JPS6045806A - Check system for working locus of industrial robot - Google Patents

Abstract

PURPOSE:To prevent previously accidents such as a collision, etc. produced in a reproduction mode by calculating a change amount of the component of velocity of each drive shaft at a joint of a straight line from the teaching data and performing the prescribed acceleration/deceleration control between straight lines to reduce a locus error. CONSTITUTION:The maximum acceleration (amax) and the maximum deceleration (dmax) are given as constants, and at the same time the maximum speed (vmax) is given for each straight line. Here a straight line P'i-1P'i has no deceleration but keeps a fixed speed until a point P'i in case vmaxi>vmaxi-1 is satisfied. While a straight line PiPi+1 has deceleration so as to obtain a speed vmaxi+1 at a point Pi+1 with vmaxi+1<vmaxi. In addition, a straight line P'i+1P'i+2 is decelerated down to zero at a point Pi+2 as long as the stop conditions are satisfied for waiting a timer or external signal. Thus a designated range ACCi of accuracy is increased to avoid a collision to an obstacle 11 despite an overshoot produced from the point P'i.

Description

DETAILED DESCRIPTION OF THE INVENTION +'The invention relates to a method for inspecting the entire motion trajectory of an industrial robot of a so-called memory and replay type.

As is well known, as a method of controlling a robot along a linear trajectory, a calculation means such as interpolation or coordinate transformation is used to find a target point on the robot's motion trajectory from time to time, and the robot is positioned at this target point. There is a method in which the position of each drive axis is given as a command value for the control system of each axis. In this method, the operating speed of the robot is controlled by controlling the magnitude of a speed vector for moving the target point. The magnitude of this velocity vector is
If the period for finding the target point is made sufficiently small, it can be changed continuously, but the direction of the velocity vector becomes discontinuous at the joints of the straight lines.

Therefore, the amount of change in the command value of the control system for each axis, that is,
The speed command value also becomes discontinuous, and in order to follow this discontinuous speed command value, an acceleration of +1 to infinity is required for each drive 11ζ11. However, since the acceleration that each drive shaft can generate is finite, a trajectory following error occurs at a straight line joint depending on the magnitude of the speed command values h'r and hIl.

It is difficult for the instructor to predict trajectory errors caused by the causes mentioned above, so in the past, it was necessary to actually reproduce all the taught data to confirm the trajectory error, and if it was large, (gender, teaching points, etc.)
A method of adjusting the teaching speed has been used. but,
In order to confirm this, the collision between the robot and the workpiece and the %-, Y, 9'<, '! There is a possibility that an accident may occur due to the impact force on the stroke end of the positive and negative axes, so a method of checking the trajectory difference without actually operating the robot has been desired.

The invention is to solve all the problems mentioned above and create a robot (
The purpose of this paper is to provide a total of 4 methods for inspecting the motion trajectory of a robot for ordinary tasks that allows estimation of trajectory errors without making H4 movements and correction of teaching data based on stiffness.

FIG. 1 shows the entire robot 7j used in this embodiment, in which each motion $ll~6 independently changes the 1'j joint angles θ1~θ6 as shown by the arrows, so that the axis 6 It is fully shown that the position and attitude of the working tool 7 attached to the tip of the tool 7 is controlled. Direct expansion end P work, P2.・
...Pi, Pi+7. Each joint angle in ... is determined by the teaching operation in Q1. It is done. These points P1゜P...
Is it possible to connect P,... with a straight line in order?29 j
1 This is the taught trajectory 8. In this robot, the position and posture of the work tool 7 expressed by each joint angle (=(θ1.θ2・,・9,θ6)−・・・i+) and the orthogonal coordinates X, Y, Zf (x, y, z , α, β, γ)T...(
2) are related to each other by appropriate coordinate exchange. This coordinate exchange formula is based on the robot's axis configuration and posture (α,
Although it differs depending on how β, γ) are taken, etc., the method of deriving the formula is well known, so the details are omitted, and the following notation is used.

t=chi(0) “-+31 0=&I(X)...(4) FIG. 2 shows teaching points P, , 1), in this example.

