WO2018206387A1

WO2018206387A1 - Robot adjustment

Info

Publication number: WO2018206387A1
Application number: PCT/EP2018/061315
Authority: WO
Inventors: Manfred Hüttenhofer
Original assignee: Kuka Deutschland Gmbh
Priority date: 2017-05-08
Filing date: 2018-05-03
Publication date: 2018-11-15
Also published as: DE102017004433B4; CN110891742A; DE102017004433A1

Abstract

A method according to the invention for adjusting a robot (1) has the steps of: - ascertaining (S30) a plurality of joint positions (q_1i, q_2i, q_3i, q_4i, q_5i, q_6i) upon starting on a specified path (g) by means of a detection means (2) using a reference (4) fixed to the robot; and - ascertaining (S40) an offset of the detection means on the basis of the ascertained joint positions.

Description

Description Robot adjustment

The present invention relates to a method for adjusting, controlling and / or monitoring a robot and a controller and a

Computer program product for carrying out the method and a

Robot arrangement with the controller.

In particular, incremental or relative encoders of a robot for detecting its joint positions can be adjusted by moving the robot into a defined adjustment pose, in which these detection means are initialized with values predetermined for this pose.

Correspondingly, the accuracy of the subsequently determined joint positions and thus in particular the accuracy of a position, determined in particular for controlling and / or monitoring the robot, based on these joint positions of a robot-fixed reference, such as the TCP, depends on this adjustment.

in particular the accuracy with which the adjustment pose is approached.

From DE 10 2007 001 395 A1, therefore, a method for adjusting a

Robot, in which this first in a so-called Urjustage brought into a precisely defined Justagepose and in these relative positions of

Characteristics are precisely measured on the robot links. to

Restoration of the adjustment position after a loss of adjustment, the robot is then moved until these relative positions are reached again.

Due to the necessary precision, this method requires a high expenditure on equipment.

An object of an embodiment of the present invention is to improve an adjustment, control and / or monitoring of a robot, in particular to reduce a measuring expenditure on equipment for the adjustment and / or to increase speed and / or precision of adjustment, control and / or monitoring , This object is achieved by a method having the features of claim 1. Claims 13-15 provide (robot) control

Computer program product for carrying out a method described here or a robot arrangement with a controller described here under protection. The subclaims relate to advantageous developments.

According to an embodiment of the present invention, the robot has a

Robot arrangement at least or exactly three, four, six or seven, joints or (motion) axes, in particular rotary or rotary joints or axes and / or translational or linear joints or axes, and / or, in particular electromotive, drives for moving or adjusting the joints or axes. In one embodiment, the robot is a redundant and / or delta robot. In such robots, the present invention can be used with particular advantage.

In general, a so-called forward transformation transforms T positions

[qi, ... qi _< ] ^T = q of the k joints or axes of the robot at positions P of a robot-fixed reference, in particular a TCP, of the robot or conversely a backward transformation TT positions P of the robot-fixed reference to q its joints:

P = T (q) (1)

q = T ¹ (P) (1 ')

A position, in particular a robot-fixed reference, within the meaning of the present invention can in one embodiment (only or also) a one, two or

three-dimensional, in particular Cartesian, location or distance and / or (only or also) a one-, two- or three-dimensional orientation (the robot-fixed reference) (relative to a reference or coordinate system), in particular be, for example, three location coordinates (P _x , P _y , P _z ) and / or have three orientation angles, in particular be determined thereby. One, in particular predetermined, path of a robot-fixed reference comprises in one embodiment a (Ab) sequence of such positions, it may be defined or formed in particular by this. In one embodiment, a (predetermined) path g may be an (ab) sequence of Cartesian locations of a robot-fixed reference point and / or an (ab) sequence of three-dimensional Orientations of a robot-fixed reference coordinate system, in particular in the case of stationary coordinate system origin, ie a mere reorientation in the case of fixed TCP, include, in particular.

When driving off a (predetermined) path g with the robot-fixed reference, their positions have a predetermined or predetermined by the path

geometric context on:

P = 9 (λ) (2) with the orbital parameter, for example, a, in particular generalized, track length, λ. This can now be exploited to the robot, in particular a

Detecting means of the robot for determining joint positions to adjust, in particular a (constant) offset, in particular offset vector, Q des

To detect detection means. Because for each determined by the detection means hinge position q, when driving off the predetermined path g with the robot-fixed reference applies:

Pi = 9 (k, P) = T (q, + Q) (3) with the (geometric) parameters p for (geometrically) defining the given path.

