JPH0816227A - Method for teaching moving route of automatic machine using position sensor - Google Patents

Abstract

PURPOSE:To control the route movement of an automatic machine capable of instantaneously responding to a change in the positioning, size, shape or the like of an individual work. CONSTITUTION:A work W having fixed cylindrical outside size on design has a stright part 6 to be a welding center line on which a temporarily welded part 8 is formed and an elliptic or circular arc part 7. A welding torch 2 and a laser sensor 3 are fixed to the tip part 1 of a robot arm. A mark 4 is the tip point of a tool for a welding robot. The tip part 1 of the robot arm is successively moved to welding line detecting positions C1, C2, D1 to D4, the positions of an individual work W on positions A1, A2, B1 to B4 are found out and the expressions of the straight part 6 and the straight line and an ellipse (or a circle) of the arc part 7 are determined by using the found positions. Then the center P0 of the arc part 7 and an intersection P1 between the arc part 7 and the straight part 6 and a position P2 separated from the intersection P1 by a set angle phi are calculated and teached to the robot. When regenerative operation is executed immediately after teaching, route movement of P0 (straight part 6) P1 (circular arc part 7) P2 is executed.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for teaching a moving path of an industrial robot used for welding, deburring, bolting, sealing, painting, cutting, boring, etc., and an automatic machine such as a CNC machine tool. More particularly, it relates to a method for determining and teaching path movement of an automatic machine using a position sensor.

[0002]

2. Description of the Related Art In an automatic machine such as an industrial robot (hereinafter, simply referred to as a "robot") or a CNC machine tool whose path operation is controlled by a numerical controller, a taught point or locus is set. The teaching reproduction method (teaching / playback method) to be followed has been widely used. A major problem in using this method is the position and posture of a work object such as a work (hereinafter referred to as “work”) (hereinafter, simply referred to as “position” in this sense unless otherwise specified). It is not possible to respond to changes in.

That is, even if teaching is performed in a state in which workpieces that represent a large number of workpieces of the same type are accurately positioned, the positioning state that exactly coincides with the time of teaching is reproduced for a large number of workpieces to be actually worked. It is difficult to do so, and when there is a non-negligible dimensional variation among the individual workpieces, a decrease in work accuracy cannot be avoided.
Further, when performing the same type of work on the types of works having the same shape but different sizes, it is necessary to redo the teaching.

In order to overcome such a problem, a sensor having a position detecting function such as a laser sensor, a visual sensor, a laser displacement sensor (hereinafter, simply referred to as "position sensor").
The following method using is adopted. [1] Teaching is performed with a work representative of the same kind of work positioned at a reference position, and the position deviation of each work from the reference position is measured by a visual sensor or the like, and the teaching trajectory is corrected based on the result. Method.

[2] Teaching is performed in a state where a work representative of the same kind of work is positioned at a reference position, and a working line or a characteristic line (for example, a welding line, a ridge line, etc.) of each work is used by using a laser sensor or the like during a regenerating operation. ) Is continuously monitored, and the route is moved while correcting the position and orientation of the automatic machine.

[0006]

However, the above-mentioned method [1] deals with the positional deviation corresponding to the parallel movement and rotation of the work, and the entire work line such as the welding line is changed by changing the size of the work. It was difficult to apply in the case where it has to be corrected in a form other than translation or rotation.

In the method [2], it is necessary for the position sensor to continuously detect the work line or the characteristic line instead of the work line. For example, in the case of welding work on a work in which many tack welds are formed. As described above, if the characteristic part suitable for detection by the position sensor does not continuously extend along the movement path of the automatic machine such as the robot, the appropriateness becomes difficult. Further, in order to perform real-time position correction using a laser sensor or the like, a large amount of data processing capability at high speed is required, which is disadvantageous in reducing the cost of the entire system including the CPU.

Therefore, the object of the present invention is to change the position (parallel movement and rotation) and size of each work, and to change the extending direction of a straight ridge between a circle (arc) and an ellipse (arc). It is another object of the present invention to provide a method for teaching a moving route of an automatic machine that can respond to any of simple modifications.

[0009]

The present invention provides, as a basic technical means for solving the above-mentioned problems, "a position sensor means is used to acquire a predetermined number of position data for a predetermined shape portion of a work object. Of the automatic machine using the position sensor, the step of specifying the equation representing the predetermined shape portion based on the position data, and the step of teaching the movement path of the automatic machine based on the specified equation. The method of teaching a moving route "(Claim 1).

