JPH0243580B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Application of the Invention] The present invention relates to a welding/cutting device for a tab plate at a tube end, which welds or cuts a tab plate used when welding the seam portion of a so-called UO steel pipe.

[Background of the invention]

UO steel pipes are made by bending a steel plate into a U-shape using a so-called U-shape press, then forming it into a cylindrical shape using a so-called O-shape press, and welding opposing edges on both sides. In order to properly weld the ends of both side edges (seam welding), a tab plate 12 is attached to the end of the UO steel pipe 10, as shown in FIG. Tab plate 12 is UO steel pipe 10
It is welded along the line connecting points A and B at the end of the line. After the seam portion 14 is welded, it is cut along a line connecting points A and B.
Conventionally, the tab plate 12 was welded or cut manually by an operator. This is because the tab plate attached to the end of the steel pipe that is being transported does not always stay in a constant position, but instead shifts three-dimensionally. Mechanizing the position measurement of the tab plate becomes extremely complicated and time-consuming. This is due to the fact that it was not automated due to increased schedules, lower productivity, and maintenance was not easy. For example, even when recognizing an object on a three-dimensional plane, many drive mechanisms and tactile senses are required, as shown in Japanese Patent Publication No. 55-3118.
Therefore, in order to determine the position of the three-dimensionally displaced tab plate, a visual function more similar to that of humans is required.

By the way, Japanese Patent Publication No. 58-38721 discloses a method in which an object is captured as an image using a camera or the like, and the image is binarized using an image processing device to determine the position and area of the object. However, even if you use an image processing device to measure the position of a large object such as a UO steel pipe, the image is made up of a limited set of pixels, so if you expand the imaging range,
The actual size per pixel increases, and measurement accuracy decreases significantly.

Furthermore, when robots are used to automate the welding and cutting of tab plates, the cylindrical shape of the UO steel pipes is deformed in various ways, even if the tab plate mounting portions of the UO steel pipes have the same nominal diameter. For this reason, the teaching playback type robot has the disadvantage that the locus of welding and cutting by the robot does not correspond to the shape of the UO steel pipe, resulting in defects in welding and cutting.

[Purpose of the invention]

The present invention provides welding and cutting of tube end tab plates that can automatically weld or cut the tab plates of UO steel pipes.
The purpose is to provide a cutting device.

[Summary of the invention]

The present invention provides an imaging means for obtaining an image of a steel pipe having a tab plate attached to the end thereof, and a first method for determining the projected position of the tab plate on a two-dimensional plane by binarizing the image outputted by the imaging means. an image processing means, a surface position detection means for detecting a circumferential surface position of the steel pipe, and a steel pipe position calculation means for calculating a center position and a radius from the circumferential surface position of the steel pipe output by the surface position detection means. , calculating the inclination and position of the tab plate by inputting the projected position of the tab plate on a two-dimensional plane output by the first image processing means and the center position and radius output by the steel pipe position calculation unit; a second image processing means that outputs the three-dimensional position information;
a teaching/storage means that stores arcuate work trajectories of a plurality of tab plates taught in advance; a robot that moves and welds or cuts the boundary between the steel pipe and the tab plate; and the second image processing means. The robot control means inputs the three-dimensional position information of the tab plate from the robot and the arcuate work trajectory of the tab plate from the teaching/storage means and outputs an operation signal for moving the robot.

[Embodiments of the invention]

A preferred embodiment of the tube end tab plate welding/cutting device according to the present invention will be described in detail with reference to the accompanying drawings.

FIG. 1 is a schematic diagram of a welding/cutting device according to the present invention. In Figure 1, roller table 1
A television camera 18 is attached to a pedestal 20 above the camera 6. Lights 22 and 24 provided on the pedestal 20
is adapted to illuminate the end of the UO steel pipe 10 including the tab plate 12.