In one embodiment, if the predetermined path is a straight line in the working space of the robot, which travels with a robot-fixed point, this path is characterized in particular by three coordinates (P _x , P _y , P _z ) of a path point and two polar angles (a, β). determining the direction v of the web in the workspace:

Pi = g (, P) = [Px, Py, Pzf + λ *, · ν (α, β); p = [P _x , P _y , P _z , a, β] (4)

A given circular path is defined, for example, by three coordinates of a center point as well as a normalized vector normalized to length 1 and the radius or a normal vector having the length of the radius, a mere Reorientation for example by appropriate Euler or gimbal angle or the like.

With a sufficient number n of determined joint positions qi, (3) results in a (above) certain equation system for the unknowns λ \, p and Q. Correspondingly, based on determined joint positions qi, the offset Q of the

Detection means, in particular by means of a

Optimization method that optimizes a quality criterion G, which in particular of deviations between positions Pj of the robot-fixed reference, which are determined on the basis of the determined joint positions qi, and positions on the

depend on predetermined path, this particular can minimize, for example:

G = £ [g (/ "P) -T (qi ₊ Q)] - [g (4, p) -T (qi ₊ Q)] (5)

i = 1

Accordingly, according to an embodiment of the present invention comprises

Method for adjusting the robot's steps:

- Determining a plurality of joint positions q, when driving off a

predetermined path g with a robot-fixed reference by means of a

adjusting detection means; and

- Determining an offset, in particular offset vector, Q of this detection means based on the determined joint positions.

As a result, in one embodiment, the adjustment can be improved, in particular an apparatus-related measurement effort for the adjustment can be reduced and / or its speed and / or precision can be increased, thus improving control and / or monitoring of the adjusted robot in a further development.

In one embodiment, the predetermined path is force, in particular yielding (keits) regulated and / or hand-held, in particular with a predetermined contact force and / or by manually guiding or actuating the robot-fixed reference on and / or departing. As a result, in one embodiment, the adjustment can be further improved, in particular an expense for the adjustment (further) reduced and / or its speed and / or precision (further) increased and thus in a development, a control and / or monitoring of the adjusted robot (further) be improved. In one embodiment, the predetermined path with constant orientation of the robot-fixed reference, in particular in a Cartesian working space of the robot or a limb, in particular a base, of the robot against which the robot-fixed reference by moving the axes to be adjusted of the robot is movable, traversed, for example by actuation of corresponding axes of a delta robot or the like. As a result, in one embodiment, a (tool) flange of the robot to which the reference may be attached, this or a congruent and / or offset parallel path from, so that advantageously a position of the robot-fixed reference relative to the flange can be disregarded , In one embodiment, the

predetermined path with varying orientation of the robot-fixed reference, in particular in a Cartesian working space of the robot or relating to a member, in particular a base of the robot, against which the robot-fixed reference by adjusting the axes to be adjusted of the robot is movable, abfahren, in particular a stationary TCP only reoriented, whereby a particularly compact path can be realized.

In one embodiment, the web is or is by a guide, in particular a forced operation, for the robot-fixed reference, in particular form or

contact-locking and / or non-positive, in particular magnetic, predetermined, which have metal in a further development, in particular may consist thereof.

Additionally or alternatively, the guide is in a development temporarily or stationary stationary to a member, in particular a base of the robot arranged against which the robot-fixed reference is movable by adjusting the axes to be adjusted of the robot or is moved, in particular non-destructive detachably or permanently or non-destructively detachable relative to the member, in particular at this or spaced therefrom, fixed or fixed.

By a, in particular positive and / or non-positive, in particular contact and / or magnetic, and / or at least partially metallic, (forced) leadership can in one embodiment, the adjustment further improved in particular their

Speed and / or precision (further) can be increased. By a stationary attachment, the precision (further) can be increased in one embodiment, reduced by a temporary attachment a restriction of the working space of the robot after the adjustment, in particular be avoided. Accordingly, in one embodiment, the guide is removed again after adjustment.

In one embodiment, the guide locks, in particular positive and / or non-positive, in particular by contact or contact and / or magnetic, at least one and / or at most two translational or spatial degrees of freedom and / or no rotational or orientation degree of freedom of the robot-fixed reference one or two-sided or puts only their Cartesian place with a

(Path parameter) degree of freedom λ fixed. As a result, in one embodiment, the

Departure of the web and / or the solution of the (over) certain equation system or optimization problem can be improved. Similarly, in one

Execution of the guide three translational or spatial degrees of freedom on one or two-sided lock, so that the TCP is only reoriented in Orttfester position or can be.

Additionally or alternatively, in one embodiment, the web may be a straight line in the

Working space of the robot, in particular his. In this way, in one embodiment, the trajectory of the trajectory and / or the solution of the (over) certain equation system or optimization problem (further) can be improved.

Additionally or alternatively, the web is in one embodiment by a profile, in particular one, in particular rectangular, angle profile or two against each other, in particular at right angles, beveled surfaces, in particular form or

contact-connected, predetermined. As a result, in one embodiment, the departure of the web can be improved.

