JPS62174804A - Learning control method for industrial robot - Google Patents

Abstract

PURPOSE:To reduce the number of times correcting a teaching value and to improve learning efficiency by adding the actual delay time of a controlled system and correcting a teaching value. CONSTITUTION:An industrial robot controller is provided with a target value memory 1, a teaching value memory 2, a reproduction value memory 4 storing information on arm position with respect to the target value detected by a position detector 3 as a reproduction value, and an arithmetic unit 5 which obtains the deviation of the target value from the reproduction value stored in the reproduction value memory 4, corrects contents (teaching value) in the teaching value memory 2 according to the deviation, reads contents stored in the target value memory 1 or the teaching value memory 2 and operates an arm through an actuator 7. For correcting the teaching value, the delay time of the arm is measured, and based on the result the teaching value in the teaching value memory 2 is corrected according to the deviation of the reproduction value allowing the delay time of the arm from the target value. In that case, an operator stores position information the same as the contents in the target value memory 1 in the teaching value memory 2 by teaching work.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a learning control method for industrial robots, repetitive positioning devices, etc.

[Prior Art] Conventionally, a controlled object is regenerated according to a taught value stored in a memory, and a deviation between a target value and a reproducing trajectory is measured during the reproducing operation, and this deviation is used as the teaching value during the next reproducing operation. A learning control method for an industrial robot is known in which the teaching value for reproducing the controlled object is sequentially corrected by adding the teaching value to the value, thereby improving the accuracy of the trajectory of the reproducing operation.

According to this learning control method, the taught value is corrected to the 1st degree according to the deviation between the target value and the actual playback trajectory, so as the playback operation is repeated several times, the taught value changes. The target value converges within the allowable error range.Therefore, the work locus of the reproducing operation can be controlled in accordance with the target value.

[Problem to be Solved by the Invention 1] However, in general, an arm or the like to be controlled in an industrial robot or the like has a certain delay time from when a target value is given to when it actually starts operating. Therefore, in the method described above, in which the teaching value to be used next time is corrected based on the deviation of the playback trajectory from the teaching value given at the current time, the teaching value is corrected by the deviation from the previous time by the delay time. Therefore, it is necessary to repeat many reproduction operations until the taught value converges within the allowable error range of the target value, resulting in a problem of poor learning efficiency.

The present invention has been made to solve these problems, and aims to provide a 6D learning method for industrial robots that can improve learning efficiency.

[Means for solving the problem 2] The present invention measures the delay time from when a teaching value is given until the controlled object starts operating, and reproduces the playback trajectory at a time delayed by a time corresponding to this delay time. The deviation between this value and the target value is determined, and the taught value for the next reproduction operation is corrected based on this deviation.

[Embodiment] FIG. 1 is a block diagram showing an embodiment of an industrial robot control device to which the present invention is applied, and the position information of each point on the motion trajectory of an arm (not shown) to be controlled is A target value memory 1 stores the target value as a target value, a taught value memory 2 stores the target value corrected by teaching correction as a taught value, and the position information of the arm relative to the target value detected by the position detector 3 is used as a reproduced value. Reproduction value memory 4
and the reproduction value stored in the reproduction value memory 4 and the target value @
a calculation device 5 which calculates the difference, corrects the content (taught value) of the taught value memory 2 based on the deviation, reads out the stored content of the target value memory 1 or the taught value memory 2, and operates the arm via the actuator 6; It is equipped with

FIG. 2 is a flowchart showing the procedure for correcting the teaching. First, in steps 10 to 15, the arm delay time is measured, and based on this measurement result, the teaching value in the teaching value memory 2 is set to the arm delay time. It is corrected based on the deviation between the playback value and the target value, taking into account.

In this case, the same position information as the contents of the target value memory 1 is initially stored in the taught value memory 2 by the teaching work by the operator.

First, in step 10, the target value at each point on the motion trajectory stored in the target value memory 1 is read out by the arithmetic unit 5, and each target value is given to the actuator 6, so that the arm moves to each target value. Correspondingly driven.

At this time, the actual operating position is detected by the position detector 3 for each target value, and is sequentially stored in the reproduction value memory 4 as a reproduction value (step 11). Once the reproduced value for the current target value is obtained, a feedback calculation is performed to add the deviation to the next target value (step 12).
, the next target value is determined by the actuator 6 based on the result of this operation.
given to. As a result, the trajectory of the arm changes in accordance with the target value stored in the target value memory 1.

When the first regeneration operation using all target values is completed in this way, the arm operation start time is detected based on the address from which the regeneration value is stored in the regeneration value memory 14 (step 14).

That is, target value memory 1. Since the memory address of the taught value memory 2 and the reproduced value memory 4 is specified by the same address signal that changes sequentially at a predetermined period, if the arm delay time is zero, the target value memory 1
The reproduced value for the target value read from is stored in the reproduced value memory 2 at the same address as the target value. However, if the arm operation start time is delayed, the data is stored at an address shifted by the amount of the delay time.

Therefore, by detecting this address deviation, the arm delay time can be detected.

If the delay time of the arm can be measured in this way, the taught value stored in the taught value memory 2 will be determined by the difference between the reproduced value and the target value at a time that is close by the arm delay time. Fixed.

In other words, with the delay time below, the target value at time t is MO(t)
, M2(t) is the reproduced value at time t, and Ml is the taught value at time t.
(t), and K is a constant, then Ml (t)' = M1
(t)+K(MO(t+T)-M2(t+T)
), the taught value M1(t) in the taught value memory is corrected, and the corrected bundle M1(t)' is used as the taught value corrected by the arm delay time.

Next, once such correction has been completed for all the taught values at each time, the corrected taught values are read out one after another, and the deviation from the reproduced value detected by the position detector 3 is determined. , a feedback calculation of adding the deviation to the taught value is performed for the taught values of all operating positions (
Steps 16-19). If the motion trajectory of the arm driven by the resulting teaching value is not within the tolerance range, the process returns to steps 16 to 19 again, the teaching value is corrected, and the teaching value is within the tolerance range. If it converges within, the teaching correction ends.

[Effects of the Invention] As is clear from the above explanation, in the present invention, the teaching value is corrected by taking into account the actual deflection time of the controlled object, so the number of teaching corrections is reduced. Learning efficiency can be improved.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing an embodiment of an industrial robot control device to which the present invention is applied, and FIG. 2 is a flowchart showing a teaching correction procedure. 1...Target value memory, 2...Taught value memory, 3...
・Position detector, 4... Reproduction value memory, 5... Arithmetic device, 6... Actuator. Applicant's agent Takahisa Kimura σ－－－－－：Su、
゛

Claims (1)

[Claims] The control target is regenerated according to the taught value stored in the memory, and the deviation between the target value and the regenerated trajectory is measured during the regenerated operation and added to the taught value during the next regenerated operation. In a learning control method for industrial robots that sequentially corrects the taught value for reproducing the controlled object, the method measures the delay time until the controlled object starts operating with respect to the taught value, and calculates this delay time. A learning control method for an industrial robot, characterized in that the deviation between a playback trajectory and a target value at a time delayed by a time corresponding to , is determined, and a taught value for the next playback operation is corrected based on this deviation.