JPH0780643A - Control method of welding robot - Google Patents

Abstract

PURPOSE:To obtain good and stable quality welding bead by sensing welding gap as using laser sensor supported with a welding robot and controlling the welding condition based on the obtained data. CONSTITUTION:Objects to be welded A, B are butt welded through a gap G of a gap width of g(X) by robot welding. In this case, a welding torch 2 and laser sensor 3 are mounted at a robot arm tip part 1, sucessively, in the region preceding to a welding point 4, by scanning with a scanning beam 5, the locus 6A-6B of laser scanning on the objects A, B is detected, the gap width data is obtained. Based on the data, the robot welding condition is controlled respectively and in real time for the change in welding gap G. Further, the control of welding condition preferably include one at least among welding current, welding voltage, welding speed and offset quantity of welding torch tip 4.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for controlling a welding robot, and more specifically, a welding torch, a laser sensor of a type for deflecting a laser beam to detect a work line position, and the like are supported by the robot. The present invention relates to a control method for a welding robot, which controls a welding condition according to a gap width between welding objects (hereinafter, referred to as "welding gap width") using a sensor output.

[0002]

2. Description of the Related Art It is known to perform a welding operation by supporting a welding torch on a tip end of a robot arm and moving the tip of the welding torch along a welding line between members to be welded. The robot used is called a welding robot. Further, a laser sensor that senses a region on the robot traveling direction side with respect to the welding torch tip position is attached to the robot to obtain welding line position data, and the robot is moved while correcting the robot position based on the data. Therefore, a control method for moving a robot along a welding line more accurately is known as a real-time tracking method.

It is certain that the adoption of this real-time tracking system can considerably improve the accuracy and efficiency of welding work. However, the accuracy and efficiency of welding work depend on the suitability of various quantities that determine welding conditions such as welding current, welding voltage, welding speed (robot movement speed), and welding torch tip position offset amount, in addition to the welding line tracking accuracy. It has the property of being greatly influenced. Therefore, in order to efficiently obtain a weld bead with more stable quality, it is necessary to set the above-mentioned various amounts appropriately.

However, since there is no method for accurately dealing with the indefiniteness of the welding gap as described below,
In actual welding work, it was extremely difficult to always set the welding current, welding voltage, welding speed, welding torch tip position offset amount, etc. to appropriate levels.

That is, the welded parts of the object to be welded are not always in close contact with each other before welding, and generally there is a slight gap (hereinafter referred to as "welding gap") between them. Moreover, the gap width is not always the same for each combination of individual objects even if the combination of the types of objects to be welded is the same, and is executed in one robot movement cycle. It has the property that it is not always constant throughout the welding operation performed.

[0006] A method for individually dealing with such indefiniteness of the welding gap and accurately determining the processing conditions has not been known yet. For example, the welding gap of a set of appropriately defined standard objects has A method has been used in which the processing conditions set as being suitable for the width are applied as they are to each of the welding operations for a large number of other similar welding object sets.

[0007]

As described above, in the prior art, it is extremely difficult to stabilize the quality of the weld bead because the working conditions are not controlled in response to changes in the welding gap. For example, if welding of a welding part having a large welding gap is performed with the welding current, welding voltage and welding speed set in accordance with the narrow welding gap, the width and thickness of the weld bead may be insufficient. Is high. Further, it is considered preferable to perform offset correction for shifting the welding torch tip position to the center side of the welding gap for a welding object or a welding site having a large welding gap width. Accordingly, since the offset amount could not be controlled in real time, there was a risk that the position of the welding bead would be biased toward one of the welding object sides.

The object of the present invention is good by controlling the processing conditions such as welding current, welding voltage, welding speed, and offset amount in the welding robot individually and in real time in response to changes in the welding gap. In order to obtain a weld bead of stable quality.

[0009]

According to the present invention, sensing is performed by sensing a region on the welding robot advancing direction side of a portion being welded by using a laser sensor supported by the welding robot together with a welding torch. The above technical problem is solved by a method of controlling a welding robot including a step of acquiring gap width data of a welding gap existing in a predetermined region and a step of controlling welding conditions of the welding robot based on the welding gap width data. (Claim 1).

