JPH0441188A - Correction method for axial dislocation of robot - Google Patents

Abstract

PURPOSE:To extremely facilitate axial dislocation correction work, by detecting the dislocation quantities of the positions in each of axial directions of X, Y, Z in a robot coordinates system from the relation of the each distance between one face of a rectangular parallelepiped and three distance sensors only by realizing a reference position, after generation of the axial dislocation, and calculating the axial dislocation quantity based thereon. CONSTITUTION:Each reference positions are reproduced by a robot to which a correction tool 7 is fitted and the translational small movement quantities which are the dislocation quantities of the terminal parts with respect to X, Y, Z axial directions are respectively found from the distances between respective distance sensors S1-S3 and faces A, B, C of a rectangular parallelopiped 8 of this time. The axial dislocation quantities of respective axes are operated based on the each translational small movement quantities and the axial dislocation thereof is corrected.

Description

DETAILED DESCRIPTION OF THE INVENTION A. Field of Industrial Application The present invention relates to a robot axis misalignment correction method, and is useful when applied to industrial robots such as deburring robots.

B. Summary of the Invention The present invention attaches a correction tool having three distance sensors to the hand of a robot, and calculates the
The correction tool is moved along any of the axes, the correction tool is parallel to one surface of the rectangular parallelepiped, and is opposed to the reference position to sandwich the reference position, and the reference position is reproduced after the axis misalignment occurs.
Based on the amount of positional deviation from the reference position detected by each distance sensor during playback, the amount of translational minute movement, which is the amount of deviation of the hand in each axis direction of x, y, and z, is determined, and based on this amount of translational minute movement, The amount of axis deviation of each axis is determined, and the axis deviation of each joint of the robot due to collisions during work is fully automatically corrected.

C2 Prior Art Currently, many types of robots are used in various industrial sectors. This type of industrial robot has a multi-jointed arm (e.g. 6 axes) with a tool attached to its tip, and moves based on data taught in advance to carry out predetermined tasks such as deburring. There is something to do.

D. Problems to be Solved by the Invention In the case of the above-mentioned industrial robots, if the applied force becomes excessive during operations where force is applied to the hands, such as collisions due to incorrect operation or deburring operations, tools, workpieces, and the robot itself may be damaged. ,
Other problems arise. One of these problems is the axis misalignment of each joint of the robot. This axis misalignment is caused by excessive force being applied to the fastening parts of the reduction gears of the motors that drive each joint, and as a result, the reduction gears move slightly, resulting in a difference between the joint position detected by the position detection sensor and the actual joint position. This is a phenomenon that causes a misalignment between the position of the joint.

When such an axis deviation occurs, a trajectory error and a positioning error occur, so it is necessary to avoid these (re-teaching or correct the axis deviation).

Correction of axis misalignment according to conventional technology is performed by resetting the reference position of the robot, but in this case, a specific position and posture must be reproduced by manual operation using an operator box, and axis misalignment correction is performed by resetting the reference position of the robot. Not only does it take a lot of time, but there are also limits to accuracy.

An object of the present invention is to provide a robot axis misalignment correction method that is easy to implement and can achieve good accuracy at the same time, based on the above-mentioned prior art.

E. Means for Solving the Problems The configuration of the present invention that achieves the above object is as follows: In an axis misalignment correction method for a robot having an arm formed by connecting a plurality of links with a plurality of joints, the distance between them is determined by a predetermined distance. A correction tool that has three distance sensors arranged at a certain value is attached to the robot's hand, and the correction tool is attached to the robot's hand section along the approach direction, which is the direction along the X, Y, and Z axes of the robot coordinate system. Approach one surface of the rectangular parallelepiped perpendicular to the approach direction, teach the reference position at one point within the measurement range of the distance sensor, and perform this teaching multiple times on multiple surfaces of the rectangular parallelepiped to determine each reference position. A teaching process, and a robot equipped with the correction tool reproduces each reference position, and the deviation of the hand in the X, Y, and Z axis directions is calculated from the distance between each distance sensor and each surface of the rectangular parallelepiped at this time. The present invention is characterized in that it includes a step of obtaining respective amounts of translational minute movement, and a step of calculating the amount of axis deviation of each axis based on each amount of translational minute movement and correcting the axis deviation.

F Effect According to the present invention having the above configuration, when there is a possibility that an axis misalignment has occurred, such as when a hand collides during work, the reference position is regenerated by the robot. If such a reference position is reproduced, the amount of axis deviation can be determined from the distance data of each distance sensor at this time, and this is corrected.

