JPS62226307A - Robot device - Google Patents

Abstract

PURPOSE:To exactly move an arm to a targeted coordinate position, by finding a deviation between the position of a commanded coordinate axis, and that of the coordinate axis of a robot arm moved according to the position of the commanded coordinate axis, and operating the robot arm according to a corrected move quantity. CONSTITUTION:A main control part 9 is operated as the one to control to find the assembling error of a robot 2, and is equipped with a deviation arithmetic part 9-1 and an error arithmetic part 9-2. The deviation arithmetic part 9-1 has a function to find the deviation (in an X-axis and a Y-axis directions) between an image pickup position (the position of a grid intersection) in a work 7 instructed to an image pickup device 5 receiving a picture data from a visual arithmetic part 8, and the position of an intersection image-picked up at the image pickup device 5. Also, the error arithmetic part 9-2 has the function to find at least the assembling errors (rotational displacement quantities) of the arms 3 and 4 of the robot 2, and the arranging position error (rotational displacement quantity and in the X-axis and the Y-axis directions) of the work 7, receiving the deviation obtained by the deviation arithmetic part 9-1.

Description

[Detailed description of the invention] [Purpose of the invention] (Industrial application field) The present invention relates to a robot device.

(Conventional technology) For example, there are two robots used in factories.

Teaching is carried out in advance to instill the working movements. In this teaching, the robot's trace source is turned on, the worker moves each arm of the robot along the exact same path as when working, and the amount of movement of each arm at this time is detected using, for example, a rotary encoder. and memorize the number of pulses. and,
During actual work, each arm moves along the taught path VCG according to the stored number of pulses.

However, when the robot is operated by sending coordinate values from a computer, a problem arises in that the robot does not reach the target coordinate position due to assembly errors of the robot.

(Problem to be Solved by the Invention) As described above, since the manipulator cannot accurately move to the target position due to assembly errors, this error is corrected by some means to eliminate the effects of assembly errors. It is required to avoid being exposed to

Therefore, the present invention aims to solve the above problems.

It is an object of the present invention to provide a robot device that can accurately move an arm to a target position by estimating and determining assembly errors.

[Structure of the invention]

(Means for Solving the Problems) The present invention calculates the deviation between the commanded coordinate axis position and the coordinate axis position of the robot arm that has moved in accordance with the commanded coordinate axis position by means of a deviation calculating means, and corrects the deviation by the calculated deviation. The robot apparatus is characterized in that the robot arm has a function of moving to the command coordinate axis position according to the moving lid.

(Function) By providing such a means, 1. The robot arm moves to the commanded coordinate axis position according to the amount of correction due to deviations such as assembly errors.

(Example) Hereinafter, an example of the present invention will be described with reference to the drawings.

FIG. 1 is an external view of the robot device. A SCARA set of articulated robots 2 are provided on a base 1. The robot 2 is equipped with arms 3 and 4, and at the tip of the arm 4 is a device composed of a plurality of solid-state imaging devices CCD. The robot 2 receives a control signal from the robot controller 6 placed on the base 1 and, for example, controls the arm 3t.
- The arm 3 moves in response to a pulse signal of the number of pulses required to move it A degree to the right. Further, a workpiece 7 is placed on the base 1, and a lattice pattern is formed on the surface of the workpiece 7. 8 is a visual arithmetic unit, which sends an imaging command to the imaging device 5 and sends an imaging command to the imaging device l! It has the function of taking in the image signal from 15 and processing the image.

Now, 9 is a main control section, which performs control to determine the assembly error of the robot 2.

In particular, as shown in FIG. 2, it has a deviation calculating section 9-1 and an error calculating section 9-2. The deviation calculation unit 9-1 receives the image data from the visual calculation unit 8 and calculates the position between the imaging position (lattice intersection point) on the workpiece 7 instructed to the imaging device 5 and the intersection point imaged by the imaging device 5. A device with the function of determining the position and Q deviation N (X-axis direction, Y, M direction)
The differential controller 9-2 receives the deviation determined by the deviation calculator 9-1 and calculates at least the assembly error (rotational deviation fi) of the robot 2 and arms 3 and 4? and workpiece 7 placement position 1
It has the function of determining the difference (amount of rotational deviation and the X-axis direction and Y-axis direction).

Next, before explaining the operation of the apparatus configured as described above, a method for estimating the assembly error of the robot will be explained with reference to FIG. Here, the lengths of the arms 3 and 4 of the robot 2 are 11 and 2, the assembly errors (displacement amounts) of the arms 3 and 4 of the robot 2 are Δθ1 and Δθ2, and the 1 difference in the placement position of the workpiece 7 ( ) in the X direction and Y direction, respectively.
ΔX, ΔY, ψ in the direction and direction of rotation.

