WO2011110278A1

WO2011110278A1 - Method for transferring the working program of a robot affected by individual errors to a second robot affected by individual errors

Info

Publication number: WO2011110278A1
Application number: PCT/EP2011/000725
Authority: WO
Inventors: Franz Ehrenleitner; Carsten Meyer; Robin Kaufmann
Original assignee: Eisenmann Ag
Priority date: 2010-03-10
Filing date: 2011-02-16
Publication date: 2011-09-15
Also published as: DE102010010919A1

Abstract

The invention relates to a method for transferring the working program of a first robot (R1) affected by individual errors to a second robot (R2) of the same series which is affected by individual errors, wherein under the control of said working program the tool centre point of the first robot (R1) is guided on a trajectory defined by a plurality of support points. For this purpose, first a functional working program working at the desired precision is provided for the first robot (R1). The individual errors of the first robot (R1) are determined by measurement and then computationally eliminated from the functional working program. In this way, a series-specific reference program is obtained. Next the individual errors of the second robot (R2) are determined by measurement and computationally included in the series-specific reference program. As a result, a functional working program which works at the desired precision is provided for the second robot (R2). In the method according to the invention, it is not necessary to know the operating program provided by the manufacturer of the robot or to intervene in said operating program. The number of measurements necessary for ascertaining the errors of the second robot (R2) can be kept small because series-specific errors do not have to be taken into consideration.

Description

Method for transferring the work program of a first robot with individual errors to a second robot with individual errors

The invention relates to a method for transmitting the work program of a first robot subject to individual errors, under whose regime the "Tool Center Point" (TCP) of the first robot is guided on a trajectory defined by a plurality of support points, to a second one Individual error-prone robots of the same series.

When using industrial robots, a distinction is made between direct and indirect programming. Direct programming, also called on-1ine programming, works mainly with teach-in procedures. Here is by means of a keyboard, z. B. a handheld programmer, the industrial robot moved in a Cartesian coordinate system or axis-specific to the target positions and the results are stored in the controller. The advantages of this direct programming method are that the industrial robot can be positioned based on the workpiece and that additional instructions can be entered directly. The disadvantage is that the industrial robot fails during program creation for production.

Therefore, attempts were made at an early stage to relocate the programming to areas outside of production and to carry out indirect programming, also called "off-line programming". For about 20 years, there are systems that the graphically interactive Allow simulation and programming of industrial robots in a three-dimensional representation. Here, a model of the industrial robot, the manufacturing cell, in which the industrial robot is arranged, and the workpieces created, with the help then

Programming tasks can be performed.

The off-line simulation systems allow the direct description of the nominal trajectory in the basic coordinate system of the industrial robot. However, a direct transfer of simulation programs into production is often not possible, because reality and ideal

Model differ greatly due to inaccuracies and modeling errors.

For this reason, it is necessary to calibrate industrial robots in order to achieve the required high absolute accuracy. The application of automated measuring methods with sufficient resolution is indispensable. To determine the robot parameters from the measured data, almost exclusively systems-theoretical approaches have been used so far. In this case, the parameters of the overall model are identified by means of numerical methods. For this purpose, the deviations of the robot from the nominal position are measured at a sufficient number of spatial positions.

Thereafter, an attempt is made to minimize the Euclidean form of the total deviation - starting from the ideal model - by varying the model parameters. The achievable limit accuracy depends on the one hand on the number and position of the spatial positions and on the other hand on the result variables, eg. As the arm lengths, angles - deviations, axle stiffness, etc. from. Especially in the automotive industry are robots often involved in a "production line". Here, the failure of a robot may lead to

Surrender of entire production stages. In order to avoid this, functionally parallel "production lines" are set up to minimize the overall impact on production in the event of a robot failure. But also at

Maintenance or repair of a robot in the same line, the robot in question should often be replaced by another robot.

Object of the present invention is to provide a method of the type mentioned in such a way that the mathematical and metrological effort is kept small.

This object is achieved in that a) a functional, working with the desired accuracy work program of the first robot is provided; b) the individual errors of the first robot are determined by measurement; c) eliminating the individual errors of the first robot from its functional work program, whereby a model-specific reference program is obtained; d) the individual error of the second robot through

Measurement to be determined; e) the individual errors of the second robot in the

built-in reference program can be counted, creating a functional, with the desired accuracy working program of the second robot is obtained.

The above sequence of the individual method steps a to e does not necessarily reflect the time sequence of these method steps. In particular, it is possible to perform the measurements given in steps b and d at an earlier time. Their results need only be available if you want to perform steps c and e.

The invention makes use of the knowledge that the large number of errors with which robots can be involved can be roughly divided into two groups:

The first group of errors includes all those determined by the series, which are common to all robots belonging to a particular series and which are subject to design by this series. These include, for example, joint elastics with and without load or elasticities. These errors are referred to here as "model-specific". The other group of errors concern individual errors of the individual robot, ie those errors in which the considered robot also differs from robots of the same series. These include, in particular, axis position errors, arm length and angle errors, temperature influences, gearbox errors and stochastic errors.

