JPS62231309A - Control method for numerically control robot - Google Patents

Abstract

PURPOSE:To execute the highly accurate numerical control to fetch a deformation into a robot by executing the control by using a means to execute the coordinate conversion in accordance with the mechanical deformation quantity of the robot itself actually measured beforehand. CONSTITUTION:A coordinate converting device 22 connected with a microcomputer 21 built in a control device 9 executes mutually the coordinate conversion between an absolute coordinate system N and a real action coordinate system R of the mechanical part of a robot based upon the design data. In the control at the time of the conversion from an N system to an R system, the mechanical deformation quantity from the design data of the robot is obtained by measuring actually beforehand and set to the device 22. The device 22 converts a command value concerning the position of a laser torch T to the command value of the R system out of inputted numerical control data, rotates motors M1-M5, irradiates the laser from the torch T to a work W and cuts along a desired channel G.

Description

[Detailed Description of the Invention] (Industrial Application Field) The present invention relates to a method of controlling a numerically controlled robot 1~, and in particular, to performing numerical control that incorporates mechanical deformation of the robot 1~ itself. Regarding the control method that can be used.

(Conventional technology and its problems) As is well known, there are two types of control methods for robots: numerical control method and playback method.
The robot is controlled based on data created using CAD or the like. In this numerical control method, it is assumed that the mechanical part of the robot has the shape and operating mechanism as designed, and the trajectory of the end effector and the like are given as numerical data based on these design data.

However, in reality, the mechanical parts of robots do not always have the shape or operation function as designed, and it is difficult to manufacture and operate each part.
Various deformations often occur due to assembly errors, aging, and deflection due to its own weight. Furthermore, when a heavy object (workpiece) is gripped and moved by a hand, such as in a transport robot, the deformation becomes even greater because the effects of these loads are added in addition to the object's own weight.

However, conventional numerically controlled robots have been controlled without taking such deformation into account, and therefore there has been a problem in that there is a limit to the control accuracy.

(Object of the Invention) The present invention is intended to overcome the above-mentioned problems in the prior art, and is capable of performing high-precision numerical control that incorporates the mechanical deformation of white meat. The purpose is to provide a control method for a control robot.

(Means for Achieving the Object) In order to achieve the above-mentioned object, in the present invention, when numerically controlling a robot manufactured based on predetermined design data, first, the design data of the mechanical part of this robot is used. Measure the deformation of

Based on the results of this actual measurement, we created (1) a first coordinate system for describing the movements of the robot 1~ assumed in the above design data, and (2) a first coordinate system for describing the actual movements of the robot in which the deformation occurred. The relationship with the coordinate system of 2 is determined in advance.

Based on the relationship obtained in this way, a coordinate conversion means is provided for converting coordinates from at least one of the first and second coordinate systems to the other, and by this coordinate conversion means, the robot The control data of this robot is converted to control this robot or the robot of Haku.

(Example) A. Overview of Mechanical Configuration of Example FIG. 3 is a schematic diagram showing the mechanical configuration of a Cartesian coordinate laser cutting robot as an example of robots 1 to 1 to which the present invention is applied and controlled. FIG.

In the figure, this laser cutting robot RB has a base 1
It has a moving table 2 which is movable in the X direction (approximately horizontal direction) by a motor M1 (not shown) on the top thereof, and a workpiece (not shown) is placed on this moving table 2. A beam 4 is installed on the top of a column 3 that stands vertically on both sides of the base 1.
A moving column 5 is provided which extends in the Y direction and is movable in the Y direction by a motor M2.

A motor M4 is provided at the lower end of the moving column 5 and is moved up and down in two directions by a motor M3.

As a result, the arm 6 provided eccentrically from the central axis of the moving column 5 rotates in the θ direction in the figure. Further, a motor M5 is provided on the side of the lower end of this arm 6, and this rotates the laser torch T in the ψ direction in the figure.

Of these, a laser beam from a laser oscillation device 7 is applied to the laser torch T through a laser guide pipe 8. In addition, the control device 9 has a built-in microcomputer, a coordinate conversion device, etc., which will be described later, and an operation panel 1.
0 is provided with a keyboard, a display, etc. Roughly speaking, the external computer 11 is for performing CAD input and the like.

The x, y, and z directions mentioned above may differ from the horizontal direction It is assumed that there is a slight deviation in the Y direction. Also, these (x,
It is assumed that displacement also occurs between respective coordinate origins (not shown) formed with coordinate axes of (y, z) and (X, Y, Z). Therefore, (X, Y.

An absolute coordinate system N (first coordinate system) formed by (Z) and (
This is different from the operating coordinate system R (second coordinate system) formed by x, y, and zL. Among these, the absolute coordinate system N (X, Y, Z) is a rectangular coordinate system, but the motion coordinate system R (X, Y, Z) corresponding to the actual motion of the robot RB is
l/, Z) is assumed to be an oblique coordinate system due to the deformation of the mechanical part.