Therefore.

These deviation amounts Δθ1, Δθ2, ΔX, ΔY, and ψ are the deviation amounts to be determined. Then, the imaging target coordinate position of the grid instructed to the imaging device 5 is expressed as Tlj(xo, yo
), and the coordinate position of the grid that was actually imaged is T1.
Let it be (XO+desire, YO+ΔY). Further, the position instructed by the imaging device 5 is assumed to be Sij, and the actual position is assumed to be Sij. Note that l is the l-th measurement point, and j is the orthogonal basis of the m-dimensional measurement space.

Here, the grid point shift *Rlj is Ttj-StjT;
h! ).

It is obtained from the image taken by the imaging device 5 from 1j4M. Furthermore, the expected value of the deviation RiJ is Tlj-Slj, which can be obtained from FIG. Therefore.

TI, = Tame + Mayuzumi + (X- person) □□□ψ1 (to Y-) -
ψ・・・・・・(1)Tl,=”40ΔY+(
X-person) thin ψ+(Y-expletive) (self) ψ...
・・・・・・(2)81,=J, Cold θ1 Anger θ,)+4 Cold Peθ1+Δθ□+02+△θ2) ・・・・・・(3)
8 1t = 4 XaLn(θ, +Δθ1)+A XS
LII(θ1+Δθ1+02+△θz) -Old-(4
), and the amount of deviation of the grid points obtained by the visual rendering unit 8 is defined as Aid, AI, ? If so.

Rl,-Ri1=AI,-(Ti,-81,)
...{5}R1,-R1,=Ai2
-(TI, -812) - Old...(6) is powerful. and.

dij = Rij −R lj, and if this dl is sufficiently small, 'l'1
j.

S1j is a good approximation. However, if dij is not sufficiently small, each deviation from dlj', ΔY, Δθ
of.

Δθ2. We must add ψ. Therefore, if dij can be converted into an edge shape for a shifted grave, the shifted grave can be determined by the method of least squares. Furthermore, in order to eliminate errors caused by linear approximation calculations, the obtained deviation proof is incorporated into the original equation and repeated calculations are performed.

In addition, in the case of the first approximation, the amount of deviation such as ΔY is expressed as rOJ
Assuming that, the H4 calculation is executed. Now, dlj=R1j
- phase j, since RlJ = Tij -5ij)11j = Tij -Slj.

dlj = ('l”ij −5iJ) −(TIJ
−8lj )=(Tlj −Tlj ) −(Sl
j −8ij) Here, dX, dY, dθ1. d
If i12 and dψ are to the left of the common deviation 1t, then TI, = dog + ΔX + ctx + (X - XO) cm (ψ
10dψ) −(y−Yo)sin(ψ→−dψ)Tl,
=Yo+Δy+tiy+(x-”Q)ain(ψ+d
ψ) +(y-Yo)cos(ψ1dψ)st1=4X
(2) (01+dθ1 to θ1+)+4×ω to θ1+Δθ
, +d0, 10θ2+Δθ20dθ,) Sl, == AX introduction θ, +Δθ, +dθ,) + 4x thin (
To θ1+, 01+dθ, +02+Δθ2+dθ,) is found, and when dX, dY, dθ1−, dθ2, and dψ are edge-shaped, we get TI, == because+”+ ax+ (X−Xo)wψc
osdψ-(X-person) sirl/ sin d9'
(Y-Yo)-nψcosd9+-(Y-Yo)co
sψ5jndp* dX+ (-(X-X,, )si
nψ−(Y−Y,)−ψ)dψ+X,,+−ff+(X
-X,,)cwψ-(Y-Y-n smcp
・= −(8) (1 in CDadψ. dψ in windψ T lx = YO+ΔY+d Y 10 (X XO)
5urtpous d9+ (X Xo) wsψ
5iud$10(Y-Yo) casψccadψ-(Y
-Y, ＞-ψ5andψ*dY+((X-X,,)a
mψ−(YYo)sL119) aψ+Y, +ΔY+(
X-X, ) sin ψ + (Y-Yo) ays'p = (9
)81, = 71an(θ1+Δθ, )ays dθ
r 4” (θ1+Δσ, ) group, dθ1+ 4ays (
θ1+02+Δθ1+Δθ2)−(ctθ,+do,)
−AaIn(θ, +θ, +Δθ1+Δθt)=(dθ,
+dθ2) (-A-(θ, +Δθ,) -4"('
s+θ2+ΔD, 10Δ0. ) ) -dll.

