JPS59193770A - Locus control method of welding robot - Google Patents

Abstract

PURPOSE:To weld adequately easily works having variance in weld lines and groove shapes by using the coordinates taught with plural representative points of master and slave works in the stage of performing multi-layer welding with a robot provided with a linear interpolating function. CONSTITUTION:The welding start points P1, P2..., end points Q1, Q2... of the respective layers of a master work consisting of materials 1-3 to be welded and the coordinates of representative points A-I indicating the groove shape are taught. The coordinates of the representative points A'-I' indicating the groove shape of a slave work having approximately the same shape as the shape of the work 1 are taught. The coordinates for the welding start and end points of the respective layers of the slave work are calculated from said coordinates and the tip of a welding torch is moved on the straight line connecting the calculated welding start and end points. Then if the coordinates of the plural representative points indicating the characteristic possessed by the groove shape of the master work are first taught, adequate welding is accomplished thereafter simply by teaching the coordinates of the plural representative points indicating the characteristic possessed by the slave work for each of the slave work.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a trajectory control method when performing multilayer welding.

Conventionally, when performing multi-layer welding with a teaching playback type welding robot, the relative position of the weld line between the robot and the workpiece changes due to variations in the size and shape of the workpiece, and variations in the accuracy of the jig. Ta. Particularly when the workpiece is a medium-thick plate, the workpiece itself is heavy and large, making it difficult to improve the accuracy of cutting and bending during the part manufacturing stage, and the weld line and groove shape of a workpiece made of parts with uneven processing. changes greatly. Moreover, the jig for setting this large and heavy workpiece also becomes large, making it difficult to guarantee the precision of the jig from a profit standpoint, and the precision of the jig is generally poor.

Conventionally, as a method of combining these low-precision workpieces, jigs, and welding robots to enable proper multilayer welding without causing welding defects in the outermost layer, we have used a "trajectory correction function using a weld line detection sensor". (Unexamined Japanese Patent Publication No. 30-/39tt
tt3) etc. have been proposed in J-7. However, although this method can follow variations in the weld line, it cannot recognize the groove shape, and the amount of deposited metal cannot be controlled, resulting in welding defects, and the outermost layer requires manual repair welding. It had disadvantages such as the need for To eliminate this drawback,
It is also possible to perform multi-layer welding by detecting not only the weld line but also the groove shape using a vision sensor using a television camera for canals, but this method requires a complicated and expensive robot system including the sensor. In addition to this, there are also problems with the environmental resistance and reliability of the sensor, and it is still at a stage where it cannot be put to practical use.

Therefore, in applications where an expensive welding robot with a shape recognition sensor is not profitable as described above, it is necessary to (1) abandon robot automation or (2) use only a simple position correction function without controlling the amount of welded metal. The only options available were to use a robot with a seven-track sensor or (3) correct the multi-layered teaching trajectory for each workpiece by teaching operation. When manually teaching all the trajectories after the first layer of multi-layer welding using a teaching pendant, it is necessary to teach while assuming the welding conditions up to the previous layer in advance, but it is necessary to teach the welding current, welding voltage, welding speed, work material,
This is difficult because the bead width of each layer varies widely depending on the plate thickness, groove shape, etc. For that reason, the second layer.

It is necessary to correct the path of the third layer by trial and error, and it is also necessary to change the welding conditions according to the change in the groove shape. Method (3), which requires a very long time and requires difficult teaching work for each work, is worthless from the point of view of automation.

The present invention was proposed in view of the above-mentioned problems, and it is possible to properly multi-layer workpieces with variations in weld line and groove shape by simply teaching a plurality of representative points representing the characteristics of the groove shape for each workpiece. The purpose is to provide a trajectory control method for welding robots that enables welding.