Further, the invention of the present application corresponds to a case where the above-mentioned automatic machine is a robot equipped with a position sensor, and said, "The automatic machine is a robot and the position sensor is installed in the robot, The method further includes a step of teaching in advance a robot position for acquiring a predetermined number of position data regarding a predetermined shape portion of the object "(Claim 2).

[0011]

As described above, when the same kind of work is repeated on a large number of works using an automatic machine such as a robot, it is necessary to assume a considerable variation in the positioning, size, shape, etc. of each work. There are many. Even in such a case, the present inventor suffices to consider the change to a very similar shape even if the shape itself of the same type work is kept substantially unchanged, or if there is a change. , We paid attention to the fact that a specific part of each work can be expressed by a common equation, at least approximately.

For example, the linear ridgeline of each work is
Even if there is a slight difference in direction and length for each work, as long as the condition that it is a straight line is not broken, the coefficient parameter value of the equation expressing the straight line in the work space and the length and angle that specify the predetermined range It is possible to describe uniformly by leaving the degree of freedom in parameters such as. In a case where flattening due to a small distortion is assumed in the arc-shaped ridge line (see the example described later), an equation expressing a general inertia circle (a circle is one of them) may be prepared with the coefficient parameter undecided.

The parameter values of the equations prepared with these coefficient parameters undecided can be determined by detecting the position of each work using a position sensor. For example, the equation of a straight line can be determined by knowing the positions of at least two points, and can be determined by knowing the positions of four points (three points in the case of holding an arc state) when assuming the transformation into an ellipse. .

Therefore, before the work originally intended by an automatic machine such as a robot (for example, welding along the ridge line of a work) is performed, the positioning and the size of each work (however, depending on the case) are performed. , It may be set for each group of workpieces whose shape is assumed to be accurate.), The position data of the required point is acquired by using the position sensor, and the equation expressing the predetermined part is determined. Based on this, it becomes possible to determine the moving path of the automatic machine during the desired work.

The type and position of the position sensor to be used may be selected according to the nature of work and environment. For example, in arc welding work using a robot, a laser sensor may be mounted on the robot and the position may be detected by sequentially moving the robot to a required position along the welding line.
Alternatively, a three-dimensional visual sensor, which is a combination of a slit light projector and a CCD camera, may be installed at a fixed position to acquire position data, and the position data may be transmitted to an automatic machine.

As described above, even if a considerable variation in positioning, size, shape, etc. of each work is to be assumed, the path of the automatic machine can be detected by performing position detection with a relatively small number of points. Is a basic feature of the present invention that can be determined in a form that immediately responds to each work. Further, according to the present invention, since it is not necessary to perform a lot of data processing at high speed unlike in the case of welding line real-time tracking, the processing load on the control unit of the automatic machine is reduced, which contributes to the cost reduction of the entire system. obtain.

[0017]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 shows a welding robot as an automatic machine to which the invention of the present application is applied, when performing arc welding work on a required welding portion along a radial direction and a circumferential direction of a work having a substantially cylindrical shape. It is a schematic diagram explaining arrangement | positioning. In the figure, W is a work to be welded, which has a cylindrical outer shape of a fixed size (radius r, height h) by design, but may have an elliptic cylindrical shape due to a minute processing error. , There may be some size error. The positioning accuracy is not strict.

Most of the robot body is omitted and reference numeral 1 is used.
The welding torch 2 and the laser sensor 3 are attached to the tip end portion of the robot arm instructed in step 1 through an appropriate attachment mechanism. Reference numeral 4 represents the welding torch tip position set as the tool tip point of the welding robot.

From the laser sensor 3, the deflected scanning beam LB
Is projected toward a region slightly ahead of the tool tip point 4 (a portion to be welded immediately after), and a light spot locus 5 is formed along a direction orthogonal to the welding center line. The laser sensor 3 has a CCD camera V for photographing the light spot locus 5 from an oblique direction with respect to the deflection scanning plane (fan-shaped plane) of the laser beam LB.
Equipped with C.