, i -11 Pi+1, Pi+2' The teaching locus 9 connected with the solid line and the point "i-1'P' i '" i +
□.

P'i+2' is connected with a broken line to indicate the command locus 10 during reproduction.

FIG. 3 shows an example of the addition/subtraction modal A mountain j at the time of reproduction with respect to the taught locus shown in FIG. This example - Cjl,
The maximum acceleration 1iamax and the maximum deceleration dmax are given as constants, and the maximum speed vmax is taught for each line.
Here, when v max > v maXi-1-(5), point 1)
'With one bite, it doesn't slow down with constant luck, and it's direct! 1-11P
1+1, vmaX1+1 (vmaX4 -'-(6) The velocity at the point Pi+1 is V f'l'l a X 1 + 1
So that it becomes impossible i・iZ L, outside if, k hair Pi
According to +I Pi+2, if there is a stop 1-1 stop 1 such as waiting for a timer or external signal at point 1]i+2, the 7c control is applied to decelerate to zero speed? 1L11-IJ:
f(, indicates that it will be done.

Such trajectory control and acceleration/deceleration control are performed, for example, at point L”V
If an overshoot occurs at point C and there is a risk that the robot will collide with the obstacle 11, (1) Even if the specified accuracy range Ace is widened and the bar shoot is set from P'1 to 1, Avoid side collision with the obstacle proboscis 11.

(2) vmax・ Reduce the total value and the speed of entry into Pi - 1 Reduce the total value. Or (:3) decrease vrnax・i and directly mA L”11
Slow down sufficiently while moving Pik.

The national difference in trajectories can be reduced by the following method.

Although a practical IiQ example of the present invention will be described for an industrial robot model that performs trajectory control and acceleration/deceleration control as described above, the present invention is not intended to be limited to robots with such a control method. Instead, the spirit of the invention should be construed based on the scope of the claims appended hereto.

A method of inspecting teaching data regarding the straight lines P1P and □ in FIGS. 2 and 3 will be described. In FIG. 4, symbols are defined as follows.

vS...Speed at point P'】-1...(7) vM...Maximum moving speed V of point P'i Pi +2
M = V max 1 = 18) vE, target speed 1 + 1 at -3 point P1 + 1...
(9) a... Acceleration during acceleration a = a max ) O-110) d... Force at speed [, speed d = d max (0-...(It)t] ・
・・Gradient [11 countries 4 o'clock ([11t2・・constant speed 11.1
r111 L3...Reduction j/profit time ■S...Each joint angle at point P'1 ■s=(θ1s, θ2S,..., θ6s)...(1
2) ■E...point" i + 1 teaching data ■E-(
θIE,..., θ6E)T...(13)O5...
・Position and orientation of point P', χS== (xs, yS, zS, αS, βS, θ5) 1
=91(O8)...(14)XE...Position of point P1+1, attitude %, E=(xE, yE, zE, αE, βE, E)"=f
f) ((E)E) =l+5)I...point P' and P1
+, distance 1 = (xs-x+:)2+(ys-yE)2+(Z
S-ZE).2-+16+At this time, the following two relational expressions hold true between the valley parameters.

vs+atl=vM=VE-dt3-'(ll+) By solving the 17th and 18th equations. In the above, tl, t2, and t3 were determined for the case where there are acceleration, constant speed, and deceleration sections as shown in Fig. 4, but for example, if t2≦0...
If this is the case, there will be no constant speed section. However, in this case as well, tl and t3 can be determined in the same way by setting t2-0 in equation 17 and equation S18.

Next, the specified accuracy range A c i 'rilL/cf point p/・l+1 position 1・i−1 attitude V,, E−(x′ E , . . . , γ′F)T ・・・(
b) Travel time T from point P' to P' and point "l"
+11 l+1 (Find the total velocity V'E at l. Point P' is PP1+
Since it is an internal division point of 1 1 1 + 1 / (bar shi.