A robot 26 is installed on the side of the roller table 16.
is mounted on the trolley 28. Robot 26 is
The arm 30 has a torch 32 and a sensor 34 at its tip. The torch 32 is electrically connected to a welding and cutting device 38 via a cable 36. This welding/cutting device 38, like the robot 26, is electrically connected to a robot control device 40. In addition, a trolley 28 on which the robot 26 is mounted
has wheels 44 that roll on rails 42, and can be moved forward and backward relative to the roller table 16 by a shift device 46.

Television camera 18 is electrically connected to image processing device 48 . The image processing device 48 has an image processing main body 50. In addition, the image processing device 4
8 is provided with a monitor television 52 that displays the image of the television camera 18, and a mount 54.
It has a console display 56 and a printer 58 located above.

The roller table 16 described above includes a lower surface measuring device 60 whose details will be described later, and a cut-out tab plate 12.
A chute 62 for receiving the air is provided. The lower surface measuring device 60 is electrically connected to the image processing main body 50 together with the robot control device 40.

As shown in FIG. 2, the image processing device 48 provides position information of the tab plate 12 to the robot control device 40 based on the image from the television camera 18 and the detection signal from the bottom surface measuring device 60. The robot control device 40 operates and controls the robot 26 based on the position information on the tab plate 12 and the teaching content on the teaching box 64. At this time, the position of the tab plate 12 is confirmed by the sensor 34 and input to the robot control device 40. Further, the image processing device 48 receives a feed pack of the robot position from the robot control device 40. The image processing device 48 then controls the shift device 46, the lower surface measuring device 60, and the roller table 16.

The bottom surface measuring device 60 is located at the point O as shown in FIG.
Four push rods 66, 6 arranged radially against
8, 70, 72. These push rods 6
6, 68, 70, and 72 are moved forward and backward by encoder-equipped cylinders 76, 78, 80, and 82 supported by the roller table 16 via a frame 74. That is, the cylinders 76, 7 with encoders
8, 80, 82 are electrically connected to the lower surface measurement control section 84, and move the push rods 66, 68, 70, 72 forward and backward according to the control signal of the lower surface measurement control section 84, and also move the push rods 66, 68, 70, The tip position of 72 is input to the lower surface measurement control section 84.

The cutting of the tab plate 12 according to the embodiment configured as described above is as follows.

The image processing device 40 detects and calculates the position of the tab plate 12, and transfers the coordinate values of the tab plate 12 to the robot control device 40. This coordinate value is expressed as a vector quantity, and the data must be transferred to the robot control device 40 as a vector quantity that the robot control device 40 can recognize. Generally, the robot control device 40 uses a coordinate system based on the direction in which the robot 26 is located, and does not have a function to match this coordinate system with an external coordinate system. Let's find out. Also,
The robot 26 does not operate with an awareness of the absolute coordinate system, but is structured to playback between taught points. Therefore, when operating the robot 26 from the outside, it is necessary to teach the relative movement amount from the taught reference position. That is, teaching is performed for two purposes: firstly, teaching is for the image processing side to know the coordinate system of the robot, and secondly, teaching is for setting a reference point for teaching the robot side the tab position.

A teaching method for knowing the coordinate system of the first robot on the image processing side is performed as follows. First, a cross mark matching the shape (width) of the tab plate 12 is set at a predetermined position on the roller table 16. This mark is, for example, a white cross drawn on a black flat plate. Then, as shown in FIG. 4, for example, the point at the center corner of the cross mark is taught to the robot control device 40, and the data is transferred to the image processing device 48. on the other hand,
The image processing device 48 captures an image of the cross mark with the television camera 18, sends it to the image processing main body 50, binarizes and analyzes the input image in the image processing main body 50, obtains the center point of the cross mark as the image origin, and The pixel size, which is the ratio between the coordinates of the image and the size of the actual measured coordinates, is determined. When the image processing device 48 receives the teaching result on the robot side, it relates the coordinate system on the image processing side and the coordinate system on the robot side using the following coordinate transformation formula.

=F・′+ ……(1) Here, is the position vector in the robot coordinate system ′ is the position vector in the image processing coordinate system,
F is a coordinate rotation matrix, and F is an origin movement amount (vector).