With the method according to the invention, the transmission of a work program from a first robot to a second robot is particularly simple because, in particular, all model-specific errors are eliminated. They can not be considered. These model-specific errors were already in the creation of the functional, with sufficient accuracy

work program of the first robot. If according to the invention (only) the individual errors of the first robot are calculated out, then first of all a reference program is created, which is here called "model-specific". This is because it is a "basic program" or just a "reference" from which all robots belonging to the same series can go out. Then only the individual errors of the second robot, to which the work program should be transferred, need to be calculated in order to obtain the functioning, sufficiently accurate work program of the second robot.

For the purposes of the present invention, it is irrelevant how the functional work program of the first robot is obtained. For example, it is possible to do this through a "teach-in" procedure. Although this determination of the first work program is then associated with the disadvantages of the "teach-in" method described above;

However, the transfer of the thus obtained, work program to other robots is then particularly simple and no longer associated with these disadvantages.

Alternatively, the functional work program

of the first robot also by measuring the model - specific errors and the individual errors and

mathematical inclusion of these errors in the of

Robot manufacturer provided ideal work program. Again, that applies to this

first determination of a functioning work program a certain amount of effort has to be made

However, in the transfer of this functional work program to other robots, the mathematical and metrological effort is extremely small. If the functional work program of the first robot is obtained in this way, in many cases the individual errors of the first robot are already known from this, so that step b) no longer has to be carried out separately.

In general, it is sufficient if only axis position and length errors are taken into account for the individual errors. Experience has shown that these errors contribute the most to the individual errors. Reducing the consideration of individual errors to Achslagen- and length errors or even only Achslagenfehler, so reduces the metrological effort again significantly; the mathematical treatment can be closed in many cases, which contributes to an increase in the accuracy and reduction of the computational effort.

In order to determine the axial position errors, it is sufficient if the robot arm is rotated about each axis and in each case three points lying on a circular arc are measured, from which the axial positions and orientations can be calculated. This means, in total, that with six axes only 18 measuring points have to be measured.

To achieve sufficient accuracy, it is desirable that the angular distance between the two furthest apart, to be measured

Points at least 90 °, preferably at least 120 °. In principle, it is conceivable that the measurement of an An individual error is influenced by other individual errors, so that the contributions of different individual errors can no longer be kept separate. This complicates the computational consideration of the individual errors in the determination of the model-specific reference program as well as in the determination of the functional work program of the second robot. Such a "superimposition" of other individual errors than the one who is about to be measured is therefore to be avoided as far as possible. In determining the axial position errors, this can be done, in particular, by the fact that the robot arm in the

Twist is held in such a Messpose in which other existing individual errors have no appreciable influence on the traversed circular path. The decisive factor is that the other individual errors, which are not known in detail, do not significantly distort the circular path as such

Position of the individual measuring points on the circular path can certainly change, without having any appreciable influence on the determined axis position.

An embodiment of the method according to the invention will be explained in more detail with reference to the drawing; show it

FIG. 1 shows a flow chart of the method according to the invention; Figures 2 and 3 Meßposen which a robot arm for

Measuring individual errors of a robot can take.

First, reference is made to FIG. The starting point of the method presented here is box 1, which stands for a functional work program of a first robot Rl. A "workable" work program is understood to be one which, with the required accuracy, is capable of guiding the so-called "Tool Center Point" (TCP) along a predetermined, specific trajectory which is defined by a plurality of support points is. Between these interpolation points, the program calculates an interpolation curve. "Sufficient" means that the sum of all errors in the movement of the TCP is within a certain tolerance window.

The way in which the "workable" work program represented by box 1 is obtained is irrelevant here. The starting point for obtaining this functional program is always the program provided by the manufacturer of the robot and not always known in detail. The series-related and individual errors of the robot Rl have been corrected in a suitable manner, for example by a "teach-in" method or by fully measuring all errors of the robot Rl and mathematical consideration by modification of the program provided by the manufacturer. It is crucial that the TCP of the robot Rl under the work program represented by the box 1, the predetermined trajectory everywhere with the desired accuracy. The work program of the robot Rl is now to be transferred to a second robot R2, either because the robot Rl has to be serviced or repaired, either because a second production line is to be set up, which carries out the same work as that production line which is the first robot ter Rl belongs.

The first step in the transmission of the work program from the robot Rl to the robot R2 is that the individual errors of the robot Rl are measured. This is shown in box 2. Details of the measurement of this individual error will be discussed below. With knowledge of the individual errors of the robot Rl gained in step 2, it is now possible to do this through the

Box 1 represented working working program of the robot Rl to convert into a design-specific reference program. The box 3 in FIG. 1 represents the arithmetic process, while the box 4 represents the type-specific reference program obtained.