−No (θ□10, 1Δ01+02)・dυ2+J
,”('I+liθ, )+J, cu-(σ1+
02+Δθ, +thD, )−(lO)Sis=J
-m(θ, +Δθt)casdθ1+ノ, O)! (
θ1+Δθ,) aindθ1detail(θ,+02+Δθ
1 + Δθ, ) (2) (dθ, +dθ2) ten, ring (θ1
+θ2+Δv1+Δθz)” (dθ, -1-d02] (), ■(θ, ten Δθ1)+for ω, θ1+02+Δθ1
+Δθ, ))・dθ10 no f gloss (θ1+θ, 10Δυ, +
Δ02)・dθ20, sin (θ, 10Δθt)+
X=(θ1+02+Δθ1+′t)−(11). Therefore, formulas (1) to (4) and (8)
From equation to equation (11).

TI□−TI, =:dX+(−(x−)stnψ−(
Y-Yo)a)sψ)-dψbe12) TI, -'l'1
2=dY+ ((X-X,,)amψ-(Y-Y,)"
ψ)・dψ −(13)81, −St, :(−)−m(
θ1+Δθ+)-X=(θ1+θ2+Δθ1+△θ2)
)×dθ1 ~ノ2aIJI(θ, 10θ2+Δθ1+Δθ2)・dθ
2...-(14)81,-8i,: (), nar ("
t+Δθ□)+4 sounds (θ1+02+Δθ1+Δθ2))
8dθ1 +120 (θ1+02+△θ1+Δσ2)・di7.
...(15) is obtained, and from these equations (12) to (15) and the above equation (7), di1 = (Ti, -TI,) - (81, -8l,) =d
X+(-(X-X,,)=ψ-(Y-Y'o)cosψ
)-dψ-(-A-irl(θ1+Δθt)-J2=(
σ1+θ, +Δθ1+Δθ, ))・dθ.

+11m (θ1+θ2+Δ01+Δθ, Edθ2・-・
(t□) di, = (Ti, -1"1.) - (S12-
81. )=aY+[(X-Xo)cosψ-(Y-Y
o) unψJ−dψ−(A−(θ, +ΔθI)+4
αTomi(θ, 10, +Δθ, 1Δθ2))・dθ1−ノ,
cos (θ1+02+Δθ1+Δθri・d02
``'C circle) is found.Therefore, by substituting dl of these equations (16) and (17) with the deviation lid angle obtained by the visual device 8, % and dl are given as Δ
Substitute Y. Note that Equations (16) and (17) are for one feeding point on the grid. Therefore, by measuring a plurality of n points on the grid, the 2n equation can be obtained.

By solving these equations, each deviation is 1α and ΔY.

Δθ1. Δθ2 and ψ are determined.

Then, equation (5), equation (6), and (16)
From equation and equation (17).

ax+(-(X-Xo)'oψ-(Y'to)=x=
ψ) dψ−(−AairI(01 to θ1+)−No m(
θ1+θ, +Δ01+Δθ, ))−dθ.

Ten no 2stn (θ, +θ2+Δθ, ten△θ2)・dθ2
= A s r (Xa 書■+ (X 'xo)
asψ−(YYo)”ψ)+ (7!, tts(0,+
Δθ, ) +J2cos(θ1+02+Δθ1+Δθ
2))dY+((X−XO)coBψ−(YYo)”ψ
)dψ−(),■(θ1+△θ1)+4th (θ,+02
+Δθ1+Δθ2))dθ.

-No2's (θ1+θ2+Δθ1+Δθ2)・dθ.

=Ai, -(Y, +ΔY+(X-X,,)amψten(Y
-YO)amψ)+(JI$In(θ1+Δθr)
+-1tx” (θ1+02+Δθ1+Δθ2))?
I can fall. Two such equations are obtained for one measurement point. Therefore, when performing nA measurement, as mentioned above, 2n
formulas are obtained. As mentioned above, at first, ΔY
, Δθ1. Approximate calculations are performed using Δθ2 and ψ. Then, dX, dY, determined by the least squares method,
dθl, dll2 j? and dψ as ta, ΔY, Δ
θ1. In addition to Δ02 and ψ, approximate calculation is performed again. Then, when the dlj value becomes biaxial, the process ends.

After this, each deviation ΔX, ΔY, Δθ, Δθ2 and ψ
When obtained, the actual (Xo+Δx, yo+ΔY) (2) position can be found. Therefore, this actual position is NX+Ny
If so.

NX=XO+ΔX+(X-Xo)-sψ-(y Yo
)”ψN,=Y,+Δγze(XXo)ainψ+(Y−
Yo) ooaψ. Then, from these values, the robot ■ Movement value θIs? and θ2.

