JPS5856784A - Method of driving robot when teaching - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a method for driving a robot during teaching.

In recent years, welding has become popular, especially in the manufacturing and processing industry.

Various industrial robots are being actively developed in order to save labor by automating tasks such as painting, assembly, and machining.

As industrial robots, intelligent robots that can act autonomously and have various movement functions, recognition functions, sensory functions, etc. similar to those of human upper limbs, for example, are becoming mainstream.

And, for such industrial robots, welding,
Painting 1 In order to obtain movement trajectory data (movement data) necessary for assembly, polishing, etc., it is possible to adopt a so-called direct teaching method in which the robot's arm is moved directly by human power to memorize the coordinates of points to be passed. be.

By the way, in order to realize this direct teaching method, each axis of the arm must be easily moved by human power.
As the reducer interposed between each axis of the arm and the drive motor, it is necessary to use a reducer with good reverse transmission efficiency (for example, a reducer configured with S/P gears).

However, this limits the number of speed reducers that can be used, and it is not possible to use a small speed reducer with good transmission efficiency (for example, product names such as Monikdra, Eve, etc.) required for the robot.

In addition, even if a reducer with good reverse transmission efficiency is used, a certain amount of force is required to move parts of the arm with relatively large inertia other than the wrist (such as the waist), which makes the robot free. There was a limit to how far I could move it.

This invention has been made in view of the above background, and in a multi-axis robot equipped with a mechanical handler at the tip of the arm, the mechanical...
A method for driving a robot during teaching is provided in which the direction of force applied to the hand is detected, and each axis of the arm is controlled so that the mechanical hand moves in the direction of the detected force. Even if a robot uses a reducer with poor transmission efficiency, the teacher can easily teach directly.

Embodiments of the present invention will be described below with reference to the accompanying drawings.

Figure 1 shows a multi-axis robot 1111 to which this invention is applied.
It is a symbolic diagram showing the configuration.

In the same figure, R+ and R2 respectively represent rotating axes from P1 to
P4 indicates the axis of rotation, and each of these axes R7
, R2, P1 to P4, that is, a force sensor 1, which will be described later, is inserted between the shaft P4 and the mechanical node MH.

Axis P51P4. The wrist portion T constituted by R2 and the mechanical arm MH are constructed as shown in FIG. 2, for example.

The force sensor 1 has a structure as shown in FIGS. 3 and 4, for example.

In both figures, a wheel-shaped base 2 is attached to the shaft P4 side, and a disc-shaped base 3 is attached to the mechanical node MH side.
and 1 are connected by a loose joint 4.

Three load cells 5 to 7 are attached to the outer circumference of the base 6 at positions shifted by 120° from each other in phase, and each position of the base 2 corresponds to each load cell 5 to 7. The pins 8 to 10 are screwed in so that the tips of the pins 8 to 10 come into contact with the load cells 5 to 7.

Note that the load cells 5 to 7 are given preload by pins 8 to 10, and the initial outputs of the valleys and valleys are made equal.

Moreover, in the force sensor 1 configured as described above.

The force applied to the center of the base 3 is applied to the axis P via the ball joint 4.
4, the force cannot be detected. Therefore, if the force is to be detected, a strain gauge may be attached to the rim portion 2a of the base plate 2.

However, in the embodiment described below, the force sensor 1 to which no strain gauge is attached is used because it is very rare to press the mechanical/hand MH in such a way that only such force is generated.

Force sense unit 1 constructed as shown in Figures 3 and 4
is interposed between the shaft P4 and the mechanical hand MH as shown in FIGS. 1 and 2, so that when the teacher pushes the mechanical hand MH in a desired direction during teaching, the method of pushing the mechanical hand MH can be easily controlled. It becomes possible to output a signal indicating the direction of the force applied to the mechanical node MH in accordance with the mechanical node MH.

FIG. 5 is a block diagram of a processing circuit showing an embodiment of the present invention.

In the figure, the arithmetic circuit 11 is connected to the mechanical hand MH.
Each load cell 5 of the force sensor 1 is pressed in a predetermined direction.
Based on the detection signals f1 to f3 outputted from 7, fα knee f++ f2/2 + fs/2 fβ=i〒/2(f2-fs) fγ2 f++ f2+f3 are calculated, respectively. However, if a strain gauge is added to the force sensor 1 as described above, f
γ is the sum of the detection signals f1 to f3 of the load cells 5 to 7, as well as the output of the strain cage 9.

The teaching pendant 12 determines the working posture of the robot. When the instructor operates the teaching pendant 12, the operation signal ST is inputted to the central processing unit 14 via the interface 16, where the operation signal ST is After the data is converted into corresponding movement data, each axis of the robot is driven via the sensor controller 15 and a drive motor for each axis (not shown). Note that during teaching, the teaching pendant 12 is also operated when the mechanical hand MH is to grip a workpiece.