B. Schematic diagram of the electrical configuration of the embodiment FIG. 4 is a schematic diagram of the electrical configuration of the bottle RB shown in FIG. 3. In FIG. 4, the following devices are connected to a microcomputer 21 built into the control device 9 via a bus BL.

■The above motors M to M5 and these motors M1 to M5
Mechanism drive system 23 including encoders E1 to E5 (not shown in FIG. 3) that detect the rotation angle of The coordinate conversion device 22 converts the absolute coordinate system N (X, Y, Z) and the motion coordinate system R (x
, y, z), and in this embodiment, the CPU (not shown)
It consists of a microcomputer that includes a computer and memory.

It is assumed that this robot RB cuts the workpiece W shown in FIG. 4 along a trajectory G.

C6 Coordinate Conversion Formula Next, in this example, the relationship between the absolute coordinate system N (X, Y, Z) based on the design data and the actual operating coordinate system R (x, y, z) of the mechanical part of the robot RB is calculated. Calculate the coordinate transformation formula. According to the present invention, a coordinate conversion means is provided that performs at least one of the conversion from an absolute coordinate system (hereinafter also referred to as "N system") to a motion coordinate system (CrR system) and its inverse transformation. However, in this example, the fourth
The illustrated coordinate transformation device 22 is configured to be able to perform both of these transformations. Therefore, in the following, both conversions will be calculated.

(C-1) Shifting from N system to R system First, with reference to FIG. 5, a conversion formula from N system (X, Y, Z) to R system (x, y, z) is derived. That is, in the N system, three-dimensional coordinates: (N1.N2.N3) ... (1
) is 1 2 in the R system.
...(2) (R, R, R3) R-R3 is expressed by N1 to N3.

Here, the R system is not necessarily a rectangular coordinate system, but in a rectangular coordinate robot, X, Y, Z (and x, y,
z) is independent, and deformation of a particular axis does not affect the movement of other axes. For example, for coordinate R3, as shown in Figure 6(a), the point It can be defined as the distance between the origin O1 and the point F3, where a plane π passing through P and parallel to the x-y plane intersects the Z axis, with a sign given. Regarding R and R1, as shown in FIGS. 6(b) and (C), the plane π31. It can be similarly defined using π12 and the intersection points F 1 and F 2 of each axis.

Therefore, first, the direction cosine that each coordinate axis x, y, z of the R system has with respect to each coordinate axis of the N system is expressed in vector form, → cl = (j!,, m-, n,), i=1 ．．
2.3・−(3). Here, i = 1.2.3 are the X axis and y axis, respectively.
The axis and Z axis are shown. Also, the coordinates of the origin OR of the R system derived from the origin ON of the N system are o = <○Pi" R2,0R3)...
(4). As described later, these are determined in advance through actual measurements.

Therefore, the equation for the i-axis (i = 1.2.3) of the R system is such that the i-axis passes through the origin OR and has a direction cosine expressed by equation (1) for each coordinate axis of the N system. Therefore, (X-〇)/A・= (Y-OR2)/m.

It can be expressed as R1+ = (Z-OR3> /n i (5). Then, if the value of this equation is set as the parameter U, then equation (5) becomes x-1, u, +OR1 ...(6) I Y = m, u ・+OR2-(7) 7-n, u, +
OR3...(8) I yell.

Next, if i, j, and k are respectively (1, 2, 3) cyclic permutation, then the plane πj parallel to the axis of j− and the one-culm equation are S・, tl
With l as a parameter, X-○ -5-1・+ti'k...(9)
11J Y-○R2R2-8i+timk...(10)
It can be written as z−oR3=s, nj+t, nk ・<11).

When the parameters S. and ti are deleted from equations (9) and (10), the equation of the plane πjk becomes: A.X+Bi Y+Ci Z11-〇...(12)■. However, → α, = (A,, B,, C,) ... (15)
It is.

As mentioned above, R1 is found based on the intersection point of the plane πjk and the i-axis, so (6) to (8)
Substituting the formula into (12), we get A・(1・U・+0R1) + B・(mu・+OR2) +C−(n・U,+0R3) +D・−〇...(16) However, by solving this for ui, U・=
-(R1・chi+D;)/immersion 1・gi; (17).

Also, according to equations (3), (13), and (15), if we set the value of equation (18) as β, then equation (17) becomes u・−(α・・OR+D,) /β, -(19).

Then, when the values of equation (19) are substituted into equations (6) to (8), the values obtained as X, Y, and Z are point F.

Since it is the coordinate value in the N system of R・-sgn (N・-0RH) −1FB −OR
1=san (N ・011) xl (j!・tJ-, m1LJ, jJLJ) 1=s
gn (N--01B> Xlu, l (j2. +m, +n, 2) 172
-sgn (N・-0H) ×1α・・oR−1−[), l, , /β1・・・・(20
) tona2ru. However, since the N system is a rectangular coordinate system, (1. 10 m. +n.) = 1...(21)
+ 1 1 was used.

This equation (20) is the conversion equation from the N system to the R system.
(14) contains coordinate information N of the point P in the N system.

(C-2) For the conversion method from R system to N system, with reference to Figure 7, in R system R=(R1°R, R3
) of the point represented by the coordinate vector N in the N system
=N, N2. Find N3).

To do this, it is sufficient to find the intersection of a plane πjH (i = 1.2.3) that passes through the point Fi and is parallel to the j- plane in the R system. Among these, the ■ coordinates (Xi, 'Yi, 'Zi) of point F in the N system are 'X1-Ri
'i +○R1...(22) L4H=Rim, +OR
2. (23) LZ, = R, n, +OR3.(2
4). Therefore, the equation of the plane πjk is (12
), A−X1B, Y0C・Z+D・=0 (25). Among these, A· to C1 are the same as those defined in equation (11). In addition, Di has the same value as , in equation (12), but it is not expressed using N as in equation (14), but in equations (22) to (24) above.
Vector defined from Lx1 to Zi in the formula; → Li = (LXi, 'Yi, '2i)...
According to (26), it is expressed as .

Therefore, the coordinates of point P in N system (N1°N2,
N3) are respectively given by X, Y, and Z obtained by simultaneous equation (25) for i = 1.2.3,
The result is: However, γ is given by.

These equations (28) to (31) are the conversion equations from the R system to the N system, and as seen from the equations (27) and (22) to (24), D(i-1,2 , 3) are expressed by the coordinate values R·(i=1.2.3) of two points in the R system.

Control including coordinate transformation from D and N systems to R system With preparations above, first control including coordinate transformation from N system to R system
This will be explained with reference to the flowchart shown in the figure.

First, in step S1 of FIG. 1, numerical control data such as a desired fusing path is created based on the coordinate values of the N system. This is done by using the external computer 1 of FIG. 3 as a CAD.

Next, after test/correction operations (step S2), when a reproduction operation is started (step S3), the numerical control data stored in the memory of the external computer 1 is read out, and the coordinates shown in FIG. transferred to the conversion device 22 (
Step S4).

On the other hand, the amount of mechanical deformation of this robot RB from the design data is determined by actual measurements in advance, and these amounts of deformation are determined by the above-mentioned directional cosine (Equation (3)) and deviation of the origin (Equation (4) Specific values of the equation) are stored in this coordinate transformation bag 'e122. This actual measurement may be performed, for example, by driving each of the motors M1 to M5 individually and determining the deviation of the locus of the torch T, or by directly determining the relative angle of each member. It may also be done by

Therefore, the coordinate conversion device 22 converts the command value regarding the position of the laser torch T out of the input numerical control data into an R-system command value using equation (20) (step S5). Then, according to the command value of this R system, the motor M
- Rotate M5 (step S6) and return from step S7 to step S4 until the final command value is reached.
Move on to reading the next numerical control data. In these processes, although not shown, the workpiece W is irradiated with a laser beam from the laser torch T, thereby cutting the workpiece along a desired path.

In these operations, the deviations (so-called "taore") due to the mechanical deformation of the robot RB are corrected by the coordinate conversion device 22, so if numerical control data is created in the absolute coordinate system (N system), the The fusing is performed along a desired path. However, strictly speaking, the attitude of the laser torch T also shifts due to the above deformation, but the effect of the attitude shift is quite small compared to the position shift, so there is no need to perform any coordinate transformation regarding the attitude. .

A first example of control using conversion from the R system to the N system in a control number including conversion from the E and R systems to the N system will be described with reference to the flowchart shown in FIG. 2A. This control is a process used in the test/correction operation (step S2) in FIG. 1 and the like.

First, in step S10, numerical control data to be input to the N system is created using CAD or the like in the same manner as described above. Then, transfer this N-system data to the coordinate conversion device 22,
This coordinate inspection device 22 converts this N system data into R system data using equation (20) (step 511
) Based on this, the motor is driven. Then, based on the determination in step S13, step 311. S
12 is repeated until the final command value is reached, and once the final command value is reached, the operation is temporarily stopped (step 514).

These operations are monitored by the operator,
If there is any inconvenience in this operation, the next step 815
Then, the operator enters the operation panel 10 and switches the control system to the teaching mode. Then, manually operate the relay amount A] - -ch T for the inconvenient part, and transfer the movement data at this time to the encoders E to [5 and J.
: is taken into the control system (step 816). Since this data is based on the actual movement of the laser torch T, the encoded signal from the encoder E-E5 is corrected data in the R system.

