JPS60134908A - Correcting method of teaching data of robot - Google Patents

Abstract

PURPOSE:To make private measuring instruments unnecessary to reduce expenses of maintenance considerably by determining the position and the attitude of a robot and correcting teaching data on a basis of a coordinate conversion matrix which is obtained by measuring this state with the robot itself. CONSTITUTION:A reference post 80 for determining the attitude of a multiaxes robot is provided in a preliminarily determined position of a working area of the robot, and an engaging terminal 81 which is engaged with this post 80 is provided, and position and attitude data of the robot in a new orthogonal coordinate system of the multiaxes robot which is set then are obtained on a basis of the operation quantity of every axis which is obtained from every axis position detector when the terminal 81 is engaged with said post 80. Thereafter, the coordinate conversion matrix between new and old orthogonal coordinate systems is obtained on a basis of these position and attitude data and those in an old orthogonal coordinate system where teaching data of the multiaxes robot are defined. Teaching data of the multiaxes robot are corrected on a basis of this matrix.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a teaching data correction method for a robot.

In order to cope with the increasing number of vehicle models and repeated model changes due to the diversification of needs, the assembly line of the company has been introducing a large number of general-purpose and quick-response robots, especially in the welding process.

By the way, these robots placed on the assembly line undergo periodic maintenance (overhaul, etc.), but each time this maintenance is performed, the robot installation position changes, so the robot is not stored in the storage device. It is necessary to change the taught data.

In this case, it may be possible to perform new teaching and re-create the teaching data, but this would take a lot of time and reduce the operating rate of the line.
Normally, the original teaching data is corrected according to the installation deviation of the robot.

Conventionally, when performing this kind of correction, the robot's installation misalignment was actually measured using a special measuring instrument.

A coordinate transformation matrix was determined based on the measured values, and the teach data was corrected using it.

However, such conventional methods require specialized measuring instruments, which is not only costly, but also requires the preparation of multiple measuring instruments when maintaining multiple robots on an assembly line at once. Otherwise, there was a problem that work efficiency would deteriorate and the operating rate of the line would decrease.

In addition, when using a dedicated measuring device, it is necessary to secure a work area for measurement around the robot installed on the line, or if such a work area cannot be secured, it is necessary to correct the teach data. There was also a serious problem that there was no such thing.

1-■ This invention has been made in view of the above background, and provides a robot teach data correction method that does not require the use of a dedicated measuring instrument for measuring installation deviations as described above. With the goal.

Structure Therefore, the robot teach data correction method according to the present invention focuses on the fact that a multi-axis robot used in, for example, a welding robot as described above is equipped with an axis position detector called an internal field sensor on each of its axis muscles. Then, do the following.

That is, a reference post for determining the posture of the robot is provided at a predetermined position in the operating area of the multi-axis robot as described above, and the tip of the multi-axis robot is engaged with the reference post to move the robots 1 to 1. An engagement terminal is provided for determining the position and orientation of the axial muscle, and the position and orientation of the currently installed After obtaining the position and orientation data of the multi-axis robot in the new Cartesian coordinate system, calculate the position and orientation data of the multi-axis robot in the old Cartesian coordinate system in which these position and attitude data and the teach data of the multi-axis robot are defined. A coordinate transformation matrix between the new and old orthogonal coordinate transformation systems is created based on the posture data, and the teach data of the multi-axis robot is corrected based on the created coordinate transformation matrix.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings.

FIG. 1 is an external view of a multi-axis robot showing an embodiment of the present invention, and FIG. 2 is a skeleton diagram for explaining the motor functions of the multi-axis robot shown in FIG. 1, represented by symbols.

In both figures, for example, a right-handed orthogonal coordinate system (○) is defined at a predetermined position of a welding line (not shown).

A six-axis welding robot 1 in polar coordinates installed at the origin O of (j, k) has a base 2, a first rotation axis 3A rotating in the θ1 direction with respect to the base 2, and a first rotation axis 6A. A'
A telescoping axis 4 that expands and contracts in three directions, and θ with respect to this telescoping axis 4
A second rotating shaft 5 that rotates in four directions, a second rotating shaft 6 that rotates in the θ5 direction with respect to the second rotating shaft 5, and a third rotating shaft that rotates in the θ6 direction with respect to the second rotating shaft 6. an engagement terminal that engages with a reference post 80 (described later) to determine the position and orientation of the welding robot 1 instead of a welding gun (not shown) that is usually attached to the tip of the third rotating shaft 7; 81 is installed.