The present invention provides the taught coordinates of the welding start point and welding end point of each layer of the master work, the taught coordinates of a plurality of representative points representing the characteristics of the groove shape of the mask work, and the characteristics of the groove shape of the slave work. The coordinates of the welding start point and welding end point of each layer of the slave work are calculated from the taught coordinates of multiple representative points representing The feature is that multilayer welding can be performed by moving the tip of the torch. In other words, first teach the coordinates of the welding start point and welding end point of each layer of the master workpiece, and the coordinates of a plurality of representative points representing the characteristics of the groove shape of the masterwork. By simply teaching the coordinates of multiple representative points representing the characteristics of the groove shape, the bath contact start point and welding end point for each layer of the slave work can be determined, and the weld line can be determined.

Appropriate multilayer welding and separation of slave workpieces with uneven groove shapes is possible.

The above is the gist of the present invention, but the present invention further provides that each layer of the mask work is determined from the coordinates of the plurality of representative points and the coordinates of the welding start point and/or weld end point of each layer, which are taught in the master work. The groove cross-sectional area of each layer of the slave work is calculated from the coordinates of the plurality of taught representative points of the slave work and the calculated coordinates of the welding start point and/or welding end point of each layer. The welding speed calculated by multiplying the ratio of these two open cross-sectional areas of each layer by the taught welding speed for each corresponding layer of the mask work can also be used as the welding speed for each layer of the slave work. When the groove cross-sectional area of the slave work is different from that of the master work, the variation in groove cross-sectional area is corrected by the welding speed,
Other welding conditions are the same as those for the master work, so that an appropriate amount of deposited metal on the slave work can always be guaranteed.

Still, the present invention is a master work taught,
The groove width of each layer of the master work is calculated from the coordinates of the plurality of representative points and the coordinates of the welding start point and/or the welding end point of each layer, and the groove width of each layer of the slave work is determined from the coordinates of the seven points of the slave work. Calculate each tip width, take the difference between the groove width of each layer of the mask work and the taught weaving width of each layer of the master work, and calculate the difference between the groove width of each layer of the slave work and the difference. Using the weaving width obtained as the weaving width of each throat S of the slave work, move the straight line connecting the welding start point and welding end point of each layer calculated above of the slave work while weaving the tip of the welding torch (weaving Welding) is also possible. This is to determine the weaving width of each layer of the slave work so that the distance between the weaving edge of the slave work and the groove width is the same as that of the master work. It is possible to eliminate excessive welding to the wall of the slave workpiece that occurs when the groove width is larger than the groove width, or insufficient welding that occurs when the groove width is too large, thereby preventing the occurrence of welding defects.

Hereinafter, the present invention will be explained with reference to drawings of embodiments.

FIG. 1 is an external view of an example of a master work in which weaving multilayer welding is performed according to the method of the present invention, and the slave work also has approximately the same shape, although there are variations in groove shape and weld line. This master work consists of three welded members/, 2.3, and the groove shapes formed by these welded members/, 2.3 (points A, B, C, D, E, F, G.
As shown in Fig. 2, the trapezoidal shape consisting of
..., multi-layer welding is performed between the n-th layer and the n-layer. The shaded area is the third layer. Point Pl.

P2. ......, Pn is the first layer, the λ-th layer, ......, respectively.

This is the welding starting point in the n-th layer. Point Ql, Q2.・
・・・・・・．

Qn is the welding end point of the first layer, the λ-th layer, ..., the n-th layer, respectively. Here, a stationary orthogonal coordinate system 0-(x, y, Z) is taken, and from now on, the coordinates of each point shall be based on this orthogonal coordinate system 0-(x, y, Z). Welding start point PI of each layer when the master work is properly welded to the outermost layer (nth layer), P2. --------
, Pn coordinates (Xpt, Ypl, Zpi)
+(Xp2.Yl)2. Zp2), ---
----------, (Xpn, Ypn.

Zpn), welding speed Vl, V2. ......,Vn
, weaving width\fl, W2.・・・・・・、W
n and the number of layers n are determined in advance through trial and error. Representative points A, B, representing the characteristics of the groove shape at the welding start point,
Each coordinate of C, D (XA, YA, ZA), (X
B, YB, ZB), (Xc, Yc, Zc
), (XD, Mn, ZD) and the representative point L + I” + G + which determines the groove shape at the welding end point
Each coordinate of I (XE, YE.

