JPH04140802A - Position control method for industrial robot - Google Patents

Abstract

PURPOSE:To alter to the interpolating method of a moving command position of a robot for realizing a smoother operation by making a rotation (a) of a second interpolation variable, and determining the resolution of a speed by the minimum time determined by a processing time of a servo-system. CONSTITUTION:A calculation period of a servo-system is set as a second interpo lation period t2, and a total second interpolation frequency N is N=>>l/vX1/t2] (l denotes a distance, and v denotes a moving speed). A calculation period of a first interpolation is set as t1, and a second interpolation frequency (a) between first interpolating positions is a=t1/t2 (a denotes an integer), and when a frequency n1 of a first interpolation is n1=N/a, n1>=2, in this case, an interpola tion xi by a first interpolation from a target position x0 and the present position xc is xi=(xo-xc)/n, and in the case of n1<2 for repeating a first interpolating calculation, a second interpolation frequency (a) between first interpolating positions becomes a first interpolation frequency n1=1, a=N and xi=xo, and a second interpolation frequency (a) between first interpolating positions becomes N and the interpolation is finished. In such a way, the frequency of a second interpolation at the time of commanding the target position xo is not a fixed number but changed to an optimal value.

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention relates to a position control method for an industrial robot that improves a method for interpolating a movement command position of an industrial robot.

(Prior Art) Generally, the commanded position of a robot is determined through teaching, and data stored in a storage device is retrieved and utilized. If the current position and this commanded position are physically distant, a route between them is provided by interpolation.

The intermediate command position determined by interpolation is determined by a straight line arc on orthogonal coordinates or by other interpolation methods. Calculation of this intermediate command position requires a processing time determined by the processing speed of the CPU, and this time is extremely long compared to the response time constant of the servo system. Therefore, if the intermediate command position obtained by this interpolation is directly commanded to the servo system, the imperfections in the command will directly affect the movement of the robot, resulting in jerky movement.

In order to avoid this problem, in the past, this command position was further interpolated and
A double interpolation method is used in which the position is commanded in detail.

In this way, the interpolation method of the robot is double interpolation, but the number of times of interpolation in the second stage has conventionally been kept constant regardless of the speed.

The number of times of interpolation n1 in the second stage depends on the processing time of CPtJ, so it is given as follows. here! is distance, ■ is speed, k is C
It is a coefficient determined by the ability of the PU, and [] represents an integer rounded up. This number of times n is the distance! and speed V
It changes depending on.

On the other hand, the number of interpolations n2 in the second stage has conventionally been set to a constant, such as n2=b (b is a constant). This means that k is C
This is a constant related to the processing time of the PU, and the time constant of the servo system has already been determined at the time of system design, so the constant is sufficient.

(Problem to be solved by the invention) The fact that the second interpolation number is a constant means that the speed resolution is C
This means that it is determined by k, which is determined by the processing time of the PU. For example, j! = 100wn,
When v = 1000 mm/s and k = 10, ■ 101000 (/5) = 1 (times), and one first interpolation, that is, no interpolation at all,
The target position will be commanded.

Next, if the speed is reduced to half V = 50 (Igm/s), n = 2, and the command will be issued twice.

Here, if v −700ms/s, n+ =
(1, 42) = 2, so the command was given twice, and in the end it was 500mm/
It can be seen that the command can only be given at a speed of s.

On the other hand, even if ■-1000IIIl/s remains, f=5
If it is set to 0mm, n+ - [0,5'l = 1, and the command should be made once. Originally, the time should be shortened by the shorter distance, but since this is not possible, the speed is 1000a+m/s, which is the desired command. half 500ms
/s.

As mentioned earlier, the velocity resolution is determined by f. For example, k=l0. If j2=100rm, the resolution is 1000mm/s, and if k=10, then 42
=50aa++, the resolution will be 2000+gaus/S. This is determined regardless of the number of times of second interpolation.

An object of the present invention is to provide a method for interpolating the movement command position of a robot that realizes smoother movement by making the number of times a of the second interpolation variable and determining the speed resolution by the minimum time determined by the processing time of the servo system. An object of the present invention is to provide an improved position control method for an industrial robot.

(Means for Solving the Problem) Therefore, the present invention has solved the problems of the prior art described above by providing the method described in the claims.