Then, the welding current, the welding voltage, the welding speed, and the offset amount of the tip of the welding torch are specified as various quantities for determining the above-mentioned welding conditions, and one or more of these various quantities are sensed by a laser sensor. It is proposed to control based on the welding gap width data acquired by (Claim 2).

[0011]

The control method of the welding robot of the present invention controls the welding conditions by utilizing the welding gap width data acquired by the laser sensor. Therefore, first, the principle of measuring the welding gap width by the laser sensor supported by the welding robot will be described with reference to two typical cases.

FIG. 1 is a schematic diagram for explaining the principle of measuring the gap at the butt welded portion using a beam scanning type laser sensor. In the figure, A and B are objects to be welded, and have a Y-direction gap G extending in the X-axis direction of the workpiece coordinate system that has been set for the robot. The gap width at a position where the X-axis coordinate value is x (hereinafter, simply referred to as “position x”) is represented by g (x). A welding torch 2 and a laser sensor 3 are attached to an end portion of a robot arm indicated by reference numeral 1 by omitting most of the robot body through an appropriate attachment mechanism.

Reference numeral 4 denotes a welding torch tip position (hereinafter referred to as "welding point") set as a tool point of the robot. The laser sensor 3 is arranged so that the scanning beam 5 scans a region preceding the welding point 4 in the robot traveling direction.

The laser sensor 3 has a built-in visual sensor device equipped with a CCD camera (or a linear sensor which replaces the CCD camera), and the loci 6A, 6B of the bright spots formed on the objects A, B by the scanning beam 5. 6B is detected. Figure 2
Bright spot loci 6A 'and 6B' seen on the pixel surface of the CCD camera
(1 cycle) is schematically shown. In each figure, the gap width g (x) at the position x is captured as a pixel proportional amount (for example, the number of pixels) of p (x) on the pixel surface. The relationship between the actual gap width g (x) and the gap width p (x) measured by the pixel proportional amount on the pixel surface can be obtained by appropriate calibration.

For example, when the posture of the welding robot is kept constant during welding, g (x) can be expressed by the following equation (1). g (x) = f1 (z) p (x) (1) Here, z is the Z-axis coordinate of the robot, and f1 (z) is a substantially linear function with respect to z. Let the robot take the same posture as when welding, and for a welding gap having a known gap width, an appropriate number of robot positions (z value = z1, z2, z).
The value of f (z) in 3 ...) f1 (z1), f1
If the calibration for actually measuring (z2), f1 (z3) ... Is executed, the numerical calculation of the above formula (1) becomes possible (the polygonal line approximation method can be used).

Next, with respect to the principle of measuring the gap width in the step-bonding welded portion using a beam scanning type laser sensor, FIG. This will be described with reference to FIG. Note that FIG.
3 and FIG. 3 are designated by the same reference numerals.

In FIG. 3, C and D are plate-shaped objects to be welded, and have a Z-direction gap H extending in the X-axis direction of the workpiece coordinate system set in the robot. The gap width at the position x will be represented by h (x). A welding torch 2 and a laser sensor 3 are attached to the tip end portion of the robot arm indicated by reference numeral 1 through an appropriate attachment mechanism. Reference numeral 4 represents a welding torch tip position (welding point) set as a tool point of the robot. The laser sensor 3 is arranged so that the scanning beam 5 scans a region preceding the welding point 4 in the robot traveling direction. Further, in this example, a state in which welding is performed with the robot arm tip portion 1, the welding torch 2, and the laser sensor 3 tilted in a direction intersecting the XYZ axes at a non-perpendicular constant angle is illustrated. There is.

The laser sensor 3 has a built-in visual sensor device equipped with a CCD camera (or a linear sensor which replaces the CCD camera), and the locus 6C, of the bright spots formed on the objects C, D by the scanning beam 5. 6H, 6D ', 6D are detected. Here, 6D 'represents a locus extending in the Z direction along the edge of the object D, and 6H represents a locus formed in the gap portion of the width h (x) at the position x. 6H does not always have a clear trajectory because it partially becomes the shadow of the object D.