ZE), (XF, YF, ZF), (Xc, Y
c, Zc), (Xr.

Y I + Zr) is manually taught using a teach box.

FIG. 3 is a diagram showing the positional relationship between the groove shapes of the master work and the slave work. Point A/, 13/, CI.

]]' and points 1C/, 171/, Q/, 1/ are grooves at the welding start point and welding end point of a slave workpiece that is misaligned with respect to the master workpiece and has variations in the groove shape. The coordinates of each of these points (X A', Y A' + Z A
' ) l (Xn ' + Y B' + Z B
' ) + (Xc', Yc', Zc' ), (
XD', YD', ZD), (XE', YE'
, ZE'), (XF', YF', ZF')
, (Xa', 'Ya', Zc') , (
Xr', YI', ZJ') are also taught manually by the teach box.

FIG. 4 is a diagram showing the groove shapes at the welding start points of the master work and slave work. Point (1°C2,...
..., Cn are the taught welding start points P+, P2. ......, Pn passes through the straight lines A and B [the intersection of the approximately parallel straight lines and the plane formed by the taught representative points A, C, and E (-t or G), point Dx , D2.・・Dn is the welding starting point P1゜1)2. ......, the taught representative point B is a straight line that passes through Pn and is approximately parallel to the straight line AB,
This is the intersection with the plane formed by D and F (or 1). Similarly, points CI', C2', ...... C, / are welding start points P 1 /, i) 2 of the slave work
/,...J-)n/, straight line A/ 7
A straight line approximately parallel to 31 and the taught representative point A'
, the intersection with the plane formed by C' and E' (Q/ at t'), points DI', D2',...
Dn' is the welding start point P1', P2', --,
This is the intersection of a straight line that passes through Pn' and is approximately parallel to straight line A'13/ and the plane formed by taught representative points +3', I)' and F/ (or 1').

The welding start point of the first layer of the slave work, the coordinates of P1' (
X'p Yuv Y'pi, Z'pi) if:, Most simply, it is independent of the mask work, that is, representative A, /, T3/, which shows the characteristics of the groove shape of the slave work.
C', f)' straight line A/B' midpoint and straight line C/I)
By dividing the straight line connecting the midpoints of ' into n equal parts, it is given by 22n 2 , In.

Coordinates of welding end point Qi' of the first layer of slave work (X
'Q i, Y'ci i, Z'Q i ) is similarly 2 2n ,! ! It is given by n 2 2n.

However, it is actually difficult to teach the height of the groove shape of the slave workpiece by dividing it into n equal parts in this way. In order to select appropriate welding conditions for each layer, in a strict sense, it may be better not to teach the welding condition divided into n equal parts. In the method of the present invention, which adopts the welding current, welding voltage, and weaving frequency of each layer of the master work as the welding conditions of each corresponding layer of the slave work, in such a case, the welding start point P1' and welding end point of each layer of the slave work are The length C1, η1 of the perpendicular line drawn from the point Q1' as described above is determined, and the welding starting point P1 of each layer is determined.
ridge height hi at welding end point Q1, embossment height k at welding end point Q1], and overall embossment height at welding start point side.

Welding end point (height k at till is heniε1+1
−C1 to 1−ηi+1−η] ∈=η1−0 The data of hj, ki, h, and k are stored in the memory. Next, the midpoint of straight line A'B'(E'P) and straight line C'y ((3
Find the length of the straight line connecting the midpoints of 'P', that is, the total height h' (k), and use the formula h'hj'==hi・− k′ to calculate the slave Welding starting point p 1/ of each layer of the workpiece
ridge height hi' at welding end point Q1', ridge height k at welding end point Q1'
Find hi, ', and ki', respectively, and connect them to the midpoint of straight line A'T3'(l'F') and straight line C/1 a (
By assigning it to the straight line connecting the midpoints of ()'I/
) is given by the coordinates of the welding end point Qj' of the first layer of the slave work (X'Q1. Y'Ql, Z'qi
−) is given.

Next, straight line A13. Straight line AC7 straight line AI! The 8 directions are (λ, μ, ν) and (a, b, c), respectively.

(a', b/, c/), these are each expressed by the following formula % Formula % Therefore, the length Ji of the straight line C1Di passing through the welding starting point P, which is the groove width of the first layer of the mask work, is , is given by .

Further, the area Sj of the trapezoid CiC +1Di-1-IDi, which is the groove cross-sectional area of the first layer of the mask work, is given by:

The groove width 1B' of the first layer of the slave work is point A.

13. Instead of the coordinates of C, E, points A', B', C',
Substitute the coordinates of 1shiI into the formula (31, +4) and +5j to obtain the direction cosine (λ, μ.

ν), (a, 1), C,), Q,), (a
', l)', C') and substitute these values and the coordinates of point P1) into equation (4) instead of the coordinates of point P1. The groove cross-sectional area S1' of the "th" layer of the slave work is also determined by an equation similar to equation (5).

The amount of welded metal per unit time of the slave work is the same as that of the mask work, and the welding contact value of the first layer of the slave work is determined from Vi' to properly weld down to the outermost layer with the same number of layers as the mask work. .

In general, the properly taught weaving width W in each layer of mask work, ld bevel width l +! Smaller Wi = l! i +1-δ engineering
The following relational expression (9) holds true. Here, δ]/
2 is the distance between the weaving end and the wall of the groove shape, and as mentioned above, when δ1 is small or negative, that is, the weaving width W1 is larger than the groove width Ji+1,
Input energy at both ends of the weaving causes over-melting of the walls, and conversely, a large δ causes under-melting of both walls. Either case causes welding defects. δ1 is usually taken to be /~3.

Master work groove width 11+1 determined by formula (9)
From the taught weaving width W1, δl in the first layer of the master work is δi = l i + t - Wl
α0). Slave work no. (
i++) groove width in layer/? The difference between 'i +1 and the weaving width Wi/ of the first layer is the above δ of the master work
If the value is set to be the same as l, the weaving width Wi/ in the first layer of the slave work is Wi' = fi + t-δ1
Demanded by wholesalers.

FIG.
FIG. 2 is a block diagram of an embodiment of a welding robot that performs weaving multiple welding of a null work using data of coordinates, welding speed (1'), and weaving Wi/.

10/ is a microprocessor that controls the robot control 1. 10. :2 represents the position of each axis of the robot based on the robot's own coordinate system in the orthogonal coordinate system 0-(X.

This is a coordinate conversion calculator that converts coordinates into Y, Z).

Reference numeral 103 denotes a coordinate transformation calculator that transforms the position iL6 of each axis of the robot expressed by the orthogonal coordinate system Q-(X, Y, Z) into the robot's own coordinate system. 70q is the above (1) ~
The coordinates of the welding start point Pl' of each layer of the slave work (X'p i, Y'p, t, Z'p i) are determined by the wholesale formula.
, coordinates of welding end point Qi'(X'Q,Y'Q 1
, Z'Q], ), welding speed ■1', and weaving width \■1/. 70 is a teach box. 10 B is a memory, as shown in Fig. 4, the coordinates of the welding start point P1 of each layer of the master work taught to the teach box 70, the coordinates of the welding end point Q], the welding speed l], the weaving width\
To'i,, Representative point A indicating the groove shape of the PM contact opening starting point
,, B, C, I) coordinates, representative point ID indicating the groove shape at the welding end point, coordinates of FI Gl■, representative point A1. indicating the groove shape at the welding start point of the slave work. B
The coordinates of /, c/, D/, the coordinates of representative points E/, p/, Q/, I/ indicating the groove shape at the welding end point, and the calculated groove cross-sectional area S1 of each layer of the mask work, The groove cross-sectional area Si' of each layer of the slave work, the difference δ1 between the groove width 11+1 of each layer of the master work and the weaving width νV1,
The coordinates of the welding start point P1', the coordinates of the welding end point Ql', the welding speed V]', and the weaving width w1' of each layer of the slave work are stored. The coordinates of each point above are converted to the orthogonal coordinate system 0- (X, Y, z) using the coordinate conversion calculator 10.2.
It has been converted into . Reference numeral 107 is a reference clock generator that generates a reference clock signal with period T. 10g is a linear interpolator, and the coordinates of the welding start point (Xpi, Ypi,
Zpi) or (X'pi, Y'pi, Z'pi
) and the coordinates of the welding end point (Xqi, , YQi, ZQi
) When the welding speed V↓ or V i ' is given, the distance m between the welding start point and the welding end point is calculated. , position increment amounts ΔX, ΔY for each reference clock signal output from the reference clock generator 10.