G example Embodiments of the present invention will be described in detail below based on the drawings.

FIG. 2 is an explanatory diagram showing an arm of a robot to which an embodiment of the present invention is applied. In the figure, 1a is a 1-axis link, 2a is a 2-axis link, 3a is a 3-axis link, 4a is a 4-axis link, 5a is a 5-axis link, and 6m is a 6-axis link.
Axis links 1a to 6a are joint parts of 1 to 6 axis drive parts 1b
~6b. Among these, the 1-axis, 4-axis, and 6-axis drive units 1b.

4b and 6b perform a turning operation, and the two-axis, three-axis, and five-axis drive units 2b, 3b, and 5b perform a rotational operation.

Each axis link 1a to 6a and each axis drive part 1b to 6 for each same axis
The combination of axes 1 and 2b are referred to as axes 1 to 6, and the positions of each axis are indicated as θ and θ6 in the figure.

FIG. 3 is an explanatory diagram showing a correction tool used in this embodiment. As shown in the figure, the correction tool 7 includes three distance sensors on a disk 7a.

Tes, s2. They are arranged so that s and sl intersect at right angles.

FIG. 4 is an explanatory diagram conceptually showing an aspect of teaching a reference position, which is one step of the axis misalignment correction method according to this embodiment.

When teaching the reference position, the correction tool 7 is first attached to the hand of the robot.

At this time, if the correction operation is to be completely automated, A T
Set C (Auto Tool Changer) to enable automatic attachment and detachment. Thereafter, as shown in FIG. 4, the robot is moved so that the hand part follows the approach direction P1,
Distance sensors S, -S, are mounted so that the disk 7a is parallel to one surface A of the rectangular parallelepiped 8 fixed at a predetermined position.
The joint positions θ, -θ6 of each axis are taught at one point in the measurable range of . Let this be the reference position WQ.

At this time, as shown in FIG. 5, the rectangular parallelepiped 8 has each surface A,
Approach directions p, , p, , p3. B, C, D are directions along the X, Y, Z axes of the robot's coordinate system; p
It is fixed at right angles to 4.

Surfaces B, C,
A reference position Q2pQ3pQ4 regarding D is taught. The reference positions Q1 to Q4 are generally Q,=(θ,.,θ2n'
θ3n'θ4n'θ5m' θ6n),
In the case of this embodiment, there are six components θ1° to θ6.

Note that hand coordinates may be taught at this time.

If there is a possibility that an axis misalignment has occurred, such as when a hand collides during work, first the robot is made to reproduce the reference position Q. At this time, the distance data obtained from the distance sensor S-5 are assumed to be j, j, and l.

Here, L; distance between distance sensors S, S, and S (
(See Figure 3) θ; Inclination between S→S direction and surface A θ; Inclination between S→S direction and surface A, then θ1=m"" (j 2 j t ) /
Let Lθ2=m-1(13-1,)/L.

Therefore, the approach direction P of the distance sensor S and the surface A
, the distance! , then 1 = l θ
・It becomes sanθ.

If the distance data of the distance sensor S1 at the time of teaching the reference position Q1 is to, and the amount of deviation of the distance sensor S in the approach direction P due to axis deviation is Δt, then Δl =1 −1 is obtained.

Similarly, reference position Q2. Q3. The amount of deviation during reproduction of Q4 is determined and set as Δl, Δl, and Δl, respectively.

Here, since the rectangular parallelepiped 8 is arranged maintaining the relationship shown in FIG.
, the amount of deviation in either the Y or Z axis direction. This amount of deviation Δ1. ~△~ is the translational minute movement amount Δ of the hand regarding each axis
X.

They are called ΔY and ΔZ.

Here, the translational minute movement amount ΔX, ΔY of the hand.

Δz1 minute rotation amount is Δδ8. Δδ7. Δδ2, and the minute movement of the joint angle is Δθ8. △θ2. Δθ3゜Δθ4. △θ
6. Δθ6 has the following relationship.

Here, Jr (each component is the robot's joint angle θ1 to θ6
) is a 6×6 so-called Jacobian matrix, and from this formula, the amount of translational minute movement of the hand, for example ΔX
(In this example, Δ1. or ΔJd) is known, Δ
One equation regarding θ1 to Δθ is established. That is, q.

One equation holds true for each point Q, Q, Q4, so if, for example, the positions of six points are taught and the amount of hand deviation is measured, a six-element simultaneous equation is obtained for Δθ□ to Δθ6, By solving this, the amount of axis deviation Δθ, ~Δθ6 can be obtained.