FIG. 4 schematically shows a locus of bright spots (one cycle) seen on the pixel surface of the CCD camera.
In this example, assuming that the CCD camera images the scanning beam trajectory from the torch side of the beam scanning surface, the trajectory image is as schematically illustrated. That is, the trajectory images 7C and 7D correspond to the trajectories 6C and 6D, respectively, and 7H corresponds to 6H. Then, with respect to the trajectory 6H of the gap portion, a partially unclear image 7H is formed in some cases.

In this case, the gap width h at the position x
Since (x) is considered to be expressed as a length L (x) of 7H on the pixel surface, the relationship between this L (x) and the actual gap h (x) is expressed by the following equation (2). Assuming that, there is a method of obtaining it by the same calibration as the case of FIGS. 1 and 2 described above. However, since f2 is not only a function of z but also a function of y, the calibration is performed with the value of f (y, z) for an appropriate number of (y, z) pairs under a known gap width. Need to plot. h (x) = f2 (y, z) L (x) ... (2) The value of L (x) is the locus images 7C and 7 on the pixel plane.
It is captured as a distance (pixel proportional amount) between the D ′ end points 8C and 8D.

When the assumption that the thickness of the object D is constant is established, the intersection 9C between the extension line of the trajectory image 7D 'and the trajectory image 7C is considered, and the distance q (x) between the points 8D and 9C is set as a gap. It can be used as an index to measure the width. If the thickness of the object D is known d, the length of the trajectory image 7D ′ is set to L ′.
As (x), the gap width h (x) is expressed by the following equation (3). h (x) = d · q (x) / L ′ (x) ··· (3) The present invention does not particularly limit the method for obtaining the gap width, and the gap width may be determined by any method other than the above. May be set. As for the laser sensor used, a laser displacement sensor using a position detection photodetector (PSD) can be used other than the beam scanning type used in the above case. For example, in the case of FIG. 3 described above, the gap width h (x) is determined by detecting the step between the objects C and D using two laser displacement sensors and subtracting the thickness of the object D from this. It is possible to ask. In this case, by utilizing the fact that the robot posture during welding is constant, the laser beams of the two laser displacement sensors are always incident on the objects C and D from directly above,
If the laser displacement sensor is installed so as to be tilted with respect to the welding torch, the step can be calculated without being affected by the change in the robot position.

In the present invention, the welding conditions are controlled according to the gap width obtained by the method described above. Typical welding conditions include welding voltage V, welding current I, welding speed v, and offset amount δ. The offset amount δ is the traveling direction of the robot (the x-axis direction).
A robot position deviation amount δy and a z-direction deviation amount δz from a teaching path in a direction orthogonal to.

A function, V (Δ), I (Δ), which is an optimum value for these welding conditions in accordance with a change in the value of gap width g (x) or h (x) (denoted by Δ). It is conceivable to define each variable separately and control each variable based on this, but in practice, the gap value level Γ, which is commonly applied to each variable, can be set in several steps as follows. It is convenient to set and specify one variable value for each level in the table data format.

Gap width level Γ V I v δy δz Γ1 = 0.0 to 1.0 V1 I1 v1 δy1 δz1 Γ2 = 1.1 to 2.0 V2 I2 v2 δy2 δz2 Γ3 = 2.1 to 3.0 V3 I3 v3 δy3 δz3 Such table data is stored in the memory of the robot controller controlling the welding robot, and it is checked whether the gap width level Γ is changed every time the gap width measurement data is obtained by sensing. As a result, it is possible to control the various amounts V, I, v, δ, etc. during the welding operation to desired levels (see Examples below).