θ, = a tan (N y/N x )−acx
s((NX +Ny+No-4')/(2xAx-IN
,” Jukyuzuki-Δθ1 θ, =: a tan ((Ny-), Hoso(0, +Δ
θs))/(Nx A-<θ, 01))-(θ,
+ΔθI)−Δθ2. Therefore, the actual movement of the robot 2 in each direction of the X1111o and Y axes m NNx p NNy is.

NNX =ノ1cosθ1+ノ,cos(θ1+θ,)
NNy = 1 x θ1 + 2 ain (θ1 + θ, ).

Next, the operation of the apparatus configured as described above will be explained with reference to the flowchart for calculating the deviation of fireflies shown in FIG. In step 81, the main controller m9 issues to the robot controller 6 a target position (2) instruction for imaging a predetermined intersection of a grid formed on the workpiece 7 with the imaging device 5. As a result, each arm 3, 4 of the robot is controlled by the robot controller 6.
The imaging device 5 moves to the target position in response to a pulse signal from the imaging device 5.
reaches. Then, in the next step S2, an imaging instruction is issued to the imaging device 5. As a result, an image signal obtained by capturing an image of the intersection of the grid is output from the imaging device 5, and this signal is sent to the visual arithmetic unit 8. When imaging for the intersections of this grid is completed.

It is determined in step S5 whether all the intersection points of the grid to be imaged have been completed. In this judgment, since one location is to be imaged, the process moves to step S6, and a position instruction for the next grid is issued from the main controller 9 to the robot controller 6. When the imaging of all imaging positions for the grid intersections is completed in this way and the image data is obtained, the process moves to step S7, and the deviation calculation unit 9-1 calculates the coordinate position obtained by imaging, for example. Shown in FIG. 3 (XI, Yl).

(X2, Y2) and the specified target coordinate position (XtY
).

Find the deviation proof from (x', y'). Next, the process moves to step S8, and the visual calculation unit 9-11c! ? The respective deviation amounts for each measurement position determined by the calculation are sent to the error calculation section 9-2, and this error calculation g9-1 substitutes dll and d12 into the above-mentioned equations (16) and (17), For example, if the number of measurement points is 4, I will obtain 8 Q equations.

Then, by calculating each of these equations, each deviation amount is determined by X, ΔY, Δ01. Find Δθ22 and ψ.

Ru. Once these deviations have been determined, the process moves to step S9, where the movements bkNN and NNa of the robot 2 relative to the target position are calculated and determined. Thus, no displacement occurs due to the movement of each arm 3°4 due to these movements @*NNx and NNy.

In this way, in the second embodiment, the deviation between the imaging position on the workpiece 7 instructed to the imaging device 5 and the imaging position of the imaging device 5 is determined, and the tray Δ171. Δθ2
Since the placement position deviations △X, ΔY, and ψ of the workpiece 7 are calculated, the correction of the present invention is performed for cases where the deviation electric current with respect to the imaging target position is 5 times as shown in the table below. By doing this, it can be kept within 90.318.

Therefore, even if there is an assembly error of the robot, this error can be corrected and each arm 3.4 can be accurately moved with respect to the target coordinate position. This correction only needs to be performed once before shipping the manipulator, and there is no need to modify the teaching. Therefore, when operating the manipulator 2, the manipulator 2 can be moved accurately simply by sending the desired movement coordinate position.

It should be noted that the present invention is not limited to the above embodiment, and can be modified without departing from the spirit thereof. For example, it can be applied not only to SCARA single-type articulated robots but also to all kinds of robots.

〔Effect of the invention〕

As described in detail above, according to the present invention, it is possible to provide a robot device that can estimate and determine 5 assembly errors and accurately move an arm to a target coordinate position.

[Brief explanation of drawings]

FIG. 1 is an external configuration diagram showing an embodiment of a robot device according to the present invention, and FIG.
A detailed functional block diagram of the controller, FIG. 3 is a schematic diagram for explaining the deviation amount estimation of the present invention, and FIG. 4 is a deviation amount estimation of the present invention device! ! It is a flowchart. 2... Robot), J, 4... Arm, 5... Imaging device. 6... Robot controller, 7... Workpiece, 8.
...Visual calculation section, 9...Main calculation m*9-1...Deviation calculation section, 9-2...Error calculation section. Applicant's agent Patent attorney Takehiko Suzue, s 1
Day figure 2

Claims (1)

[Claims] A function in which the deviation between the commanded coordinate axis position and the coordinate axis position of the robot arm moved according to the commanded coordinate axis position is determined by a deviation calculating means, and the robot arm moves to the commanded coordinate axis position according to the movement amount corrected by the determined deviation. A robot device comprising: