JPS59114604A - Acceleration and deceleration controlling system of industrial robot - Google Patents

Abstract

PURPOSE:To change smoothly a speed by executing a control by an operation of a pulse moving extent which moves between two point of a teaching point train, and an operation of a weighted moving mean of these continuous time series data of a prescribed number. CONSTITUTION:A pulse moving extent, etc. of X-Z axes are operated through a PTP operating circuit 2, etc. basing on a data of a pulse number, etc. between teaching points stored in a teaching point data memory 1, by an industrial robot of an orthogonal coordinate type, and an acceleration and deceleration control of digital servo-systems 6-8 of each axis is executed. As for this operation, the pulse moving extent is operated from the number of clocks and the number of increment pulses stored in the memory 1 by a PTP operating circuit 2, and also a weighted moving mean of the pulse moving extent of each axis is operated 3-5. This mean is provided as a pulse moving extent at a time tn, to the servo-systems 6-8 of each axis.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention provides discrete teaching point data,
The present invention relates to an acceleration/deceleration control method for industrial robots in which the pulse movement amount of each axis is calculated at regular intervals from this teaching point data.

Conventionally, no-acceleration/deceleration control of industrial robots is based on P↑P.
Intermittent trajectory control in this system, that is, stopping at each teaching point, is performed only for positioning, for example, a spot welding robot. Figure 1 generally shows the teaching point sequence, and shows the teaching point sequence at times tn-1, tn, tn.
+1, teaching points Pn-1 and Pn, respectively
, Pn+1.

FIG. 2 is an example of the above-mentioned acceleration/deceleration control for a Cartesian coordinate type industrial robot, and shows each X and Y of a certain axis at each time.

When it approaches the vicinity of the teaching point pn, it is linearly decelerated, and at time Mtn, the pulse pulse movement amount becomes zero and stops at the teaching point Pr+. Then, it is linearly accelerated from time tn', reaches a constant speed at a certain time, and moves toward the next teaching point pn+1. In the vicinity of teaching Pn+1, linear acceleration/deceleration control is similarly performed. Since industrial robots generally have multiple axes, the acceleration/deceleration control described above is complicated. Even if the speeds of the respective axes at constant speed are different, acceleration/deceleration control is performed so that the speed patterns of all the axes are similar, that is, have a constant proportional relationship. Even in this simplest linear acceleration/deceleration control, calculation of pulse movement amount at constant speed, calculation of first-order difference pulse movement amount during acceleration/deceleration, acceleration → constant speed 1. A switching/switching determination calculation and a constant speed → deceleration switching determination calculation were required.

In contrast to the intermittent trajectory control in the PTP method described above, there is continuous trajectory control in the PTP method in which the speed changes without stopping at each teaching point.

An example of an industrial robot that performs this continuous trajectory control is an arc welding robot. In such industrial robots, speed changes occur on each axis at each teaching point, and these speed changes are completely independent of each axis, with some axes accelerating, some axes decelerating, and some axes having no speed change. Since the results are disjointed, the linear acceleration/deceleration control described above is performed.
However, it is virtually impossible to complete pulse distribution for all axes at the same time for each teaching point online due to limitations in calculation speed. Figure 3(a)(b)
(c) shows an example of a temporal change in the amount of movement of the gururus with respect to the X, Y, and Z axes of such a continuous trajectory control industrial robot. As is clear from the figure, the velocity change before and after the teaching point was different from the case of the intermittent orbit control shown in FIG. 2, and was performed in steps, causing vibrations.

That is, in a multi-axis industrial robot that performs continuous trajectory control, a method for smoothly controlling speed changes at the teaching point has remained unresolved.

The present invention was proposed in view of the above-mentioned problems, and is applicable not only to multi-axis industrial robots that perform continuous trajectory control, but also to multi-axis industrial robots that perform intermittent trajectory control, and can smoothly change speed. The purpose of this invention is to provide an acceleration/deceleration control method for industrial robots.

The acceleration/deceleration control method for the industrial robot of the present invention is as follows: (1) For each axis, when moving between two adjacent points in the teaching point sequence at a constant speed, N, pulse movement amount 5n-j (j=N-1, N-2--1
, O), and performs acceleration/deceleration control using this weighted moving average Sn as the pulse movement amount at time tn.

The above principle of the present invention will be specifically explained. Figure 7(a)
moves between teaching points at a constant speed with a pulse movement amount S=2, reaches the teaching point at time t3, and thereafter,
The method of the present invention is applied to an axis that moves at a constant speed with a pulse movement amount of 5-12 until the next teaching point. In this example, N=4Wo=0. 'l, Wl = 0
．． 3, W2 =02, Wl =θ/.

Below, each time t3 y t4 v t5 r t6 p
S3 outside the pulse movement amount at t7. s4. s5. s6
．． S7 weighted (i) when t=t3 S3 = S2
=S1 =s□ =2 Thea 6 Kara S3 =
2 (ii) t = t, when S4=/j, 53
=S2=3□-2 and Luka 34 =O,ll×/2
+0.3x2+0.2x2+0. /X2=Otsu(iii)t=t5(7) Toki55=S4=/2,
53=S2=2”?: Because there is s5=θl×72+0
3x/2+0.2x2+0. /×2=9 (IV) When t=j6(7) 56-85-84-12
, 53-2, so 56=O,ll×/2+0
3×/2+0.2X/2〇, lX2=//.

(v) When t=t7 57=S6=S5"'S4-
/j while S7 =O,'l×/
2+Q, 3×/2+'0.2X/2+0. /×/2=I
It becomes 2. If this is illustrated, the curve will be in a step shape as shown in Figure (a), and since the period is small, it will be a curve similar to the approximate curve diagram. Fig. 5(b) shows the case of an axis where the pulse movement amount does not change between teaching points, and the weighted moving average of the pulse movement amount is, of course, the same as the actual %/res movement amount.

w6=o, os, same figure (b) HaN=7, Wo
=Wl =W2 =O,,2.

This is a case where the injuries are equal. Therefore, in each case, the weighted moving average of the pulse movement amount takes a step shape as shown in the figure, and is approximated by a curve (a straight line in the case of (C)).

As described above, by appropriately determining the number of cycles of the pulse movement amount (N-/) and the weighting value Wj (j = 0.N-/), any arbitrary, for example, straight line, It is possible to select acceleration/deceleration patterns such as an S-shaped curve, an exponential curve, and a broken line. - Next, FIG. 6 shows a block diagram of the main parts of an orthogonal coordinate type industrial robot as an embodiment to which the industrial robot acceleration/deceleration control method of the present invention is applied. Teaching point data memo IJ
l is the number of clocks M required to move between teaching points,
Teaching data such as the number of increment pulses LX, Ly, and LZ between teaching points on each of the X, Y, and Z axes is stored.

FTP Performance I! In circuit -, teaching point data memory 1
The number of clocks M1 incremented by /l/
From the data of "X L'/ r L z"
, pulse movement amount of each Z axis 5xn=Lx/M, 5yn=L
V/M+5zn=Lz/M is calculated o 3. 'l,
s is a weighted moving average of the X-axis pulse movement amount 3Xn, the Y-axis pulse movement amount syn, and the Z-axis pulse movement amount Sin calculated by the FTP arithmetic operation circuit, SXn, Syn, and Szn. Arithmetic circuit, Y
These are a moving average calculation circuit with axis pulse movement weighting, and a moving average calculation circuit with weighting of two-axis pulse movement amount. Regarding the operation of the X-axis pulse movement amount weighted moving average calculation circuit 3, see the seventh section.
The explanation will be based on the flowchart shown in the figure. Each time tn
X-axis pulse movement amount SX'n' (n==o, l, 2..., n) of (n'70.l, 2..., n)
. . , n) are input in series from the FTP calculation circuit 2 at a constant cycle.