In one embodiment, the robot for adjusting, in particular for an initial or pre-adjustment and / or for traversing the predetermined path, in one

Execution by hand method, (in each case) into one, in particular the same or different, predetermined output poses (s) (pre) positioned in a

Training, in particular for the initial or pre-adjustment, by a, in particular multi-part, robot-proof marking, for example, mutually associated brands of mutually movable robot members, may be specified. As a result, in one embodiment, the adjustment, in particular their

Convergence and thus their speed and / or precision can be improved. In one embodiment, the detection means is based on this predetermined

Starting poses, in particular coarse, initial or pre-adjusted.

In one embodiment, the guide is only after starting the predetermined

Starting poses arranged or fixed, which in particular the approach of Ausgangsspose and so the speed and / or precision of the adjustment

can improve.

In one embodiment, the robot-fixed reference is stationary or permanent or non-destructive releasably or (only) temporarily or non-destructively releasably attached to a, preferably end or distal, member of the robot. Additionally or alternatively, the robot-fixed reference may comprise metal, in particular consist thereof, and / or have a completely or partially spherical outer contour, in particular for contact with the (positive) guide.

By attachment to a distal robot member, in particular

Tool flange, in one embodiment can be adjusted all the axes of the robot that move this member. By a stationary attachment, the precision (further) can be increased in one embodiment, reduced by a temporary attachment restriction of the robot after the adjustment, in particular be avoided. Accordingly, in one embodiment, the reference is removed again after adjustment. By an at least partially metallic and / or spherical reference, the adjustment can be further improved in one embodiment, in particular their speed and / or precision (further) can be increased.

In particular, an at least partially spherical (outer) contour of the reference can compensate orientation changes around its center, so that in one embodiment it can determine the position of the reference.

In one embodiment, at least two, in particular base-next and / or rotational and / or translational, preferably all, axes of the robot, in particular together or together, in particular synchronously adjusted. As a result, in one embodiment, the adjustment can be further improved, in particular its speed and / or precision (further) can be increased. Similarly, in another embodiment, only certain axes of the robot and / or axes of the robot can be adjusted sequentially. In one embodiment, the, in particular multi-part, detection means detects the Achsbzw. Joint positions of one or more, preferably all, axes or joints of the robot, in particular incremental or absolute, or is set up for this purpose. In one embodiment, the detection means may include one or more relative,

In particular, incremental encoders, in particular for the detection of

Have relative positions, in particular relative position changes, one or more to be adjusted axes of the robot, in particular consist thereof. For adjusting such detection means, the present invention may be used to particular advantage without being limited thereto. Accordingly, the detection means may also comprise, for example, resolvers or the like. According to equation (3), each determined joint position q _; d equations, where d denotes the dimension of a position given by the orbit, in cartesian locations, for example, three equations (d = 3). N denotes the number of joint positions determined, D the dimension or number of parameters p for (geometrically) defining the predetermined path, and k the number of axes (or joints) of the robot to be adjusted or the dimension of a determined joint position and thus of the to determined offsets Q, the (d + k + n-1) unknown p, Q and λ _\ = _2, 3,, " _{n give} the (dn) equations a particular system of equations, if n = (D + k - 1) / (d - 1) (6) or an overdetermined system of equations, as soon as n> (D + k-1) / (d-1) (6 '), for example in a six-axis robot (k = 6) Straight line in the

Working space (D = 5) and Cartesian locations (d = 3) n = or> 5. Advantageously, only (n - 1) path parameter values λ _{1 = 2} , _3, ... _{n are} unknown if the parameters p used for a joint position λι, for example, one of the determined positions as a straight line point.

Accordingly, in one embodiment, the number n of the determined joint positions is equal to or, preferably, greater than (D + k-1) / (d-1) with the minimum

Parameter number D for geometrically defining the predetermined path, the number k of the joints of the robot to be adjusted and the dimension d of a position of the robot-fixed reference predetermined by the path, preferably at least twice as large (n> 2 (D + k-1) / ( d - 1)), preferably at least three times as large (n> 3 (D + k-1) / (d-1)). Additionally or alternatively, in one

Execution number n of the determined joint positions at least 10, in particular at least 15, in particular at least 30. As a result, in one embodiment, the adjustment, in particular their convergence and thus their speed and / or precision, (further) can be improved.

In one embodiment, a starting value of the optimization method is determined on the basis of the predefined starting position. As a result, in one embodiment, the

Adjustment, in particular their convergence and thus their speed and / or precision, (further) to be improved.

If the robot is adjusted, in particular determines the offset of the detection means in a manner described here, in one embodiment by means of

Detection means are determined based on or taking into account the determined offset Solloder actual joint positions of the robot and / or setpoint or actual positions of the or another robot-fixed reference, in particular according to

Equations (1), (1 ').