Therefore, this R-system data is coordinate-transformed into N-system data in the coordinate transformation device 22 based on equations (28) to (31) (step 517), and the N-system data thus obtained is The original numerical control data (N-system data) is corrected by (step 818). Therefore, if the reproducing operation shown in FIG. 1 is executed based on the data corrected in this way, more accurate numerical control will be performed.

FIG. 2B is a flowchart showing a second control example including conversion from R system to N system. The difference between this process and the process shown in Fig. 2A is that instead of correcting numerical control data created by CAD, etc., numerical control data is directly created based on teaching, and based on that, the robot or the robot in question is numerically adjusted. It's in the point of control. That is, the second
In step 831 of Figure B, the robot RB is set to teaching mode, and teaching is performed in step S32. This teaching data is converted from the R system to the N system in the coordinate conversion device 22 (step 533), and is stored in the memory in the external computer 11 (step 5).
34).

Since the data obtained in this way is N-system data, it is data that does not depend on the amount of deformation of the robot. Therefore, this data or data modified as appropriate can be used as numerical control data for other types of numerical control in the robot concerned or for numerical control of other robots (steps 835 and 836). ).

In other words, by performing coordinate transformation through the coordinate transformation device 22, it is possible to freely transform data that takes into account the amount of deformation unique to each robot and numerical control data that is common to each robot. It is.

F. Modifications Although the embodiments of the present invention have been described above, the present invention is not limited to the above-described embodiments, and for example, the following modifications are possible.

■ In the above embodiment, the structure was configured to have a mutual conversion function between the N system and the R system. If only the conversion function to the system is provided, in the playback operation,
Furthermore, if only the reverse conversion function is provided, desired effects can be obtained in the numerical control data creation process. Also, the coordinate transformation is the above J: sea urchin rough 1
It is not necessary to perform this separately, and it may be configured as a dedicated hardware circuit.

(2) In the above embodiment, a Cartesian coordinate robot was used as an example, but the present invention can also be applied to a Cartesian coordinate robot (for example, a cylindrical coordinate robot). In an articulated robot or the like, the coordinate transformation means is complicated to some extent, but if, for example, the amount of deformation of only the arm near the end effector is to be taken into account, it is sufficient to perform coordinate transformation to the same extent as in the above embodiment.

■ When incorporating deformation due to the load of a transport robot, etc., rather than deformation due to the robot's manufacturing process or its own weight, the amount of deformation will differ depending on the weight of the work being handled. In this case, a plurality of types of deformation amounts may be actually measured using the weight of the work to be handled as a parameter, and during actual playback, the weight may be specified or measured and coordinate transformation may be performed accordingly.

(Effects of the Invention) As explained above, according to the present invention, control is performed using a means for performing coordinate transformation according to the amount of mechanical deformation of the robot itself, which is actually measured in advance. Highly accurate numerical control can be applied to the robot.

[Brief explanation of drawings]

FIG. 1 is a flowchart showing a control operation including coordinate transformation from an N system to an R system in an embodiment of the present invention, and FIGS. 2A and 2B show a control operation including a coordinate transformation from an R system to an N system. FIG. 3 is a schematic perspective view showing an example of a robot suitable for implementing the embodiment of the present invention; FIG. 4 is an electrical configuration of the robot shown in FIG. 3; 5 and 6 are explanatory diagrams of the coordinate transformation formula from the N system to the R system. FIG. 7 is an explanatory diagram of the coordinate transformation formula from the R system to the N system. RB...Laser cutting robot, T...Laser torch, 9...Control device, 11
...External computer (CAD), 22...Coordinate conversion device, N...Absolute coordinate system, R...Movement coordinate system

Claims (2)

[Claims] (1) A control method for numerically controlling a robot manufactured based on predetermined design data, the method comprising actually measuring deformation of a mechanical part of the robot from the design data, and based on the result of the actual measurement, Determining in advance the relationship between a first coordinate system for describing the motion of the robot assumed in the design data and a second coordinate system for describing the actual motion of the robot in which the deformation has occurred, and determining the relationship. A coordinate transformation means is provided for performing coordinate transformation from at least one of the first and second coordinate systems to the other based on the coordinate transformation means, and the coordinate transformation means coordinately transforms the control data of the robot so as to transform the control data of the robot or the robot. A method for controlling a numerically controlled robot characterized by numerically controlling another robot. (2) The robot is a Cartesian coordinate robot, and the relationship between the first and second coordinate systems is determined by an amount depending on the direction cosine between each coordinate axis and an amount depending on the deviation of the origin. A method for controlling a numerically controlled robot according to claim 1, which is obtained based on the following.