[0025]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 5 is a block diagram of the essential parts showing an example of a system configuration that can be used when implementing the welding robot control method of the present invention. To explain this, 10 is a robot controller having both a robot control function and an image processing device function, and has a central processing unit (hereinafter referred to as CPU) 11, and the CPU 11 has a memory 12 composed of a ROM, Teaching operation including a memory 13 including a RAM, a non-volatile memory 14, a laser laser sensor 3, and a general-purpose interface 15 connected to the welding torch controller 40, a frame memory 16, an image processor 17, and a liquid crystal display (LCD) 18. A robot axis control unit 20 connected to a robot body 30 via a board 19 and a servo circuit 21 is connected to each other via a bus 23.

In the ROM 12, the CPU 11 includes the robot body 30, the laser laser sensor 3, the welding torch controller 40.
Also, various programs for controlling the robot controller 10 itself are stored. The RAM 13 is a memory that can be used for temporary storage of data and calculation. Into the non-volatile memory 14, various parameter setting values and a teaching program created offline are input / stored from the teaching operation panel 19 or an offline program creating device (not shown). Here, it is assumed that the teaching program designating the path along the welding line corresponding to the case shown in FIG. 3 is already stored. As the teaching points, as shown in FIG. 3, P1 (air cut point immediately before the welding line start point), P1
Two points (air cut point immediately after the end of the welding line) and P3 (air cut point separated from the welding line) shall be set.

Laser The laser sensor 3 is supported by the robot hand 1 as described in the section of description of the operation, and the laser light source having the deflection scanning irradiation area in the area closer to the robot moving direction than the tool tip point 2 is the laser light source and the deflection. It is of a known type with a CCD camera with a scanning irradiation area in the field of view.
An image signal captured by a CCD camera that captures an image in synchronization with the deflection scanning of the laser is converted into a grayscale grayscale signal via a general-purpose interface 15 and stored in a frame memory 16. The image information read from the frame memory 16 is processed by the image processor 17 to detect the presence or absence (start point and end point) of a welding line.

Although the configuration and function described above are basically the same as those of the system including the robot controller for the conventional welding robot, the system of the present embodiment is used to implement the method of the present invention. In particular, the following preparations have been made.

(1) The non-volatile memory 14 stores a calculation program for obtaining the gap width and related required data by the method described in the section of description of the operation.
Here, the above equation (3) h (x) = d · q (x) / L ′
A program and data for controlling the image processor 16 to obtain q (x) and L '(x) as distance data on the sensor coordinate system in order to calculate (x) are stored. And

(2) In the ROM 12, the relationship between the gap width levels Γ1 to Γ3 and the quantities V, I, v, δy, δz is set in the table data format, and the thickness data d of the object D is input. Display the welding condition setting screen for
The operator sets the levels of Γ1 to Γ3, V1, I
1, v1, δy1, δz1; V2, I2, v2, δy2, δz
2; Store a program that enables numerical value setting of V3, I3, v3, δy3, δz3 and screen input of the thickness value d.

(3) The setting and input of each data by the operator is completed using the above welding condition setting screen.

(4) The welding torch controller 40 receives a control signal from the CPU 11 via the general-purpose interface 15 to turn on / off the welding torch, and at the same time, set the welding voltage V and the welding current as specified in the above table. Must have the function of supplying I to the welding torch.

(5) For the teaching points P1 and P2, the offset amounts ε1y, ε1z and ε2y, ε2z are input in advance from the teaching operation panel 19 independently of δy and δz which are controlled according to the gap width Δ. That

(6) In accordance with the teaching program, the offset amount ε designated for each teaching point, the welding gap sensing result and the setting table data, the program for executing the processing shown in the flowchart shown in FIG. Various setting values are stored in the ROM 12 or the non-volatile memory 14
Be stored in.

Below, assuming these preparations, FIG.
In the example, P0 (home position) → P2 → P
The processing of the CPU 11 when executing the control of the welding robot in which 3 is the teaching path will be described with reference to the flowchart of FIG.