Of these X-axis pulse movement amounts, N X-axis pulse movement amounts Sxn, 5X(nl)S during (N-/) periods
X (n-2) , =-・-8X(n-N+1) (n
=o, l, 2,...-p) are always stored in the memory. Then, N weights W Or W 1 which are set in advance together with the value of N and stored in another memory are
. . . Value of WN-1 and N X-axis pulse movement amounts 3Xn stored in the above memory.

5X(n-1), 5X(n-2)-,・Song・, 5X(n
2N+1), the weighted moving average SXn of the X-axis pulse movement amount at time tn is calculated, and as a digital value, the rotation motor detected by the rotational position detector PGX is sent to the control device O' of the X-axis digital servo system A. It is output together with the signal X indicating the rotational position of MX. Similarly, the Y-axis pulse movement weighted moving average arithmetic circuit and the X-axis pulse movement amount weighted moving average arithmetic circuit j are respectively syn.

The digital servo is output to the system control device g'. Furthermore, X
My, PGy, and V are the rotary motor and rolling displacement of the Y axis, respectively.

In the present invention, the calculation of the amount of pulse movement that moves at a constant speed between two points in the teaching point sequence and the weighting of a predetermined number of consecutive time series data of these pulse movement amounts are performed using only a very simple calculation called a moving average. Any pattern of acceleration/deceleration control can be realized. Furthermore, this acceleration/deceleration control is automatically performed only for axes with speed changes, and acceleration/deceleration control that could previously only be performed when stopping at each teaching point has been changed from Since it has desirable functions not found in conventional acceleration/deceleration control, such as being able to control the speed even when the robot is in the becomes possible.

[Brief explanation of the drawing]

Fig. 1 is a diagram generally showing a teaching point sequence, Fig. 2 is a diagram showing an example of acceleration/deceleration control of an industrial robot with intermittent trajectory control, and Fig. 3 is a diagram showing the pulse of a conventional industrial robot with continuous trajectory control. Figure 1 shows changes in the amount of movement, Figure 1 is a diagram showing an example of calculating the weighted average of the amount of pulse movement according to the present invention, Figure 1 is a diagram showing an example of calculating the weighted average of the amount of pulse movement according to the present invention, and Figure 1 is a diagram in which the acceleration/deceleration control method of an industrial robot according to the present invention is applied. FIG. 7, which is a block diagram of the main part of the industrial robot/embodiment, is a flow chart showing the operation of the X-axis pulse movement amount weighted moving average calculation circuit of FIG. 3. /...Teaching memory 1 2...PTP calculation rotation 3...X-axis pulse movement amount weighted moving average calculation circuit t...Y-axis pulse movement amount typed moving average calculation circuit 1 5. -! rz-axis pulse movement weighted moving average calculation circuit, B...X-axis digital servo system, 7...Y-axis digital servo system, g...2-axis digital servo system. Figure 3 Figure 2 Figure 4 Figure Procedural Amendment (Method) April 5, 1980 Commissioner of the Patent Office Sir 1, Indication of Case 1981 Patent Application No. 223942
No. 2, Name of the invention Acceleration/deceleration control method for industrial robots 3, Person making the amendment Relationship to the case Applicant: Yaskawa Electric Manufacturing Co., Ltd. 4, Agent address: 1-9-20-5 Akasaka, Minato-ku, Tokyo, Amendment Date of Order (1) IIA Specifications, page 11, line 6, saku is corrected to saku 5.

Claims (1)

[Claims] Discrete teaching point data is given, and the pulse movement amount of each axis is calculated from the teaching point data at regular intervals. In an industrial robot driven by these digital servo systems, for each axis, N pulses are generated during the (N-1) cycle period (7) before time tη in the second adjacent teaching point sequence. Movement amount 3n-j' (j=IJ-1, N-2,... Out-,
An acceleration/deceleration control method for an industrial robot characterized by performing acceleration/deceleration control as a pulse movement amount in a weighted moving average of l, O)