ΔZ or ΔX/, ΔY, and ΔZ' are calculated. 109 is a servo mechanism for each axis muscle of the robot. /10 is the servo mechanism 10 for each axis muscle of the robot
This is a servo drive circuit that drives 9.

/// is a position detector that detects the positions S'g], S2, S'g3°, . . ., S'gk of each axis of the robot. For convenience, the servo mechanism 109 and the servo drive circuit/101 position detector/// are shown together, but in reality they are independent of each axis of the robot. //2 is a switching circuit that transmits the command from the teach box 10 to the servo drive circuit 10 when in the teaching mode, and converts the coordinates to the robot-specific coordinate system using the coordinate transformation calculator 103 when in the playback mode. Converted position increment amount △X, △ for each reference clock signal
Y, △Z (△X', △Y', △Z/) are input commands for each axis of the robot Sgt, 8g2.8g3. .......

It is transmitted to the servo drive circuit/10 as Sgk.

//, 2 is a weaving constant setter, //3 is a weaving drive circuit, //q is a weaving actuator,
//ri is a welding torch. Figure 7 shows the robot arm//Weaving actuator attached to B//
It is an external view showing an example of IA and welding torch/15. Of the above, the coordinate transformation calculators 10.2, 103 and the calculator 70 are shown separately for convenience of explanation, but they are usually composed of seven microprocessors.

Next, the operation of the welding robot having the above configuration will be explained.

First, teaching work for mask work is performed. On the teaching box 10, the switching circuit //2 is set to the teaching mode, the servo mechanism 109 of the robot is moved, and the working point of the tip of the welding torch //S is positioned with respect to the mask work. As a result, the taught coordinates of the welding start point Pl and the taught coordinates of the welding end point Ql of each layer of the mask work in the robot-specific coordinate system are converted to the orthogonal coordinate system 0 by the coordinate transformation calculator 70.2.
−(X, Y, Z), respectively Pl(X
pl, Ypx, Zpl), P2 (Xpz, Y
l:12. Zn2), ...---, Pn (Xpn
, Ypn, Zpn), Ql (XQI, YQI
, ZQI).

Q2 (XQ2. YQ2. ZQ2), ・---
・, Qn (Xqn, Ycin, Zqn) is stored in the memory 10B. At the same time, the coordinates of the representative points A, B, C, I) and E+J''TG, I, which indicate the groove shape at the welding start point and welding end point of the mask work, are also taught, and the orthogonal coordinate system 0(x, y, z) coordinates (X
A, YA, ZA), (XB, Yn.

ZB), ......, (Xz, Yl, Zr)
It is stored in memory 10B as . Next, using the weaving constant setter//2, weave widths Wl, Wz, ---, fn and welding speeds Vl, V2 of each layer of the master work are set. ...-, Vn are set, and these data are also stored in the memory 10B. The teaching work for the master work described above, including teaching of welding voltage, welding current, and wire feed amount, is repeated by trial and error until these values become appropriate values. Thereafter, the microprocessor 10/ stores the coordinates (Xpl,
Ypl+ Zpl )+ (Xp2+ Yp2+ Z
n2) and the coordinates of representative points A, , B, C, D≠=≠(X
A, YA, ZA), (XB.

