch follows an extension line of a previous welding target line when the correction control is not performed.

Images