Generally, axis misalignment is more common in the three axes of the hand, so Δθ1=△θ=
If Δθ = 0, it is only necessary to teach the positions of three points to solve the simultaneous equations of three elements, which becomes very simple.

After calculating the axis deviation regarding each axis as described above, axis deviation correction is performed based on the amount of each axis deviation.

The above-mentioned axis misalignment correction method is illustrated in a flowchart as shown in FIG.

FIG. 6 is a block diagram showing an apparatus for realizing this embodiment. In the figure, 11 is a man-machine processor;
2 is! 11a is a processor, 13 is a job memory for storing a program (JOB) that describes the motion of robot I, 14 is an operating box for operation, and 15 is a C
RT and keyboard, 16 is man-machine processor 11
and a RAM shared by the control processor 12, 17 an interface for controlling each joint of the robot 1 via a servo driver 18, 20 a CRT and keyboard interface, 21 a pass line, and 22 a distance sensor S.
This is a crowd interface for importing S3 distance data.

In such a device, first, JOR3 for correction, that is, the first
JOB corresponding to the algorithm shown in the figure! 1 Beresian box 14, HL < HA CRT and keyboard 1
5 to the job memory 13. This JOB first includes a command to attach the correction tool 7 to the hand by ATC.

Next, before the axis is shifted, surface A of the rectangular parallelepiped 8 is manually operated using the operation box 14.

Distance sensors S1 to S3 are connected to B and C. These distance sensors S1
~The reference position Q is made to face each other so that the output of S3 is the same.
,, Q2. Q3 is taught, and the reference positions Q1 to Q3 are
is stored in the job memory 13. In addition, the reference position Q,
, Q2. The output of the distance sensor S-S3 in Q3 is input via the CRT and keyboard 15 and stored in the RAM 16.
to be memorized.

On the other hand, a calculation formula for calculating the minute translational movement of the hand from the distance data of the distance sensor S-8 and a calculation formula for calculating the amount of axis deviation from simultaneous equations are built into the control processor 12, and can be called within the JOB. Let's configure it like this.

In this state, when the axis deviation correction job is started from the operating box 14 after the robot collides, the axis deviation is automatically corrected in the procedure shown in FIG.

Hl Effects of the Invention As specifically explained above with the embodiments, according to the present invention, after axis misalignment occurs, simply by realizing the reference position,
Based on the distance relationship between one surface of the rectangular parallelepiped and the three distance sensors, the amount of positional deviation in each axis direction of X, Y, and Z in the four-bot coordinate system is detected, and the axis deviation is determined based on the amount of deviation in B. Since the amount is calculated, it not only makes it extremely easy to correct the axis misalignment, but also allows the amount of axis misalignment to be detected with a certain degree of accuracy. Become.

[Brief explanation of the drawing]

FIG. 1 is a flowchart showing an embodiment of the present invention, FIG. 2 is an explanatory diagram showing a robot to which the embodiment is applied, FIG. 3 is an explanatory diagram showing a correction tool used in the embodiment, and FIG.
The figures are an explanatory diagram showing the manner in which the reference position of the robot shown in Fig. 2 is taught; 1 is a block diagram illustrating an apparatus for implementing an embodiment. FIG. In the drawings, 18 to 6a are 1 to 6 axis links, 1b to 6b are 1 to 6 axis drive units, 7 is a correction tool, 8 is a rectangular parallelepiped, si, s2. s3 is a distance sensor. Figure Figure

Claims (1)

[Claims] In an axis misalignment correction method for a robot having an arm formed by connecting a plurality of links with a plurality of joints, three distance sensors are arranged to ensure a predetermined distance between them. Attach a correction tool to the tip of the robot and approach one surface of a rectangular parallelepiped perpendicular to the approach direction along the approach direction, which is a direction along the X, Y, and Z axes of the robot coordinate system, and A step in which a reference position is taught at one point within the measurement range of the sensor, and this teaching is performed multiple times on multiple faces of a rectangular parallelepiped to teach each reference position; a step of reproducing the distance sensor and calculating the amount of translational minute movement, which is the amount of deviation of the hand in each of the X, Y, and Z axis directions, from the distance between each distance sensor and each surface of the rectangular parallelepiped; A robot axis misalignment correction method comprising the steps of calculating the amount of misalignment of each axis based on the amount of movement and correcting the misalignment.