The six axes from the two-axis component 6 to the third rotating shaft 7 in this welding robot 1 are each driven by a drive device (not shown) for each axis, which is composed of a hydraulic actuator, an electric motor, etc., and the amount of each operation is It is detected by an axial position detector such as an absolute encoder provided on each axial muscle.

In FIG. 1, 9 is a control device for the welding robot 1, and 10 is a teaching box.

In addition, in FIG. 2, the lengths of the base 2 and the first rotating shaft 3A are c'1. c2 (constant), the combined length of the first rotation axis 3B and the telescopic part 4 is L (variable), and the bending position of the second rotation axis 5 and the second rotation axis 6 (90° bending position) is C- (constant n), and the combined length of the remaining part of the second rotating shaft 6 and the third rotating shaft 7 and the engagement terminal 8
Let the lengths of 1 be C4 and G'3 (constant), respectively.

Furthermore, the following OL+ 02+'3+04+0s
Let +&e (hereinafter abbreviated as "01 to O6") indicate the amount of operation (displacement variable) of each axis.

Next, the structure of the reference post 80 and the engagement terminal 81 will be explained with reference to FIGS. 3 to 7.

First, in FIGS. 3 to 5, the reference post 80 for determining the posture of the welding robot 1 consists of a base plate 80a, a post plate 80c held upright on the base plate 80a by a splint plate 80b, and a post plate 80c of the post plate 80c. It consists of grooves 80d and 80e drilled in a cross shape in the upper part, and if it can be permanently installed at a predetermined position in the operating area of the welding robot 1 shown in FIGS. 1 and 2, or if it can be made so that no installation error occurs, Temporarily installed during robot maintenance.

Note that this reference post 80 is not very large and can be permanently installed in the operating area of the robot without interfering with the movement of the robot, and in this embodiment, it is permanently installed. do.

Further, the direction of the surface of the post plate 80c of this reference post 80 and the cross-shaped groove 80d with respect to the longitudinal direction of the surface, for example,
The attitude of the welding robot 1 is determined by the rotation angle of the soe (in this embodiment, the angle at which the longitudinal direction of the post plate 80c and the groove 80e are parallel), and the intersection of the cross-shaped grooves 80d and 80e is located at the position when the post is installed. The position of the welding robot 1 is determined by:

Next, in FIGS. 6 and 7, the engagement terminal 81 is attached to the mounting plate 81. and the cylindrical body 81 fixed to this mounting plate 81a.
b, and the grooves 8 of the reference post 80 mentioned above are fixed around the cylindrical body 81b at every 90° along its generatrix direction.
The four engagement plates 81c to 81f that engage with the welding robot 1 and the soe are arranged in a row and are attached to the tip of the welding robot 1 (the tip of the third rotating shaft 7) by the mounting plate 81a.

Note that the engagement terminal 8 attached to the tip of the welding robot 1
Although the posture of the welding robot 1 determined when the engagement plates 810 to 81f in No. 1 are engaged with the grooves 80d and soe of the reference post 80 is arbitrary, in this embodiment, the grooves 80d and soe on the permanently installed reference post 80 are determined. The direction vector V determined by soe (see Figure 2) is set as the direction in which the welding gun grips the workpiece, and the direction vector in the direction perpendicular to the surface of the post plate 80c (see Figure 2) is set as the position P (see Figure 2) of the intersection of groove 80d and soe. This is the approach direction of the welding gun (see Figure 2).

Also, the position P of this intersection and the direction vector u, 1
0D(11,1)) can be expressed as three angles using Euler angles), welding robot 1
The position and orientation of are uniquely determined.

Furthermore, if the rotatable range of the third rotating shaft 7 exceeds 90 degrees, the engagement plates 810 to 81f of the engagement terminal 81 may be color-coded so that they can be identified.
~81f and the grooves 80d and 8os of the reference post 80 can always be engaged in the same state.