If we give two points corresponding to equation (1), we can get F from the following equation
can be determined.

ΔP＝F・Δ′ ……(2) F＝Δ・(Δ′) ^-1 ……(3) Or, =−F・′ =−ΔP・(Δ′) ^-1・′ ……(4) becomes. As a result, it is possible to obtain F and which determine the conversion formulas for both coordinate systems.

Teaching to set the reference point for teaching the second tab position to the robot side is an operation required when using a teaching playback robot. This is done by teaching basic movements using a tube. Standard tubes with tab plates are prepared for each tube diameter, tab plate width, and plate thickness, and operation reference points are taught and stored in each of the image processing device 48 and the robot control device 40.

When the above teaching is completed, an execution command is input to the image processing main body 50 via the console display 52. The image processing device 48 then drives the conveyance rollers of the roller table 16 to convey the UO steel pipe 10. UO steel pipe 10 is
It is stopped when the end reaches below the television lamella 18 by the operation of a limit switch (not shown) or the like. The television camera 18 images the tab plate 12 and sends an image signal to the image processing main body 50. The image processing main body 50 displays the image on the tab board 12 on a monitor television 52 and also binarizes this image. As shown in FIG. 5, the image captured by the television camera 18 has a large amount of light reflected from the tab plate 12 and the end of the UO steel pipe 10 due to the light from the illumination lights 22 and 24.
A high brightness portion 86 is formed, and the other portion is a low brightness portion 88 where the amount of reflected light is small. The image processing main body 50 binarizes this grayscale image using an appropriate threshold value, and represents the high brightness portion 86 with "1" and the low brightness portion 88 with "0" as shown in FIG. Thereafter, the image processing main body 50 displays the stored standard pattern, for example, the vicinity of the bases Q ₁ and Q ₂ of the tab plate 12 in FIG. 5 with "0" and "1" as shown in FIG. Scan the area in Figure 6 to find the base Q ₁ ,
Find the position of Q ₂ and perform image matching.

On the other hand, the lower surface measuring device 60 includes cylinders 76, 78, 80, 82 with encoders as shown in FIG.
is activated, and the tips of the push rods 66, 68, 70, 72 are brought into contact with the lower surface of the UO steel pipe 10. As described above, the push rods 66, 68, 70, and 72 are arranged radially around the point O, and the distances ₁ , ₂ , ₃ , and ₄ from the point O to the tip of the push rod are the encoder-equipped cylinders 76, 78, 80, and 82 convert it into an electrical signal, and send it to the image processing device 48 via the lower surface measurement control section 84. Image processing device 48
calculates the radius R and central axis O _p of the UO steel pipe 10 in the following manner based on the signal from the bottom surface measuring device 60.

As is well known in geometry, the locus of a circle can be found if there are three coordinate points. Therefore, if it is Honjo, UO can be moved by three push rods.
The radius R of the steel pipe 10 can be determined. However, as shown in Fig. 8 or 9, one of the push rods
may hit the weld bead 90. Therefore, in the case of Fig. 8, a circle C 1 passing through the points P ₁ , P ₂ , P ₄ and a circle C ₂ passing through the points P ₁ , P ₃ , P ₄ which are equal to the outer diameter of the UO steel pipe ₁₀ are obtained. In the case of Fig. 9, UO steel pipe 1
A circle C ₃ passing through points P ₁ , P ₃ , P ₄ equal to the outer diameter of 0,
A circle _C4 passing through points _P1 , _P2 , and _P4 is obtained. When two circles are calculated in this way, the diameters of the two circles are always compared, and the one with the larger diameter is selected to capture the surface of the UO steel pipe 10 and calculate the radius R.

The coordinates (x ₀ , z ₀ ) of the central axis _OP of the UO steel pipe 10 are:
Ask as follows. However, points P ₁ , P ₂ , P ₃ , P ₄
Let the coordinates of (x ₁ , z ₁ ), (x ₂ , z ₂ ), (x ₃ ,
z ₃ ), (x ₄ , z ₄ ). At this time, points P ₁ , P ₂ , P ₄
The center coordinates (x ₀ , z ₀ ) of the circle passing through are becomes. Further, the radius R can be determined by R=√( ₁ , ₀ ) ² +( ₁ , ₀ ) ² (7).