YB, ZB), (Xc, Yc, Zc), (X
D, YD, ZD) and the weaving width W1 of the first layer are read, and these data are input to the arithmetic unit 10il.

Arithmetic unit 1ott is (3+ (41451 formula, straight line A
Direction cosine of B (λ, μ, ν), direction cosine of straight line AC (a
, b, c), direction cosine of straight line AE (a/, b/
, c'), the groove widths of the first and second layers of the master work are determined by equation (6). Find I, 12. Furthermore, the computing unit 70t calculates the obtained groove width/? l and 1
2 and welding start point PL, coordinates of P2 ≠ engineering (Xpl, Yp
1, Zpl), (Xpz, Yp2.Zn2), and the direction cosine (λ, μ, ν), the groove cross-sectional area S1 of the first layer is determined from the equation (7), and δ1 is obtained from the αO) equation.

The thus obtained groove cross-section profiles Sl and δ1 of the first layer are stored in the memory 10B. The above two operations are repeated for the n-th layer, and the groove cross-sectional areas 81, 82, . . . . Differences between Sn and the groove width and weaving width of each layer δl, δ2. ..., δn is stored.

Next, teaching work for the slave work is performed. Representative points A', B', C', D' and E showing the groove shape at the welding start point and welding end point of the scrape work
The coordinates of /, P/, G/, and I/ are taught on the left side of the teach box 70. These coordinates are also transformed from the robot-specific coordinate system to the orthogonal coordinate system Q-(X, Y, Z) by the coordinate transformation calculator 10-2, and the coordinates (
XA', YA', ZA'), (Xn', Yn', Z
n..., (Xr', YI',
ZI') is stored in the memory 10B. Next, the microprocessor 10/ calculates the coordinates (XA'l) of the representative points A', B', ..., 1' from the memory 10B.
Y A' l Z A' ) l (X n' +
Yn' + Zn') +...

Read (XI', Yr', Z■'),
These data are input to the calculator 1041. Arithmetic unit 1
04t are these representative points A', B', ......, ■
/ coordinates (XA', YA', ZA'),
(Xn', YB', ZB'),...
, (XI', Yz', Zr') and from (
1), (2), the coordinates of the welding start point P1', P2', =-=-, Pn' of each layer of the slave work are (
Xlpl, Y'pl.

Z'pl ), (X'p2+ Y'p 2. Z'
p2), ・-・・-, (X'pn, Y'pn.

Z'pn) and welding end points Ql', Q2',...
, Qn' coordinates (X'QI, Y'QI, Z'QI
), (X'Q2. Y'Q2. Z'Q2),...
....

Find (X'cin, Y'qn, Z'cin). These coordinates are stored in memory 10B. Above, the representative points of the slave work found are A/ , B/ ,
Coordinates of C/, E/ and welding start point I) 1/,
From the coordinates of p 2/, the groove width 7?1' and the groove cross-sectional area Sl' of the first layer are determined by the calculator /θg in the same manner as in the case of the master work. The first part of this slave work
The groove cross-sectional area 81' of the layer and the groove cross-sectional area S1 of the first layer of the master work stored in the memory 10B. Welding speed Vl
From equation [8], the welding speed V+' of the first layer of the slave work is determined by the calculator 1011-, and is stored in the memory 1.
Stored in 0. Still, from the groove width Z2/ of the first layer of the slave work and δ] stored in the memo +)10B, the weaving width of the first layer of the slave work~■1' is also sent to the calculator using formula 09. Asked for in 〇gu, memo l)
/ Stored in 0B. The above operation is repeated for the nth layer, and the welding speed V]' of each layer of the slave work is stored in the memory 10B.

V2', .----, Vn' Towie hinge width WJ', W2', .=..., Wnt are stored.