FIG. 8 is a block diagram of the control system of the welding robot 1 shown in FIGS. 1 and 2.

In the control device S in the figure, 11 is a robot control microcomputer (hereinafter referred to as the "control microcomputer"), and 12 is a teach data correction microcomputer (hereinafter referred to as the "correction microcomputer"). It is.

The control microcomputer 11 performs the following functions depending on the mote selected by, for example, a teach mote/playback mode selection switch and a correction mode changeover switch 13 provided on a robot control panel (not shown).

(a) When the teaching mode is selected and the correction mode selector switch 16 is off, when drive commands for each axis of the robot are input from the teaching box 10 via the interface circuit 14,
The control microcomputer 11 performs the first
．． The first rotation axis 3 in the welding robot 1 shown in FIG.
A drive device 1 that drives each axis from A to the third rotating shaft 7, respectively.
5A to 15F as the interface circuit 14. Each is independently controlled via D/A converters 16A to 16F and drive circuits 17A to 17F.

Therefore, each axis of the welding robot 1 moves according to the drive command, and if the welding gun is attached, the position and orientation of the welding gun and the engagement terminal 81 are changed if the engagement terminal 81 is attached as shown in the figure. be adjusted.

Therefore, if the instructor has attached the welding gun by appropriately operating the teaching box 10, he or she can position it at the welding point position on the workpiece in a predetermined posture, and also attach the engagement terminal 81 as shown in the figure. If so, move the engagement plates 81c to 81f to the reference post 8.
It can be positioned and engaged within the grooves 80d and 3[le of 0.

Then, when the memory switch of the teaching box 10 is turned on during 1st positioning, the control microcomputer 11 detects the amount of movement of the first rotation axis 3A to the third rotation axis 7 in the welding robot 1, which is detected by the axis position detectors 18A to 18F. After taking in θ1+02+'3+ 04+ 05,06 via the interface circuit 14, it is stored in the teach data memory 20 via the interface circuit for one day. The muscle movement amounts 01 to θ6 are also used for positioning control during shaft drive.

(b) When the teach mode is selected and the correction mode changeover switch 13 is on, it is almost the same as when the correction mode changeover switch 13 is off, except that the captured operation amounts 01 to θ6 are stored in the teach data memory 20. Interface circuit 1 without memorizing
4 or 19 to the correction microcomputer 12 (described later).

(c) When the playback mode is selected, the control microcomputer 11 refers to the teach data stored in the teach data memory 20 and updates the interface circuit 14. The drive devices @15A to 15F are controlled via D/A converters 16A to 16F and drive circuits 17A to 17F, thereby moving the welding gun attached to the welding robot 1 according to the teach data.

At this time as well, the motion amounts 01 to θG of each axis muscle from the axis position detectors 18A to 18F are used for positioning control during axis driving.

Next, the correction microcomputer 12 operates the first. The sampling switch 21 is operated when an installation error occurs during maintenance of the welding robot 1 shown in FIG. The correction switch 22 is a switch that is turned on when sampling the movement amount of each axis obtained when 810 to 81f are engaged, and the correction switch 22 is a switch that is turned on when starting a correction calculation of the teach data after the sampling is completed. The operation status of both switches is determined by the correction microcomputer 12 via the interface circuit 23.
checked by.

The operation of this correction microcomputer 12 will be explained in detail below, including some operations of the control microcomputer 11, but it is assumed that maintenance is being performed on the welding robot 1 and its installation position has not yet been determined. .

First, operate the teach mode/playback mode selection switch 1 on the robot control panel (not shown) to set the teaching mode, and turn on the correction mode changeover switch 1B.

In this way, the control microcomputer 11 performs the steps from 5TEPI, 2 to 5TE of the processing program flow diagram shown in FIG.
Proceed to P3 and wait for a drive command from the teaching box 10.

Next, the instructor operates the teaching box 10,
1st. When a drive command for driving each axis from the first rotation axis 6A to the third rotation axis 7 in the welding robot 1 shown in FIG. 2 is input to the controller, the control microcomputer 11 proceeds to 5TEP 4. Drive unit based on the drive command @1
Performs drive control of 5A to 15F.