The welding center line is composed of a straight line portion 6 extending in the radial direction on the top surface of the work W and an arc portion (or an elliptical arc portion, hereinafter the same) 7 along the peripheral portion of the top surface. , The starting point P0 of the straight line portion 6 is at the center of curvature of the circular arc portion 7 (in the case of an elliptical arc, it is considered to be the midpoint of two focal points, as will be described later), and the end point is the starting point P1 of the circular arc portion 7. Match.

The arcuate portion 7 is assumed to extend from the starting point P1 to the ending point position P2 which is separated by an angle φ in the direction shown. It should be noted that the straight line portion 6 of the welding center line has a V-shaped valley shape, and temporary welding parts 8 are formed in advance at some points along the straight line portion 6 and the arc portion 7 of the welding center line. There is. After all, the total length of the welding center line is P0 → (straight line portion 6) → P1 →
(Arc part 7) → P2.

As described above, under the condition that the shape error, size error, positioning error and the like of the work W are assumed, the positions of the points P0, P1 and P2 on the welding center lines 6 and 7 are the work W. It is not always constant every time. Therefore,
Generally, P0, P1 and P2 are not fixed points on the world coordinate system Σw set for the robot.

FIG. 2 is a block diagram showing the outline of the entire system including the robot 1 and the laser sensor 3 used in the above arrangement. Explaining this, 10 is a robot control device having both a robot control function and an image processing device function.
CPU. ) 11 and the CPU 11 has an RO
A memory 12 including M, a memory 13 including RAM, a non-volatile memory 14, a general-purpose interface 15 connected to the laser sensor 3 and a power supply unit 30 for the welding torch, a frame memory 16, an image processor 17, a liquid crystal display device (LCD). ) 18 and a robot axis controller 20 connected to the robot body 1 via a servo circuit 21 are connected via a bus 23.

The laser sensor 3 is provided with a laser beam projection unit and a CCD camera VC, and performs deflection scanning of the laser beam LB and photographing of a light spot locus according to a command from the CPU 11. The image signal captured by the CCD camera is converted into a grayscale grayscale signal via the general-purpose interface 15 and stored in the frame memory 16. The image information read from the frame memory 16 is processed by the image processor 17. The power supply unit 30 also has a function of controlling the welding voltage and the welding current supplied to the welding torch in accordance with a command from the CPU 11.

The ROM 12 stores various programs for the CPU 11 to control the robot body 1, the laser sensor 3, the power source unit 30 and the robot controller 10 itself. To this end, a program for setting various conditions related to execution of multi-layer welding in a screen input format in the manner described below and a light spot locus image are analyzed to obtain a rise h of the light spot locus image, A program for converting this to the thickness of the weld layer (denoted by γ) is included. RAM1
Reference numeral 3 is a memory that can be used for temporary storage of data and calculation.

The non-volatile memory 14 stores various programs and set values that define the operation contents of the robot system. Here, in particular, a program for executing a series of processes (see FIG. 3) described later, related set values, equation expression data, and the like are included. Given the system configuration and functions described above, the following preparations are made to carry out the method of the present invention.

[1] Teaching of welding center line position detection position (see FIG. 1) Assuming a position where the work W is most likely to be arranged, at least two points C1 around the straight line portion 6,
C2 (generally, an appropriate number of two or more, here two), and at least four points D1, D2, around the arc portion 7.
D3, D4 (Generally, an appropriate number of 4 points or more, here 4
It is a point. However, if the deformation into an ellipse is not considered, it is possible to set at least 3 points. ) To teach the robot position.

The position teaching is performed by operating the teaching operation panel 19.
The robot 1 is jogged, the laser sensor 3 is activated, and the robot is stopped at a position where the laser beam LB is deflected and scanned so as to cross the welding center lines 6 and 7 at a position sufficiently distant from the tack welding portion 8, Teaching / playback method is used. Thereafter, the robot positions C1, C2, D1, D2, D3, D for the arbitrary work W
The detection points when the position detection by the laser sensor 3 is executed at 4 are represented by A1, A2, B1, B2, B3, and B4, respectively. Note that FIG. 1 illustrates that the position B4 is detected at the robot position D4, and B3 and D3 are not shown (hidden behind the robot 1).

[2] Setting of parameters for determining welding range In the present embodiment, the welding start point P0 is the center point of the circular arc portion 7 and the direction change point P1 is the intersection of the straight line portion 6 and the circular arc portion 7. To determine the range, the angle φ from the position P1 to P2 is set in the non-volatile memory 14.