When the predetermined data has been stored in the memory 10B as described above, the switching circuits // and 2 are switched to playback mode. The microprocessor 10/ starts from the memory 10B to the welding start point Px'(X'
p1. Y'p1. Z'ps) and welding end point (X'
Q+, 'Y'Q J, , 'Z'Q 1) Then, the welding speed Vl' is read out and set in the linear interpolator 10g, and at the same time, the weaving width W1 is also read out and set in the weaving constant setter //2. The linear interpolator 10g calculates positional increments △X/, △Y/, and △Z' for each reference clock signal output from the reference clock generator 10 using the Equation 1, and outputs these to the coordinate transformation calculator 103. do.

The coordinate transformation calculator 103 converts the orthogonal coordinate system Q-(X.

Incremental work amount △X/ , △Y', △ expressed as Y, Z)
Coordinate transformation of Z' is performed to the position of each axis of the robot based on the robot's unique coordinate system. Position input commands for each axis of the robot obtained in this way Sgx, Sgz, 8g3.・・Self・
・.

Sgk passes through switching circuit //2, position detector ///
The detection signals S'gl, S'g2. S'g3. ++
++++, S'gk The toe difference is given to the servo drive circuit/10 of each axis of the robot. Servo drive circuit/
10 performs PID compensation and the like on these differences, then amplifies the power and controls the position of the servo mechanism 709. At this time, the welding torch /15 is controlled by linear interpolation, and at the same time the weaving actuator /41 is controlled by the weaving drive circuit /
/3 and starts a weaving motion with a width Wl. This operation is repeated until the welding end point 917 points are reached.

Layer λ, third layer of slave work.

・・・・・・Welding of the nth layer is also at each welding starting point P
2′.

p3/, ,・, ,,, pnl, welding end point Q2', Q3', -----,,.

Qn', welding speed v2', V3', -・-, Vn
', weaving width W2', W3',...
..., are similarly controlled based on WH'.

In the above example, the groove shape of the workpiece is an inverted trapezoid, but it is a curved shape as shown in Fig. g (1) or a curved groove shape as shown in Fig. g (2).
In the case of a workpiece having an L-shaped groove shape, there are three representative points indicating the groove shape, so it is sufficient to set =Z1'. In addition, in the case of a workpiece with a U-shaped groove shape as shown in Fig. g (3), the point with the best linear approximation is the representative point C.
, D.

In FIG. 7, an example is shown in which a weaving actuator //4t is attached to the drive shaft //B of the robot, and this weaving actuator // performs welding)-f//3VC weaving motion, but the gist of the present invention is is to find the weaving width of each layer of the slave work from the relationship between the groove width of each layer of the master work and the calculated groove width of each layer of the slave work, and any means to achieve this can be used, and the robot Naturally, so-called soft weaving, in which weaving motion is performed as a composite motion of each drive shaft, is also applicable to the present invention.

The welding speed Vi' for each layer of the slave work was determined based on the ratio of the groove cross-sectional areas Si and Si' including the welding start points Pi and Pi' of the master work and slave work. Workpiece welding end point Q1
It may be determined by the ratio of the groove cross-sectional area including rQ''.The groove width IH, /!' may also be determined at the welding end point Q, Qi' side.

In addition, in the example, data such as the coordinates of the welding start point and welding end point of each layer of the mask work, the coordinates of the representative point indicating the groove shape, and the welding conditions (welding speed, weaving width) are taught using a teach box. However, for example, data obtained in advance from a design drawing or the like may be key inputted.

As explained above, the present invention enables proper multilayer welding of workpieces with variations in weld line and groove shape by simply teaching the coordinates of a representative point representing the characteristics of the groove shape for each workpiece. As a result, it has become possible to use robots to automate the welding of workpieces produced in small quantities of a wide variety, which is difficult to manage from a profit standpoint, both in terms of workpiece precision and jig precision. In addition, the accuracy of the workpiece and the jig are not managed with care, and since the workpiece can be rough, the jig is simple and the economic effect is significant. Furthermore, simply by having a normal welding robot with a calculation function for coordinate transformation perform the above-mentioned multiplication, it is significant that variations in the target workpiece can be corrected at low cost without the use of special expensive sensors.