The interface 13 receives calculation results fα and fβ output from the calculation circuit 11 according to instructions from the central processing unit 14.
, fγ into digital signals and transmits them to the central processing unit 14.

The central processing section 14 is composed of a microcomputer or the like, and performs processing as shown in FIG. 6, for example, at 70--.

That is, first, the calculation result fα is output from the calculation circuit 11 via the interface 13 at 5 TEPI.

Take in fr, fr in the form of digital signals, 5TEP2
So f7 is set IIi! fo Check whether it is very large.

Then, if fr≦f, return to 5TEP 1, fr
>fo, proceed to 5TEP3. In addition, 5TEP2
Checks such as these are performed for mechanical hand handling.

In 5TEP3, the direction cosine tH, mH, and nH of the force F (pressing force against the mechanical hand MH) in the wrist coordinate system are determined based on the acquired fα, fr, and fr.

That is, as shown in FIG.
The force F applied to the point of force P is F(fα, fr.

Since it can be expressed as fr) [force vector], the wrist coordinate system whose origin is, for example, the fulcrum Q (corresponding to the center of the ball joint 4) of the wrist surface 16 including the base 2 of the force sensor 1 (FIG. 4) F(fα. in (z, yes).

The direction cosine tH, mH + 7LH of fr, f, ) is
It can be obtained by calculating も9.

In 5TEP4, based on the position and direction data of the fulcrum Q of the wrist surface 16 in the absolute coordinate system of the
Direction cosine zn, m in the wrist coordinate system obtained in P3
)I, nH to the direction cosine 10 + tnO+ nO in the absolute coordinate system of the robot.In addition, this direction cosine ZO+ ? FlO* tio is mechanical
・It shows the direction of the force F applied to the hand MH in the robot's absolute coordinate system.

Further, the position and direction data of the fulcrum Q of the wrist surface 16 in the absolute coordinate system of one robot can be obtained from the output of an internal sensor (such as a yometer) provided on each axis of the robot.

In 5TEP5, the direction obtained in % 5TEP4 ('o 1
fiQ + tio), the mechanical hand MH sets the set speed υ by a predetermined set value δ. The movement data for each axis is output to the servo controller 15 to move the mechanical hand MH under the above conditions. However, when controlling the drive of each axis, the working posture of the robot determined by the teaching wave dant 12,
That is, movement data for each axis is calculated such that the direction of the wrist surface 16 does not change.

Then, in 5TEP6, after the mechanical hand MH moves, the moving acid of each axis is obtained by reading the output of the inner world provided on each axis, and after storing it in a teaching memory (not shown), the process returns to 5TEPI. .

Since the central processing unit 14 operates as described above, the instructor pushes the mechanical hand MH in a desired direction every time the mechanical hand MH moves and stops (this pushing does not directly move the robot by force). ) can teach the robot's work movements.

Incidentally, since the load cells 5 to 7 of the force sensor 1 are subjected to gravity loads including themselves or the gripped workpiece depending on the posture of the mechanical hand MH, the load cells 5 to 7 of the force sensor 1 are If the above output is read and used as the initial output of each load cell 5 to 7 to perform the above-mentioned processing, the accuracy of teaching will be improved.

As described above, according to the present invention, when the teacher pushes the mechanical hand in a desired direction during teaching 77, each axis of the robot is automatically driven so that the mechanical hand moves in that direction. This not only makes it possible to use a reducer with poor reverse transmission efficiency, but also eliminates the need to directly move parts with relatively large inertia.

[Brief explanation of drawings]

FIG. 1 is a symbolic diagram showing the axis configuration of a multi-axis OX throat to which the present invention is applied. FIG. 4 is a sectional view taken along line IV-IV in FIG. 3; FIG. 5 is a block diagram of a processing circuit showing an embodiment of the present invention; FIG. FIG. 7 is an explanatory diagram for explaining the processing flow diagram shown in FIG. 6. 1... Force sensor 2, 6... Base 4... Ball joint joint 5-7... Load cell 8-10...
Pin 11... Arithmetic circuit 14... Center, processing section
15...Sir? Controller MH...Mechanical hand Figure 6 Figure 7 Order: 46

Claims (1)

[Claims] 1. In a multi-axis robot equipped with a mechanical hand at the tip of an arm, during teaching, the direction of the force applied to the mechanical hand is detected based on the output of a force sensor interposed between the tip of the arm and the mechanical hand. and, based on the detection result, drive-controls each axis forming the arm to move the mechanical hand in the direction of the detected force. Robot movement method.