The processing of the CPU 1 senses the welding gap and the main processing for controlling the robot movement and welding conditions.
A sub-process (task process) for creating data representing the gap width is included, and a flag F for controlling the sub-process and a gap width level index β having a value corresponding to the gap width level number are set. And F is 0
It also takes a value of 1 (the initial value is F = 0), and β is Γ 1 to Γ
It takes a value of 1 to 3 corresponding to 3 (the initial value is β
= 1).

The main process and the sub process are started at the same time. In the main processing, the CPU 11 first sets the first 1 of the teaching program.
The block is read (step M1), and the robot is placed on a trajectory toward a position (to be represented by P + ε1) where the teaching point P1 is shifted by ε1y in the Y-axis direction and ε1z in the Z-axis direction (step M2). On the other hand, as soon as the sub-process is started, the system waits for the flag F to be reversed (step S
1). Immediately before the robot reaches the position P1 + ε1, the next block of the teaching program is read (step M
3) The flag F is inverted to 1 (step M4).

The CPU 11 detects this in the sub-process, starts the sensing by the laser sensor 3, and measures the welding gap width (step S2). The method for obtaining the welding gap width is as described in the section for explaining the action. When the gap width is obtained, it is compared with the table data to determine the corresponding gap width level Γ (step S3), and it is determined whether or not the value of the gap width level index β needs to be updated (step S4). ). If it needs to be updated, it is immediately rewritten (step S5).

For example, when the gap width is 1.4 mm, which corresponds to the level Γ 2 in view of the table data, the value of the gap width level index β is set to the initial value 1 to 2.
To update. After confirming that the flag F is not inverted to 0 (step S6), the process returns to step S2 again,
The gap width is measured, and thereafter, steps S2 to S6 are repeated until the flag F is inverted to 0.

On the other hand, the robot moves the position P2 in the Y-axis direction by ε according to the contents read in step M3 of the main processing.
It is put on a trajectory toward a position (to be represented by P + ε2) shifted by 2ε and 2ε in the Z-axis direction (step M5). Then, the first reading of the gap width level index β is executed (step M6), the table data corresponding to the read value of β is accessed, and the welding torch controller 40 is instructed via the general-purpose interface 15. And the welding torch is energized under the conditions of the corresponding welding voltage V and welding current I (step M7). Further, the robot position command value is corrected by the offset amount δ designated by the table data (step M8), and the welding speed (robot moving speed) v needs to be changed (step M9). The value is corrected to the correct value (step M10).

When the processing corresponding to the first reading of the gap width level index is completed, P2 corresponding to the end point of the welding line is set.
After confirming that the robot has not reached the previous position (step M11), the second gap width level index β is read (step M12). It is determined whether or not the gap width level index β matches the previous read value (step M13), and if they match, the process returns to step M11. If the gap width level index β has changed from the previous read value, the quantities V, I, v, and δ that determine the welding conditions are changed accordingly to the values according to the table data corresponding to the corresponding gap level ( Step M14). Each change process includes steps M7 to M9.
It is performed according to the processing of.

That is, the welding voltage V and the welding current I are changed by giving a command to the welding torch controller 40 via the general-purpose interface 15, and the offset amount δ is changed in consideration of the offset amount ε. This is performed by changing the shift correction amount for the robot trajectory that is inserted and interpolated to a newly designated δ. Also,
Changing the welding speed v is achieved by changing the distribution period of the position pulse for each axis.

Eventually, the gap width measurement by the laser sensor 3 is periodically repeated throughout the welding execution section (flag F = 1 section), and the value of the gap width level index β is changed every time the gap width level Γ is changed. Will be updated, and the welding conditions will be changed / controlled accordingly. When the robot reaches the position immediately before the teaching point P3 defined corresponding to the end point of the welding line, a yes determination is made in step M11, the next one block of the teaching program is read (step M15), and the flag is set. F is inverted to 0 (step M16), and the welding torch is deactivated (step M1).
7). The order of the processes of steps M15 to M17 may be changed as appropriate.

Finally, the robot moves to the teaching point P3 and completes one work cycle. Even in the sub-process, F = 0
If it is detected in step S6, the gap width index β is reset to 1 (initial value) (step S6).
7), the process ends.