After the above settings are made, the work W is sequentially supplied and positioned, and the welding work to which the control method according to the present invention is applied is started. Hereinafter, the automatic driving process executed by the CPU 11 of the robot controller 10 will be described with reference to the flowchart of FIG. In addition to FIGS. 4 and 5, the position detection method by the laser sensor 3 will be referred to.

First, the laser sensor 3 is activated (step S1), and the robot is moved to the teaching position C1 (step S2). Next, at the robot position C1, the intersection position A1 between the linear portion 6 and the laser beam LB is detected (step S3). Here, the outline of the method of detecting the intersection position A1 will be described with reference to FIG.

FIG. 4 shows that when the position of the linear portion 6 is detected, the linear portion 6
C at the time when the laser beam LB is deflectively scanned
It is an example of a video image captured by a CD camera VC. FIG.
In the above example, the image of the light spot locus to be photographed is composed of straight line image portions 5H and 5H ′ corresponding to the flat portions on both sides of the work W and 5H ″ corresponding to the V-shaped valley portion. The position corresponding to A1 is, for example, the bottom 5J of the V-shaped valley portion or the inner end point 5 of the straight line image portions 5H and 5H '.
We can consider the midpoint 5K "of K, 5K '.

By the image processing using the image processor 17, the position on the pixel surface of 5J or 5K ″ is obtained,
This is converted into data on the sensor coordinate system, and the position of the point A1 is obtained using the calibration data of the sensor coordinate system and the position data of the robot position C1 (how to use such a laser sensor is well known). .

Similarly, in steps S4 and S5,
The robot is moved to the position C2 and the position of the point A2 is obtained. Next, the robot is moved to the teaching position D1 (step S6), and the (elliptical) arc portion 7 is formed at the robot position D1.
The crossing position B1 of the laser beam LB and the laser beam LB is detected (step S7).

The outline of the method of detecting the position B1 will be described with reference to FIG. FIG. 5 shows C at the time of detecting the position of the (elliptical) arc portion 7.
It is an example of a video image captured by a CD camera VC. FIG.
In the above example, the image of the light spot locus to be photographed is composed of the linear image portion 5M corresponding to the flat portion of the top surface of the work W and 5N corresponding to the side surface portion. As the position corresponding to the detection position B1, the bending point 5T to which the straight line image portions 5M and 5N are connected may be considered.

By the image processing using the image processing processor 17, the position of the bending point 5T on the pixel surface is obtained, converted into data on the sensor coordinate system, and further calibration data of the sensor coordinate system and the robot position. The position of point B1 is obtained using the position data of D1 (how to use such a laser sensor is also well known).

Similarly, steps S8 to S13.
To move to robot positions D2 to D4 in sequence and move from point B2 to
Find the position of B4.

Next, after turning off the laser sensor 3 (step S14), the equations of the straight line portion 6 and the (elliptical) arc portion 7 are determined (steps S15 and S16).
The equation of the straight line portion 6 can be calculated from the three-dimensional position data of the two points A1 and A2. In addition, when using the position data of three or more points, for example, an approximate calculation by the least square method may be performed. Also, the equation of the (elliptical) arc part 7 is
It can be calculated from the three-dimensional position data of two points B1 and B4. When using the position data of five or more points, similar to the case of the straight line portion 6, an approximate calculation by the least square method or the like may be performed.

In the following step S17, step S16
The center position (in the case of an ellipse, the midpoint position of the two focal points) is calculated from the equation of the circle (ellipse) obtained in step 1, and the position P0 is determined. In addition, due to the measurement error of the laser sensor 3, the position P0
However, in some cases, the equation represented by the straight line portion 6 may not be strictly satisfied. In that case, a perpendicular line (which should be extremely short) is dropped from the position P0 to the straight line portion 6 as necessary, and the intersection point is again set to P0. Also good. Alternatively, the detection point A1,
It is also possible to substitute one of A2 at the center position of the circle (ellipse) (or the midpoint position of the two focal points).

The position P1 is determined by solving the equation represented by the straight line portion 6 and the equation represented by the (elliptical) arc portion 7 as simultaneous equations. Further, P2 is obtained as a position separated from the position P1 by an angle φ along the (elliptical) arc portion 7. The position data of these positions P0, P1 and P2 are
It is stored in the non-volatile memory 14.

