JPH0320804A - Robot reproduction control method - Google Patents

Abstract

PURPOSE:To obtain a locus with fidelity to route fluctuation by decreasing a correction value fitting in the magnitude of the speed fluctuation and acceleration fluctuation of a robot. CONSTITUTION:A robot axis 4 can be moved by sending a command to move the robot axis 4 to a correction targeted position Pi' from a CPU 1 to an actua tor 3. The CPU 1 reads out the next teaching position Pi from a memory 2, and reads out teaching positions Pi-a...Pi-1 before that, and teaching positions Pi+1...Pi+b behind that. Next, a weight factor can be obtained by operating the mean value Pia of preceding teaching positions and the mean value Pib of succeeding teaching positions, and calculating a mean route seconds of arc thetai by utilizing the inner product of a vector connecting the position Pia to the Pi with a vector connecting the position Pi to the Pib. The CPU 1 operates the correction targeted position Pi' based on the teaching position, the preceding teaching position, the succeeding teaching position, and the weight factor. Since the correction value is decreased when a large amount of route fluctuation exists, the locus with fidelity to the route fluctuation can be obtained.

Description

[Detailed Description of the Invention] Industrial Application Fields The present invention relates to a method for regenerating a robot IIJrIg, and is designed to maintain a smooth movement path and make the movement path, especially at corners, close to the original teaching position. This is what I did.

Note that the robots referred to in the present invention include ordinary robots,
Includes manibulator, work i[, M feed iw, etc.

BACKGROUND ART FIG. 1 is a block diagram showing the main parts of a robot, for example a painting robot. This robot uses the teaching play pack method to teach work. In general, the teaching playback method has the advantage that the teacher's skills can be directly taught.

Here, the teaching reproduction operation will be explained in more detail.

When teaching, the teacher must move the robot axis 4 (moving axis)
move directly. Therefore, the teaching path L of the robot 4
is as shown in FIG. 5, for example. During this teaching, a large number of teaching positions P are sampled from the teaching path L at fixed time intervals or variable time intervals, and these teaching positions P are stored in the memory 2. Note that the reason why the teaching light @L is in the form of a polygonal line is that unintentional hand movements by the teacher may be included as teaching information.

When performing a reproducing operation, the central processing unit (CPU) 1 reads the taught position P, from the memory 2, and performs a correction operation on the taught position P, to obtain a corrected target position P,'. This correction calculation is performed to obtain a smooth correction path l from the polygonal taught path L, and is expressed by the following equation (1).

The CPUI sends operation commands to the actuator 3 to operate the robot axis 4 so that the robot axis 4 sequentially moves to each corrected target position P'.

At this time, the position information of the robot axis 4 is fed back to the CPUI for feedback control. Thus, the robot axis 4 sequentially moves toward each correction target position p, /, and its reproduction path becomes a smooth correction path l.

Problems to be Solved by the Invention> By the way, in the above-mentioned prior art, as shown in FIG.
When moving in a straight line, the trajectory of the robot axis 4 (correction 111
1! The trajectory of the robot axis 4 (
The correction path (Ril) has the disadvantage that it goes inside the original teaching position and deviates from the expected trajectory (the deviation is indicated by α in FIG. 5).

SUMMARY OF THE INVENTION In view of the above-mentioned prior art, it is an object of the present invention to provide a regeneration control method for a robot that can maintain a faithful trajectory even when the path changes at corners and the like.

Means for Solving the Problems> The present invention solves the above problems as follows: In order to find the corrected target position to which the moving axis will go next, first, the teaching position corresponding to the desired corrected target position and the required number of points before this are determined. Find the average path displacement angle formed by the line connecting the average position of the teaching positions of , the average found! ! A correction weighting function with a value corresponding to the value of the road displacement angle is determined, and the correction calculation formula is calculated using the obtained correction weighting function, a teaching position corresponding to the desired correction target position, and teaching positions before and after this teaching position. The feature is that the desired corrected target position is obtained by substituting .

Function> The magnitude of the path fluctuation is determined from the average path displacement angle, and when the path fluctuation is small, such as when moving in a straight line, the correction is increased to obtain a smooth trajectory, and when the path fluctuation is small, such as when moving in a corner, the correction is increased to obtain a smooth trajectory. When the route variation is large, the correction is made smaller to obtain a trajectory that is faithful to the route variation.

Embodiments FIG. 1 shows the main parts of a robot (painting robot, etc.) to which the method of the present invention is applied, and FIG. 2 is a flow diagram showing the main parts of the method of the present invention.

The teaching operation is similar to the conventional one, in which the teacher directly moves the robot axis (moving axis) 4, samples the teaching position P from the teaching line RiL (see Figure 4), and stores a large number of sampled teaching positions P1 in the memory 2. Remember. In the memory 2, the initial value of the weighting function K, and the numbers of reference points a and b are set. Note that the initial value of the weighting function K and the reference points a and b may not be stored in the memory 2, but may be readable by the central processing unit (CPU) 1 from a setting device such as a preset switch.