In the above embodiment, the welding torch is turned on.
The turning-off is performed at the timing when the robot reaches the position corresponding to the teaching points P2 and P3. However, the processing for causing the laser sensor 3 to detect the starting point and the ending point of the welding line is executed in the sub-processing, and when the starting point is detected. It is also possible to adopt a method of energizing the welding torch and deactivating it when the end point is detected.

It is also possible to combine a control system in which the laser sensor 3 senses the welding line position at the same time as measuring the welding gap width level and performs real-time tracking correction. For example, a trajectory in which a position of 8D is captured on the trajectory image in FIG. 4 and position correction is performed so that the robot has a specific positional relationship with the lower edge position of the step portion of the object D in the case of FIG. It is conceivable to adopt a control method in which the offset position for correcting the robot position by δy in the Y-axis direction and δz in the Z-axis direction is further corrected according to the gap level.

[0047]

According to the present invention, various amounts that determine welding conditions can be controlled in a form that immediately responds to the indefiniteness of the size of the welding gap existing between the objects to be welded. As a result, the quality of the weld bead is prevented from becoming unstable, and the quality of the weld bead is constantly improved.

[Brief description of drawings]

FIG. 1 is a schematic diagram for explaining the principle of measuring a gap at a butt weld portion using a beam scanning laser sensor.

2 schematically shows a locus of bright spots formed on the pixel surface of the CCD camera in the arrangement of FIG.

[Fig. 3] Fig. 3 shows the gap width in the step-bonded welded portion.
It is a schematic diagram for explaining the principle of measurement using a beam scanning type laser sensor.

4 schematically shows a locus of bright spots formed on the pixel surface of the CCD camera in the arrangement of FIG.

FIG. 5 is a principal block diagram showing an example of a system configuration that can be used when implementing the welding robot control method of the present invention.

FIG. 6 is a flowchart outlining the processing of the CPU when carrying out a welding operation in the case depicted in FIGS. 3 and 4 using the system shown in FIG.

[Explanation of symbols]

1 Robot Arm Tip 2 Welding Torch 3 Laser Sensor 4 Welding Torch Tip 5 Scanning Beam 6A, 6B, 6C, 6D, 6D ', 6H Scanning Beam Trajectory 7A, 7B, 7C, 7D, 7D', 7H Scanning Beam Trajectory Image 8C End point of locus image 7C 8D End point of locus image 7D '9C Intersection of extension line of locus image 7D' and locus image 7C 10 Robot controller 11 Central processing unit (CPU) 12 Memory (ROM) 13 Memory (RAM) 14 Nonvolatile Memory 15 General-purpose interface 16 Frame memory 17 Image processor 18 LCD 19 Teaching operation panel 21 Robot axis control unit 22 Servo circuit 23 Bus 30 Robot body 40 Welding torch controller A, B, C, D Welding object G, H Gap L Distance between end points 8C and 8D L '(x) at position x The length of the mark image 7D 'P1, P2, P3 teaching point g (x), h (x) gap width p (x) video path 7A, between the inner end points of 7B distance q (x) 8D, 9C distance

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Yuuki Makibata Yuki Makihata 3580 Kobaba, Oshino-mura, Minamitsuru-gun, Yamanashi Prefecture FANUC Co., Ltd. 3580 FANUC CORPORATION (72) Inventor Yoshitaka Ikeda Oshinomura, Minamitsuru-gun, Yamanashi Prefecture Kobaba, 3580 Address FANUC CORPORATION

Claims (2)

[Claims] 1. A welding existing in the sensed region by sensing a region on the welding robot advancing direction side of a region being welded using a laser sensor supported by the welding robot together with a welding torch. A method of controlling a welding robot, comprising: obtaining gap width data of a gap; and controlling welding conditions of the welding robot based on the welding gap width data. 2. The control of the welding condition includes control of at least one of a welding current, a welding voltage, a welding speed and an offset amount of the welding torch tip. Control method for welding robot.