Acc・〉! ... For the ringing of 0[Yes], it is SoE-SoS...(c). ]゛, V'E is accelerated at point P'i+1,
(...) Depends on which section of deceleration it is in, as shown below.

i) Accl)/...(a) When T-0...Kan vE=vS...blade When 1-Goshizuku---Ko-j to μ-AC°i),,,,.

v E = vS 10aT...; (3t!i) 12
3≧Ac c ]≧v M t3+-dj32=13
=-', 3'j when VE=VM...G■ 1V) 13≧Acc≧0...(V11 or more, position of 9 points P'1+□, attitude x E and speed 10: v'p: Next, find the total amount of change in the minute speed of +1 for each movement iii at point P'. Just before reaching point P', the minute speed of 1+1 [with] IN-(2111N, ..., δ61N )T・
・・C))) An appropriate minute distance Δl (5 in this example)
rnm ), where 'y, 1N-iE-Δl (X E-Ys)/L ・
...(4]). The command value of the minute speed immediately after reaching the point P'i+1 = (S OUT ,..., h'60UT)
T...The position and attitude of point P, +2 are Xi. Close X0
UT=X'E+Δl! (lI+2'X'E)/L'...
・(Foundation)...It's an egg. Therefore, the elementary velocity IW of each motion axis at point P'i+□ is expressed as follows. The square of this amount of change (Δ0)2 is the point P'1+□
Using an appropriate coefficient Kj (j-=1.6), the j-th !I is approximately proportional to the overshoot amount in
An overshoot of (Δshij)2Kj occurs due to the influence of l+. Since the teaching point is P, the overshoot amount A, which is l+1 for Pi+□, is the value obtained by subtracting Aeci, which is Ai-(Δmij)2Kj-Acci+□...-1. In this example, we set an appropriate orchid ε and (Δlj
j) 2K j −Ac c・〉ε ...0.

In the case of 1+1, it is determined that overshoot will occur, and the instructor is notified of the teaching point number where overshoot occurs and the cause of all overshoot as a replacement report for the ω axis. All systems are in place to provide information for making decisions such as changing the position or slowing down. By adding the travel time Ti of the Ti@ line, the entire cycle time of the reproduction operation can be predicted in advance. In each of the above-described embodiments, a method was adopted in which the industrial robot was checked in advance without operating the entire robot, but a method in which the robot was checked while operating the robot could easily be considered.

As described in detail above, according to the present invention, it is possible to estimate problems such as overshoot during playback of taught data in a short time without actually operating each arm of an industrial robot. This makes it possible to completely prevent accidents such as collisions during playback operations.

[Brief explanation of drawings]

No. 1 No. 1 is the configuration k of the operating shafts 1 to 6 of an embodiment of the present invention.
Figures 1 and 2 show the teaching trajectory 9 and the command 1 during playback.
FIG. 3 is a graph for explaining all the acceleration control, and FIG. 4 is a graph for explaining all the parameters when performing all the acceleration control. 1 to 6...operation 1i11, P, P・, p-p・1-
1'll+1tsu l+2 Agent Patent Attorney Kei Nishi Figure 1 Figure 2 Cause Figure 3 Multi-Gravity Ryo Kyoto

Claims (1)

[Claims] Divide the entire series of movement trajectories of the working end into multiple straight lines, store the positions and movement speeds at the end points of the straight lines as all teaching data, and make sure that the working end moves on all the memorized straight lines at the memorized movement speeds. In an industrial robot that fully controls the position and speed of each drive axis, the amount of change in the minute speed of each drive axis at the joint between shifts and positions is calculated from the teaching data above, and when each drive axis is operated, h < = t + A method for inspecting the motion trajectory of an industrial robot, which is characterized by estimating all the errors between the taught motion trajectory at the @ line joint and the actual motion trajectory 1,',)1. which can be reproduced during playback.