As described above, the temporary (te
mporary) teaching work is completed. Therefore, in the following step S18, the robot is moved to the position P0 and the welding torch 2 is energized at an appropriate timing. It should be noted that the robot path P0 to P2 itself is set by the set amounts Δ1 to Δ3 from the path.
If it is desired to shift the position, the position shift amount of Δ1 to Δ3 may be added to determine the robot movement position.

Then, the robot is moved between P0 and P1 (or between the shift positions thereof) in a linear motion (step S19), and when the position P1 (or its shift position) is reached, the robot is moved. Position P2 (or its shift position)
Up to the step S20, the robot is moved while sequentially calculating the route along the arc or elliptic arc using the equation calculated in the step S16. Thus, one work cycle for one work W is completed. If multi-layer welding is necessary, it is sufficient to shift the path by an appropriate amount and repeat the path movement.

When the work for one work W is completed,
When the next work W is supplied, steps S1 to S20 are executed again. In principle, the movement path of the robot is determined based on the detection results of the positions A1 to B4 executed for each work. Therefore, when there is a considerable variation in the positioning, size, shape, etc. of each work. Even if there is, a robot path that corresponds to each work is realized.

As described above, the case of performing the arc welding work of the required welding portion along the radial direction and the circumferential direction of the work having a substantially cylindrical shape has been described on the assumption that the welding robot equipped with the laser sensor is applied. As an automatic machine, for example, a machine tool having an XYZ table whose position is controlled by CNC is assumed, and the same idea is applied even when a slit light projection type three-dimensional visual sensor is used as a position sensor in a fixed arrangement, and an individual work piece is applied. It will be clear that it is possible to carry out route teaching that is responsive to.

[0045]

According to the present invention, even when a considerable degree of variation in positioning, size, shape, etc. of individual works is expected, the automatic machine can be operated by detecting the position with a relatively small number of points. It is possible to set the route in a form that responds to each work.

Further, in the present invention, since it is not necessary to perform a large amount of data processing at high speed unlike in the case of welding line real-time tracking, the processing load on the control unit of the automatic machine is reduced and the cost of the entire system is reduced. It is possible to plan.

Further, even in a work including a portion where it is difficult to apply the real-time tracking method, such as a welding work for a work having a large number of tack welds, the movement path of the automatic machine can be rationally and easily. There is also an advantage that it can be set.

[Brief description of drawings]

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 illustrates a welding robot as an automatic machine to which the invention of the present application is applied, and an arrangement for executing arc welding work of a required welding portion along a radial direction and a circumferential direction of a work having a substantially cylindrical shape. FIG.

FIG. 2 is a block diagram showing an outline of an entire system including a robot and a laser sensor used in an embodiment.

FIG. 3 is a flowchart illustrating a process executed in the embodiment.

FIG. 4 illustrates an image captured by a CCD camera VC when a laser beam LB is deflectingly scanned on a linear portion 6.

FIG. 5 illustrates an image captured by a CCD camera VC when a laser beam LB deflects and scans an arc portion 7.

[Explanation of symbols]

 1 Robot body (arm tip) 2 Welding torch 3 Laser sensor 4 Welding point (welding torch tip) 5 Light spot locus 5H, 5H ', 5H ", 5M, 5N Light spot locus image 5J Valley bottom corresponding point 5K, 5K 'End point of light spot locus image 5K "5K, 5K' midpoint 5T Bent portion of light spot locus image 6 Welding center line (straight part) 7 Welding center line (arc part or elliptical arc part) 8 Temporary welding part 10 Robot Control Device 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 Control Panel 21 Robot Axis Control Unit 22 Servo Circuit 23 bus 30 power supply unit LB deflection further scanning laser beam VC CCD camera W workpiece to be welded

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁶ Identification code Internal reference number FI technical display location G05D 3/12 L

Claims (2)

[Claims] 1. A step of obtaining a predetermined number of position data on a predetermined shape portion of a work object by using a position sensor means, and a step of specifying an equation representing the predetermined shape portion based on the position data, A method for teaching a moving path of an automatic machine using a position sensor, which comprises a step of teaching a moving path of the automatic machine based on an identified equation. 2. The automatic machine is a robot, and
The position sensor is mounted on the robot, and the method further comprises the step of previously teaching a robot position for acquiring a predetermined number of position data regarding a predetermined shape portion of the work object. The method described.