On the other hand, when performing a reproducing operation, the CPU performs the following processing and then obtains each corrected target position P.

Here, the CPU executed before finding the corrected target position P'
The processing operation of I will be explained with reference to FIGS. 2 and 3.

■ When reproducing the robot, the CPU 1 reads the next teaching position P to be reproduced from the memory 2 (step i) o ■ The CPU 1 reads the teaching position pl-m before the teaching position P
'...p, -,, p. , as well as teaching position @P, later teaching position teaching position II P141'P1 and 2,...
- Read P1+b (step 2).

■ Next, the previous teaching position @ P, -,,...Pl-2j
Average position of Pl-1 Pl. and the later teaching position fiI P
++t'P l+1'... Average position of Pl+k+ [P,
Calculate ゎ using the following formula (21 (31) (step 3
).

Σ P P =゜"1-゜...
(2) 1 a Σ PP = 21211 ... (3)
IIb ■ Calculate the average path displacement angle θ using the following equation (4) using the inner product relationship of the vector P,,P, connecting positions P1 and P, and the vector PIPlb connecting positions P, and Pll. (Step 4).

(2) Substitute the calculated average path displacement angle θ, into the following equation (5) to calculate the weighting function KJ used in the correction calculation equation (1).

K=f(θ) (5) j f(θ) is a function that determines K based on θ1, and is set depending on the object to be controlled (step 5).

■ The CPUI is at the teaching position p. , previous position P ,...
P+-2, P+-1, later position p: ++'P,. 2,・
...P,+1,, the calculated weighting function K is corrected using the calculation formula (1
) to calculate the corrected target position PI' (step 6) o Σ K, After performing the processing of ■ to ■ above, the CPU directs the robot axis toward the corrected target position P, Sends a movement command to the actuator 3 so that the actuator 4 moves (
Step 7).

In case B, the movement of the robot axis 4 is controlled by feedback control. Therefore, the robot axis 4 moves smoothly along the correction path l shown in FIG. Moreover, as mentioned above
Since the processing described in (2) is performed, when the path variation is small, such as when moving in a straight line, the correction is increased to obtain a smooth trajectory, and when moving around a corner, as in the case of moving in a corner, 1! When road fluctuations are large, the correction is made small to obtain a trajectory that is faithful to the route fluctuations. In other words, as shown in FIG. 4, the correction path t, which is the locus of the robot axis 4, is smooth and is close to the original taught position even at the corners.

Therefore, the deviation α shown in FIG. 5 is eliminated, and the fidelity of the reproduced trajectory is improved. For example, in a painting robot, there is no unpainted area, and the quality of painting is improved.

Effects of the Invention> As specifically explained above in conjunction with the embodiments, according to the present invention, when calculating the corrected target position during playback, by incorporating the path fluctuations before and after the correction target position, the path fluctuations are reduced at the corner portion. It is possible to reduce the correction according to the size of the object and move the movement axis faithfully and smoothly along the taught path, and in the linear movement part, it is possible to increase the amount of correction and move the movement axis even more smoothly. In addition, since smooth movement is possible, position speed is also averaged, speed fluctuations and acceleration fluctuations are reduced, and machine life is improved.

[Brief explanation of the drawing]

Fig. 1 is a block diagram showing the main parts of the robot, Fig. 2 is a flow chart showing the calculation procedure of the method of the present invention, Fig. 3 is an explanatory section showing the method of calculating the average path displacement angle, and Fig. 4 is the method of the present invention. FIG. 5 is an explanatory diagram showing a route obtained by the conventional method. In the drawing, 1 is the central processing unit, 2 is the memory, 3 is the actuator, 4 is the robot axis) P is the teaching position, P' is the correction target position, L is the teaching path, I is the correction path, K is the weighting function, θ is the average path displacement angle.

Claims (1)

[Claims] A large number of teaching positions are sampled and stored from a teaching path that is the path of the robot's movement axis during teaching, and when the movement axis is regenerated, the playback path is smoothed. In a robot regeneration control method in which each taught position is corrected using a correction calculation formula to obtain a corrected target position corresponding to each taught position, and the movement axis is sequentially moved toward each corrected target position. To find the corrected target position to which the axis will go next, first, draw a line connecting the taught position corresponding to the desired corrected target position and the average position of the required number of previous taught positions, and the line that corresponds to the desired corrected target position. The average path displacement angle formed by the line connecting the teaching position and the average position of the required number of subsequent teaching positions is calculated, and then a correction weighting function of a value corresponding to the value of the calculated average path displacement angle is calculated. and substituting the obtained correction weighting function, a taught position corresponding to the desired corrected target position, and taught positions before and after this taught position into the correction calculation formula to obtain the desired corrected target position. A method for controlling playback of robots.