When the drive control in ST'EP4 is performed, the welding port bot 1 moves and the engagement terminal 81 attached to its tip moves in a required direction.

If the memory switch and end switch in the teaching box 1o are not turned on, 5TEP3, 4,
5, 7.3 is formed, the teacher repeatedly operates the teaching box 1o to adjust the engagements 181c to 81f of the engagement terminals 81 to the reference post 8D (] 1180d.

80e, and when the memory switch is turned on when positioning, the control microcomputer 11
is 5TEP 6, and during the positioning, the axis position detector 18
Movement amount θ of each robot axis detected by A to 18F respectively
l to θ6 are taken in through the interface circuit 14, and the taken-in operation amounts oI to θ are transferred to the correction microcomputer 12 through the interface circuit 14 or 1 day.

Then, when the end switch is turned on when the capture of the operation amount is completed, the control microcomputer 11 ends all processing.

Note that the teaching data memory 2o stores the welding robot at that time before maintenance or when teaching a welding point by attaching the engaging terminal 81 to the robot instead of the welding ring and performing almost the same operation as described above. 1
It is assumed that the motion amount data for each axis of the robot regarding the position and orientation of the robot is stored.

On the other hand, if, for example, the teach mode is selected and the sampling link switch 21 is turned on at the same time as the correction mode switch 13 is turned on, the correction microcomputer 12 proceeds to steps 5TEP8 and 9 of the processing program flowchart shown in FIG. So 5
After checking whether the transfer of operation amount data from the control microcomputer 11 has been completed at TEPIO, if it has not been completed, the controller waits until the transfer data is received at 5TEP11.

Then, the correction microcomputer 12 receives the motion amount θ+' + for each axis of the robot, which is transferred from the control microcomputer 11.
fla' + βB' y 04' + θ, / , (
1, 1 is received, those received data are 5TEP1
After temporarily storing the data in the internal RAM of the correction microcomputer 12 in step 2, the process proceeds to step 5TEPL3, where the following analysis process is performed.

In other words, the new installation position of the welding robot 1 is set to the origin Q.
The coordinate position P and direction vector of the intersection (see Figure 2) in the new orthogonal coordinate system (Or, %, jrk'), where V is P p (P + u + 1)) = [M +]
It can be determined by the conversion formula XPO(θ1+('2+'3+04°05+0a)).

Here Pp is P, u,? ! 7 is the new orthogonal coordinate system (○′;
"+"-+"), and Po indicates that θ and θ6 are data in the local coordinate system of each axis of the robot.

Further, [M+] is a coordinate transformation matrix, which can be determined by the axis configuration of the welding robot 1.

Specific examples of the above conversion formulas are as shown in the following formulas (L) to q).

t nahchi-E”, E”, E”, E”,
B'' are the local coordinates (i, jl
Let the t-axis and J-axis in k) be rotation transformation matrices (rotation tensors) around the axes, and let go be the reference norm vector (direction vector indicating the grip l direction of the welding gun when all 0s are [0]). , P(x, y, z)=C', +c! H+E"'E""(
已) 10E'.・E″6″E6″(C′4) +E′.’E″E″E″’ (C4+Gj)=−・−−
-・-Q)u(ytl, u 2 +"3)=E"'
E"'E"''E"' (Cm +G5) −
・・■? ! 7(t'+lυ2#u3)=E”°E4,
, aO4E@0rEA(,,.)・------@i
llλ (C1+C2+L3 +C4, Cs +G
6, see Figure 2).

Therefore, in 5TEP13, the operation amount e'1+ O;tβ3. which was stored in the RAM in 5TIEP12. θ≦, θ−+
06 is read out, analysis processing is performed using equations 0 to 0, and the coordinate position data p'(z',H') of the position P in the new orthogonal coordinate system (○': i'j', /i'') is obtained.

2'), and the direction vector '(?L'!, 1L2
, ? t'3) r M) (υ'1.υ;, υ3
).

Next, in 5TEP14, P', u' obtained in 5TEP13
.

After storing U' in the internal RAM, the process advances to 5TEP15.