(Embodiment) Next, an explanation will be given with reference to the drawings showing an interpolation calculation flowchart of a position control method for an industrial robot according to an embodiment of the present invention.

In block 1, the calculation cycle of the servo system is set as the second interpolation cycle, and this is defined as t, and the total number of second interpolations N is first determined.

here,! is the distance, and ■ is the moving speed.

Also, [ ] is a function that rounds up the decimal point.

Next, in block 2, the calculation period of the first interpolation is set as tl, and the period ratio with the second interpolation period tl, that is, the number of times a of second interpolation between the first interpolation positions is calculated as a − t + / t * (a is an integer )...(
2), the number of times of first interpolation, n, is determined by n,=N/a (3).

Then, in block 3, the first interpolation number n, is compared with 2, and if n≧2, the process moves to block 4, where the target position is set to Xo, the current position is set to xe, and the first interpolation is performed. Find the position X.

nl In block 5, the number of partial second interpolations of the second interpolation performed between the first interpolation positions is set as a, and N=N-a (6) nl-n
l 1 ...(7)X c =
X i ...(8), then return to block 3 and repeat from the first interpolation calculation formula (4). However, if n, <2, proceed to block 6 and calculate the target position using formula (5). The command has been completed, so it will end.

If nl <2 in block 3, then block 6X;=
Xo...(5), the number of times a of second interpolation between the first interpolation positions is set to N, and the process ends.

In this way, the number of times of second interpolation when commanding the target position X0 is not a fixed number (it is changed to an optimal value).

In the flowchart of FIG. 1, nl-2, which is the condition for terminating interpolation, is not the point of the present invention, but if n<1, the time for the number of times of second interpolation is sufficient to perform the first interpolation calculation. If you have to wait for the calculation to finish before issuing a command, there is a risk that the smoothness of the movement will be lost.

(Effects of the Invention) The present invention makes the number of times a of the second interpolation variable, determines the speed resolution by the minimum time determined by the processing time of the servo system, and determines the movement command position of the robot to achieve smoother movement. This is a position control method for industrial robots that improves the interpolation method.

For example, in the drawings, k=10. v=41゜ram/
If s, l = 100mm, 1+ = 100
as.

When Ts=lOms and n1=2.43 instead of nl-3, conventionally, a=10 and the first interpolation number n is set.

is 3, which is 3 times, and the total number of second interpolations N is 3a-3
0 times. Therefore, in the past, the actual speed was 410 mm/s.
However, the present invention has the following advantages:
The aim is to bring it closer to 410 yam/S. That is, in the above example, it is a-10, so if n+=2.43, the first time is a=10, and the second time is (transition from block 5 to block 6 due to n, <2). ), by setting a=15, the total number of second interpolations is N=25. Compared to the conventional 30 times, the speed is 400n+m/s, which is closer to the more accurate 410mm/s. Therefore, in the conventional interpolation method in which the number of second interpolations is fixed, the time resolution is determined by the first interpolation period t, which causes a large difference between the desired speed to be commanded and the speed actually commanded by interpolation.
At a high speed of more than m/s, 100I! When teaching with about ll11 pinch, in ts-o,i (k=10),
The speed resolution is 500 mm/s or more, and there is a problem in that the speed can fluctuate by more than double due to a small error, and the operator needs to consider whether to lower the speed or increase the pitch.

On the other hand, by making the number of interpolations of the second interpolation variable according to the present invention, the time resolution is determined by the second interpolation period t8, so a resolution that is 10 to 20 times better than that of the conventional method can be obtained. , it becomes possible to make the robot's movements smoother. In order to substantially achieve the high speed required for modern robot movements, simply improving actuator performance is insufficient, and the interpolation technology of the present invention is an important key. Robots cannot be made faster.

[Brief explanation of drawings]

The drawing shows an interpolation calculation flowchart of a position control method for an industrial robot according to an embodiment of the present invention.

Claims (1)

[Claims] A method of interpolating a command position of an industrial robot, wherein a first interpolation determined by an interpolation calculation cycle is performed, and a second interpolation determined by a calculation cycle of a servo system is performed until a target position given by the first interpolation is performed. An industrial robot characterized in that in a two-stage interpolation method that performs interpolation, the calculation cycle of the servo system is set as a second interpolation cycle, thereby changing the second interpolation number as necessary. position control method.