Similarly, the center coordinates (x ₀ , z ₀ ) of the circle passing through points P ₁ , P ₃ , and P ₄ are becomes. The formula for calculating the radius R in this case is:
This is the same as equation (7) above.

In this way, the center coordinates of the UO steel pipe (x ₀ , z ₀ )
Once obtained, the central axis for point O shown in Figure 10 can be determined.
The deviation amounts Δx and Δz of O _P in the x and z directions can be easily determined. And the base of the tab plate 12
Angles β ₁ , β ₂ that Q ₁ , Q ₂ make with the center line L of the UO steel pipe 10
is obtained as follows. The X-direction positions (projected positions on the horizontal plane) of the root portions Q ₁ and Q ₂ shown in FIG. 11 are obtained as distances x _l1 and x _l2 from the reference point OG by the pattern matching method described above. At this time, the distances y ₁ and y ₂ of the tube end from the reference point OG can also be obtained by the pattern matching method. Then, since the deviation amounts Δx and Δz are found, the root part
Shoe distance in the x direction from the central axis O _P of Q ₁ , Q ₂ x _L1 , x _L
2 can be found. Therefore, this x _L1 , x _L2
The angles β ₁ and β ₂ can be determined from the value of and the radius R, and the inclination and position of the tab plate 12 can be recognized.

When the image processing device 48 recognizes the position of the tab plate 12, it inputs the position information to the robot control device 40 and drives the shift device 46 to move the robot 26 forward along the rail 42 and stop it at the working position. . On the other hand, the robot control device 40
drives the arm 30, and detects the end face position of the UO steel pipe 10 by the sensor 34 provided at the tip of the arm 30, as shown by the solid line in FIG. this is,
As described above, when measuring a large object such as the UO steel pipe 10 by image processing, the actual size per pixel becomes large, and the cutting line of the tab plate 12 is determined using the measurement values y ₁ and y ₂ described above. This is because there is a risk that the accuracy will be insufficient if it is determined.

When detecting the end face position of the UO steel pipe 10, the robot control device 40 moves the sensor 34 via the arm 30 as shown in FIG. That is, first, after detecting point A on one side of the tab plate 12, the sensor 34 is moved in a substantially U-shape, and the point A on one side of the tab plate 12 is detected.
Point B on the other side of 2 is detected, and it is determined whether the end face of the UO steel pipe 10 is inclined. When the detection of points A and B is completed, the robot control device 40
As shown by the two-dot chain line in the figure, the torch 32 (cutting torch) is moved onto the cutting line connecting points A and B. Next, the robot control device 40 controls the actual UO steel pipe 1.
0. Select the operation pattern of the robot 26 from the teaching data according to the type of the tab board 12,
The torch 32 is operated according to a predetermined cutting teaching pattern to complete the work. Then, the type of steel pipe, the content of the work, etc. are printed out and recorded by the printer 58.

In addition, when welding the tab plate 12 to the end of the UO steel pipe 10, the tab plate 12 is temporarily attached to the end.
By transporting the UO steel pipe 10 on the roller table 16 and using a welding torch as the torch 32,
This can be done in the same manner as the cutting described above.

As described above, in this embodiment, the image processing device 48 that recognizes the end of the UO steel pipe 10 in a state projected on a plane is equipped with a simple tube position detector that detects the circumferential position of the UO steel pipe 10. By adding the lower surface measuring device 40, it is possible to recognize the tab plate 12 that is positioned while changing three-dimensionally. Therefore, the device for welding and cutting the tab plate 12 can be simplified, the device can be automated, and the time required to measure the tab plate can be shortened. In addition, the sensor 34 provided at the tip of the arm 30 of the robot 26 can compensate for the insufficient accuracy of position recognition of the tab plate 12 by the image processing device 48, so that an accurate locus of welding or cutting can be obtained.