[Brief explanation of drawings]

Fig. 7 is an external view showing seven examples of mask work to explain the trajectory control method of the welding robot of the present invention, and Fig. 2 shows the groove shape of the mask work for each layer to be multilayer welded, the welding start point, and the welding end. Figure 3 is a perspective view showing the groove shapes of the mask work and slave work, Figure 9 is a diagram showing the cross section of the groove on the welding start point side of the master work and slave work, and Figure S is a diagram showing the groove shape of the master work and slave work. A block diagram of an embodiment of a welding robot to which the method of the invention is applied, FIG. 4 is a diagram showing data stored in the memory 10B, FIG. S diagram is a diagram of a work (mask work, slave work) showing another groove shape. 1? Engineering, IF5. ..., Pn: Welding start point of each layer of the master work. Q ' + Q '' +..., Qn: Welding end point of each layer of the master work. Vl, V2...., Vn: Welding speed of each layer of the master work. Wl, W2...., ~■n : Weaving width in each layer of master work. A,, B, C, D, E, F, a, I:
A representative point that shows the master work groove shape. 11, e2. ..., /n: Groove width of each layer of master work. 81.82. ..., Sn: Groove cross-sectional area of each layer of master work δJ, δ2. ..., δn: Difference between groove width and weaving width of each layer of master work. AI, B/, C', I)', Iv, I", (3
', I/: Representative point indicating the groove shape of the slave work. 1) I/l) 2/,...'pn/: Welding start point of each layer of slave work. Q1', Q2',...Qnl; Welding end point of each layer of slave work. 11', p2',..., Jn': groove width of each layer of slave work. V1', Vz', ..., Vn': Welding speed in each layer of the slave work. Wl', W2',..., WB2: Weaving width in each layer of slave work. Patent applicant: Yaskawa Electric Co., Ltd. Flute 6
Figure 8 Figure 117

Claims (1)

[Claims] 1. When performing multi-layer welding in a welding robot equipped with a linear interpolation function, the taught coordinates of the welding start point and welding end point of each layer of the mask workpiece and the characteristics of the groove shape of the mask workpiece are Calculate the coordinates of the welding start point and welding end point of each layer of the slave work from the taught coordinates of the plurality of representative points representing the characteristics of the groove shape of the slave work, and the taught coordinates of the plurality of representative points representing the characteristics of the groove shape of the slave work, A trajectory control method for a welding robot, characterized in that multilayer welding is performed by moving the tip of a welding torch on a straight line connecting the calculated welding start point and welding end point of each layer. 2. The groove cross-sectional area of each layer of the master work is determined from the taught coordinates of the plurality of representative points of the master work and the coordinates of the welding start point and/or welding end point of each layer. The coordinates of the plurality of representative points and the calculated welding start point of each layer and/or
- Or calculate the groove cross-sectional area of each layer of the slave work from the coordinates of the welding end point, and multiply the ratio of these two groove areas for each layer by the taught welding speed for each corresponding layer of the mask work. The trajectory control method for a welding robot according to claim 1, wherein the calculated welding speed is used as the welding speed for each layer of the slave work. 3 The taught coordinates of the plurality of representative points of the master work, the welding start point of each layer, and /-! The groove width of each layer of the mask work can be calculated from the coordinates of the welding end point or the welding end point.
The coordinates of the plurality of taught representative points of the slave work and the calculated welding start point and/or of each layer.
Alternatively, calculate the groove width of each layer of the slave work from the coordinates of the welding end point, calculate the difference between the groove width of each layer of the master work and the taught weaving width of each layer of the mask work, and calculate the groove width of each layer of the slave work. The weaving width obtained by taking the difference between the groove width of each layer and the above-mentioned difference is the weaving width of each layer of the slave work. The trajectory control system for a welding robot according to claim 1 or 2, characterized in that the tip of the welding torch is moved while weaving.