At 5TEP15, it is checked whether or not the correction switch 22 is turned on. If it is not turned on, the process returns to 5TEP 9, and if it is turned on, the process goes to 5TEP16.

If the sampling switch S is also turned off when the correction switch 22 is turned off, the loop of 5TEP9, 15 is repeated, and if the data transfer is completed even if the sampling switch 9 is turned off, the loop of 5TEP9°10, 15 is loop is repeated.

Therefore, when the correction switch 22 is turned on, the correction microcomputer 12 proceeds to 5TEP16 and sequentially calculates the movement amount 01 to θ6 of each axis of the robot regarding the reference post 80, which was taught before the installation error occurred, from the teach data memory 20. After reading, the same analysis process as in 5TEP13 is performed based on the operation amounts 01 to 06 in 5TOP17, and the plane orthogonal coordinate system in which the teach data stored in the teach data memory 2O is defined, that is, this embodiment Now, the orthogonal coordinate system (o; i) of the welding robot 1 before maintenance.

Coordinate position data p (j, k) (see Figure 2)
Oc', y, z) and the direction vector u(lt,
.

tt2. u3) and v(υIl v21 v3).

Then, the coordinate position data P, u, v obtained are 5TEP.
At step 18, the data is temporarily stored in the internal RAM and the process proceeds to 5TEP19.

In 5TEP19, the plane orthogonal coordinate system (0; i, , z,
k) and the new orthogonal coordinate system (016t' L, J l
, kl ) Note: e represents the inner product of the base vectors of the plane coordinate system, ΔX,
Δy and Δ2 represent the amount of parallel movement in each axis direction from ○ to O'.

To find out, we perform the following calculations.

That is, the simultaneous equations stored in the internal RAM at 5TEP14.18 are set up, and by solving these equations, the coordinate transformation matrix M2 is created and stored in the internal RAM.

Next, in 5TEP20, the teach data in the plane orthogonal coordinate system that was used before the installation error occurred and is stored in the teach data memory 20 [the amount of movement of each robot axis θIj+02'r+'3L+04t+O5i0ei( ""l to n, n is the number of points in the teach data)] is read from i = 1, and analysis processing based on the 0 to 0 formula described above in 5TEP21 is performed to create the surface orthogonal coordinate system (Oi' + Jy) described above. ) at Pi, (
zi, yi,, zi,), ui (ail.

u' 2 + tz L 3 ) r U L (u
Add L 1+ u L 2 r u L 3.

Then, the coordinate transformation matrix M2 stored in the internal RAM is read out in the 5TEI) 22, and the coordinate transformation matrix M2 is read out as Pi (Pi).

ui, vi) = M2 Pi-'(PL', ui
'.

1171,'), that is, by setting up the following coordinate transformation formula, the new orthogonal coordinate system (0/;,", a/
.

P i' (z i,' , L', z
i,').

LL'l-' (tj L1' + u L2
' r u 'l-' ) g vL 'and therefore P L', 1. L7. ' , vL' 5TEP2
By performing the following synthesis process in step 3, new teach data in the new plane coordinate system θ 1L
' , θ 2j' y '2j' + 194j'
+05λ'.

θei' ('=1 to n)].

That is, P I = (01L' + 02L'HI2
2゜′.

04. ,'r θs% + OGL') = f (sword i,', ML'.

7%, 111. (', 1Liz', uL5'
The function f in , ui, (' , ui%υL3') is a complex nonlinear form, but in the above-mentioned insertion, C', +c: =R-12("+z+ !f+? 7+
2) E"'E"' (W) = R34(" 34
＋! f 341 7 34) x, :'01 E 7'
E"" (C4) "'R56("ss+y661
zss) E"'E"'E"E"'(C'5 +
G'a)=R? 8(”7B'l !p 781778
), then Ol, C2, p3 are.

0 1"fan -' (,!f 34/ "34)C
2='tan-1(-4-Z34/f)72
3:6-7 can be expressed as Z4.

Also, E""""r:k(-'1)R78=("
k+! l k+ zk), then 04+05 is.

(14==tan-' (y k / Guo k,)0
5:t,an-' (±zk/m).