"Type-specific reference program" is understood to mean a program which is freed from the individual errors of the robot R 1 but still contains the construction-site-specific errors. These model-specific errors require no consideration in the method according to the invention, since they are probably the robot Rl and the robot R2 own and in the creation of the

functional work program (box 1) of the robot Rl have already flowed.

The model-specific reference program (box 4) is now applied to the robot R2, which comes from the same series as the robot Rl. The robot R2 also possesses individual defects that need to be taken into account in addition to the type-specific errors that it has in common with the robot Rl. These individual errors of the robot R2 must be measured, which is represented by the box 5 in FIG. The individual errors thus determined are calculated (Box 6) for the purpose of modifying fication of the construction-specific reference program used, whereby a functional work program (box 7) for the robot R2 is created. In carrying out this method, it is clearly not necessary to gain knowledge about the manufacturer's own program of robots. The correction of the individual error occurs in that the robot R2 inputting bases whose coordinates do not coincide with those of the desired interpolation points of the trajectory, but whose input causes the TCP of the robot R2 to approach the desired interpolation points due to the individual errors. To carry out the mathematical calculations, the

are represented by the boxes 3 and 6 and in which the individual error of the robot Rl from the

Functioning work program (box 1) are eliminated in step 3 or the individual error of the robot R2 in step 6 in the construction-specific

Reference program (Box 4),

For computational reasons, the number of fault types considered should be reduced as far as possible. Measurements and tests have shown that 80 to 90% of the axis position errors alone contribute to the individual errors of the robot, while arm length and angle errors contribute 5 to 10% of the total error. It is therefore generally possible to measure in steps 2 and 5 for the preparation of mathematical steps 3 and 6 only the errors mentioned. This not only has computational advantages; Of course, it is also metrologically favorable to have to carry out only a few measurements in the determination of individual errors. When identifying the individual error should be on it be respected by taking appropriate

As far as possible, only one type of error is detected in the measuring position of the robot arm, ie that a mixture of different error influences in the measurement result is avoided. How this can be done in the measurement of Achslagenfehlern is based on the figures 2 and

3 clearly:

First, the robot is the manufacturer of the

Provided system, z. B. brought by the robot manufacturer set zero marks in the zero position. Then the individual deviations of the robot are determined by single-axis measurement. This happens because the axes of the robot are moved individually and from three measuring points on one

Circular arc around the real axis, the respective

real axis position is determined.

The measuring techniques used here in detail are not in the present context of

Interest. In order to achieve sufficient accuracy, the three measuring points measured for each axis lie at an angular distance of 60 ° from each other. FIG. 2 shows the measurement of the first axis of the robot arm. The measuring pose is chosen so that

errors other than deviation of the position of the first axis, although the actual position of the circular path, which is traversed during the measurement, possibly deviates from the expected circular path. Nevertheless, the path traversed by the TCP remains a circular path, the axis of which largely coincides with the real position of the axis 1.

Stochastic errors, e.g. B. in the individual joints, affect in this way practically not. FIG. 3 shows how the axis 3 of the robot can be measured. Even with the measurement pose shown here, it can be seen that, in the presence of errors other than errors in the position of the axis 3, the individual measurement points do not ideally match the desired measurement points; However, the actual measuring points remain essentially on the desired circular path, so that the determined position of the axis 3 largely coincides with the actual position of the axis 3.

Claims

claims

1. Procedure for transferring the work program

of a first robot subject to individual errors, under whose regime the "Tool Center Point" (TCP) of the first robot is guided on a trajectory defined by a plurality of support points, to a second individual error-prone robot of the same series, characterized in that providing a working work program of the first robot (Rl) operating with the desired accuracy; b) the individual errors of the first robot (Rl) are determined by measurement; c) the individual errors of the first robot (Rl) are calculated from its functional work program, whereby a construction-specific

Reference program is won; d) the individual errors of the second robot (R2) are determined by measurement; e) calculating the individual errors of the second robot (R2) into the model-specific reference program, whereby an operable work program of the second robot (R2) working with the desired accuracy is obtained.

2. The method according to claim 1, characterized in that the functional work program of the first robot (Rl) is obtained by a "teach-in" method.

3. The method according to claim 1, characterized

that the working work program of the first

Robot (Rl) by measuring the model-specific errors and the indivudual errors under mathematical

Including these errors in the ideal operating program provided by the robot manufacturer is obtained.

4. The method according to any one of claims 1 to 3, characterized in that in the individual errors only

Axial position and length errors are taken into account.

5. Method according to one of the preceding claims,

characterized in that for determining the

Axis position error, the robot arm is rotated about each axis and each measured three points lying on a circular arc, from which the axis position and the axial direction can be calculated.

6. The method according to claim 5, characterized

that the angular distance between the two furthest apart points to be measured is at least 90 °, preferably at least 120 °.

7. The method according to claim 5 or 6, characterized in that the robot arm is held during the rotation in such a Messpose, in which other existing individual errors have no appreciable influence on the traversed by the "Tool Center Point" (TCP) circular path.