In the above embodiment, the case where the lights 22 and 24 were arranged above the tab plate 12 was explained, but as shown in FIG. Good too.
Further, in the embodiment, the bottom surface measuring device 60
Although the embodiment has been described using a push rod and a cylinder with an encoder, a linear sensor, a laser sensor, or an ultrasonic sensor may also be used. Instead of providing a plurality of sensors, one sensor may be used to measure multiple points in the circumferential direction of the tube, or may be caused to continuously scan in the circumferential direction. By doing this, the recognition accuracy of the tube position can be improved,
Equipment can be simplified. Furthermore, the bottom surface measuring device 60
Since it is desirable to measure the shape near the tube end, the UO steel tube 10 may be configured to be movable in the tube axis direction depending on the stopping position when the UO steel tube 10 is carried in.

〔Effect of the invention〕

As described above, according to the present invention, the tab plate of a UO steel pipe can be automatically welded or cut, and the labor required for welding or cutting the tab plate can be saved.

[Brief explanation of drawings]

Fig. 1 is a schematic configuration diagram of an embodiment of a pipe end tab plate welding/cutting device according to the present invention, Fig. 2 is a block diagram of the embodiment, and Fig. 3 is a detailed diagram of a bottom surface measuring device of the embodiment. , Fig. 4 is a flowchart showing the procedure for coordinate alignment between the robot side and the image processing side, Fig. 5 is an explanatory diagram of an image captured by a television camera, and Fig. 6 is an example of binarization by an image processing device. Figure shown, 7th
The figure shows an example of a binarized standard pattern.
Figures 1 and 9 are explanatory diagrams of how to determine the pipe diameter when the push rod of the bottom surface measuring device hits the weld bead.
Figures 0 and 11 are explanatory diagrams of how to determine the position of the tab plate, Figure 12 is an explanatory diagram of how to detect the tube end face using a sensor installed at the tip of the robot, and Figure 13 is an illustration of how to determine the welding line or cutting line. An explanatory diagram of the method, Fig. 14 is an explanatory diagram of an example of arrangement of lighting for a television camera, Fig. 15 is a perspective view of the end of a UO steel pipe, and Fig. 16 is an explanatory diagram of an example of the arrangement of lighting for a television camera.
FIG. 2 is a plan view showing a part of the end of the UO steel pipe in cross section. 10...UO steel pipe, 12...tab plate, 18...TV camera, 26...robot, 32...torch, 34...
Sensor, 38... Welding/cutting device, 40... Robot control device, 48... Image processing device.

Claims (1)

[Scope of Claims] 1. An imaging means for obtaining an image of a steel pipe having a tab plate attached to the end thereof; and 2.
a first image processing means for determining the projected position of the converted tab plate on a two-dimensional plane; a surface position detection means for detecting the circumferential surface position of the steel pipe; and the steel pipe outputted by the surface position detection means. a steel pipe position calculating means for calculating a center position and radius from a circumferential surface position of the tab plate; and a projected position on a two-dimensional plane of the tab plate outputted by the first image processing means and outputted by the steel pipe position calculating means. a second image processing means that inputs the center position and radius, calculates the inclination and position of the tab plate, and outputs the three-dimensional position information; a robot that moves and welds or cuts the boundary between the steel pipe and the tab plate; and three-dimensional position information of the tab plate from the second image processing unit and the teaching/storage unit. A pipe end tab plate welding/cutting device comprising robot control means for inputting an arcuate working trajectory of the tab plate from the means and outputting an operation signal for moving the robot. 2. The pipe end tab plate welding/cutting device as set forth in claim 1, wherein the robot is provided with a protrusion at its tip on the steel pipe side to abut against and confirm the end surface of the steel pipe. . 3. The surface position detection means includes a plurality of push rods arranged radially, a drive device that moves the push rods forward and backward, and a tip that detects the position of the tip of the push rod and outputs a detection signal to the steel pipe position calculation means. Welding and welding of a tube end tab plate according to claim 1 or 2, characterized in that it is equipped with a position detector.
Cutting device.