Furthermore, by transforming the above-mentioned equation G), E″+(Jl″
'E"-")E""")E1'c'-0°ゝ(υ)=E
”' (side 0), and since the left side of this equation can be assumed to be known, we can write this as (Sc+ + !11+ zt )
and vo (':'o + !for Zo), C6 is θe-"tan-1[(Zo !fx
+y0 Zx)/(yo yl +zo z,)].

Therefore, new teach data can be obtained by calculating each of the above equations.

Next, in 5TEP24, after storing the new teach data obtained in 5TEP23 in the teach data memory 20,
At 5TEP25, check whether the process is completed with i = H, and if it is not completed, return to 5TEP20 and perform 5TEP.
Each process from 20 to 24 is repeated, and if completed, the correction process is complete.

Then, after completing the correction of the teaching data mentioned above, select the playback mode and start the welding robot 1.
The same welding work as before maintenance can be performed.

In the above embodiment, the teaching data of a welding robot is corrected, but this is not limited to this, and this method can be applied to correcting the teaching data of any playback type robot. The invention is equally applicable.

Furthermore, the present invention can be similarly applied to correction of teaching data when teaching a robot at a location other than the installation location or when performing automatic teaching on data using a CAD system that has been actively developed in recent years.

However, when performing such correction, the unique position and orientation data of the robot in the orthogonal coordinate system in which the tee data obtained by each teaching is defined,
That is, it is also necessary to teach the data determined by the reference post and the engagement terminal described above.

As described above, according to the present invention, the position and orientation of the robot are determined using a reference post that determines the position and orientation of the robot, and an engagement terminal that engages with the reference post,
The determined state is measured using the robot itself as a measuring instrument, a coordinate transformation matrix is determined, and the teaching data is corrected based on the coordinate transformation matrix. This eliminates the need for equipment, which not only greatly reduces maintenance costs, but also allows maintenance of multiple robots on an assembly line at once, or even if the assembly line itself is modified, it is not necessary to correct the old teaching data or perform other tasks. Teach data can be easily corrected based on pre-teaching work at the location, without significantly reducing the line operating rate.

Furthermore, the teaching data can be corrected even if a sufficient work area cannot be secured around the robot.

[Brief explanation of drawings]

FIG. 1 is an external view of a multi-axis robot showing an embodiment of the present invention, and FIG. 2 is a skeleton diagram for explaining the motor functions of the multi-axis robot shown in FIG. 1, using graphical symbols. Fig. 3 is a plan view of the reference post, Fig. 4 is a front view of the same, Fig. 5 is a side view thereof, and Fig. 6 is a plan view of the engagement terminal. 7 is the same (front view thereof, FIG. 8 is a block diagram of the control system of the welding robot 1 shown in FIGS. 1 and 2, and FIG. S is the microcomputer for controlling the robot shown in FIG. 8. FIG. 10 is a flow diagram showing an overview of the processing program of the correction microcomputer shown in FIG. 8. 1...Welding robot S...Control device 10...・Teaching box 11...Microcomputer 12 for robot control
・Correction microcomputer 16...Correction mode changeover switch 14.19.23...Interface circuit 18A~
18F...Axis position detector 20...Teach data memory 21...Sampling switch 22...Correction switch 80...Reference boss 81...
・Engagement terminal Fig. 1 Fig. 2 1・

Claims (1)

[Claims] 1. In a multi-axis robot equipped with an axis position detector on each axis muscle, a reference post for determining the posture of the robot is provided at a predetermined position in the operating area of the multi-axis robot, and a reference post is provided at the tip of the multi-axis robot. An engagement terminal that engages with the reference post to determine the position and orientation of the robot is provided, and when the engagement terminal is engaged with the reference post, each axis position detector obtains each axis position. After obtaining robot position and orientation data in the new orthogonal coordinate system of the currently installed multi-axis robot based on the amount of motion, these position and orientation data and the teaching data of the multi-axis robot are defined. A coordinate transformation matrix between the old and new orthogonal coordinate systems is created based on the position and orientation data corresponding to the position and orientation data in the old orthogonal coordinate system, and based on the coordinate transformation matrix, A robot teach data correction method characterized by the ability to correct the teach data of a multi-axis robot.