According to one embodiment of the present invention, a (robot) controller, in particular hardware and / or software, in particular program technology, is set up to carry out a method described here and / or has the following:

Means for determining one or the plurality of joint positions when driving one or the predetermined path with one or the robot-fixed reference by means of one or the detection means; and Means for determining an offset or the offset of the detection means on the basis of the determined joint positions.

In one embodiment, the controller or its agent has force-controlled and / or hand-guided means for starting and / or stopping the predetermined path and / or with constant or varying orientation of the robot-fixed reference.

In one embodiment, the robot arrangement has a guide or guide for the robot-fixed reference, by which the path is or is predetermined or which is set up or used for this purpose.

In one embodiment, the controller or its agent has:

Means for (pre) positioning the robot for adjusting, in particular for an initial or pre-adjustment and / or for traversing the predetermined path, in a, in particular by a robot-fixed marking, predetermined output poses; and or

Means for adjusting at least two, in particular base-next and / or rotational and / or translatory, axes of the robot, in particular together; and or

Means for optimizing a quality criterion, preferably of deviations between positions of the robot-fixed reference, based on the determined

Joint positions are determined, and positions on the given path depends, by means of an optimization method for determining the offset; and / or means for determining the offset of the detection means according to one here

described method; and means for determining at least one desired or actual joint position and / or position of a robot-fixed reference on the basis of the determined offset. A means in the sense of the present invention may be designed in terms of hardware and / or software, in particular a data or signal-connected, preferably digital, processing, in particular microprocessor unit (CPU) and / or a memory and / or bus system or multiple programs or program modules. The CPU may be configured to execute instructions implemented as a program stored in a memory system, to capture input signals from a data bus, and / or Output signals to a data bus. A storage system may comprise one or more, in particular different, storage media, in particular optical, magnetic, solid state and / or other non-volatile media. The program may be such that it is capable of embodying or executing the methods described herein so that the CPU may perform the steps of such methods and thus, in particular, the robot or its

Adjust detection means and / or control and / or monitor the robot.

In one embodiment, one or more, in particular all, steps of the method are completely or partially automated, in particular by the controller or its (e) means.

Further advantages and features emerge from the subclaims and the exemplary embodiments. This shows, partially schematized ,:

1 shows a robot arrangement with a robot, a guide and a controller according to an embodiment of the present invention; and FIG. 2 shows a method for adjusting the robot according to an embodiment of the present invention.

Fig. 1 shows a robot assembly having a six-axis robot 1, a guide in the form of a rectangular metal profile 5 and a controller 3 according to an embodiment of the present invention, which performs a method for adjusting the robot 1 described below according to an embodiment of the present invention.

In a first step S10, the robot 1 by hand method is relatively coarse in a Ausgangspose q-ι = [q ^, q _2, i, q _3, i, q _4, i, qs.i, qe, i] ^T positioned which is predetermined by mutually aligned marking elements or features of a robot-fixed marking on mutually movable members of the robot 1, of which in Fig. 1 by way of example two mutually associated marking elements 6 are indicated. On the basis of this predetermined output pose qi, the controller 3 adjusts

Detecting means with relative encoders 2 for determining the joint positions initially coarse, they can, for example, zeros or occupy predetermined for the initial poses of non-zero initial values.

In a step S20, as a preferred embodiment of a robot-fixed reference, a metal ball 4 is firmly fixed to a distal tool flange 11 of the robot 1 and the metal profile 5 is fixed relative to the proximal base 12 of the robot 1.

In the output pose qi, the metal ball 4 or its center, which is activated as TCP, is positioned in the Cartesian position P = [P _{1 iX} , Pi, _y , Pi, _z ] ^T , which is indicated by dashed lines in FIG.

Subsequently, in a step S30, the metal ball 4 fastened to its tool flange 11 is guided along the guide 5 or the metal profile 5 in the case of the robotic robot 1, so that its center or the TCP of the robot has a straight line that is predetermined by the metal profile 5 g descends in the working space of the robot.

In this case, in certain, in particular equidistant, time intervals, n joint positions q; = [Q ij q _2, i, q ^ i, q _4, i, sensing amplifier QS j, qe, i] ^T determined, one of which by way of example one is illustrated in FIG. 1.

In a step S40, the controller 3 determines estimated values for the positions Pi,..., P _n by means of the detection means coarsely adjusted on the basis of the output positions determined by the marking 5 with the relative rotary sensors 2 in accordance with equation (1).

Then it solves the optimization problem according to equation (5). In this case, it uses as starting values of the iterative optimization for the parameters p = [P _x , P _y , Ρ _ζ , α, β] ^T for specifying the predetermined path g as point [P _x , P _y , P J ^T for determining the path g the estimated value for the position P-ι and for the direction v (a, ß) for determining the path g the difference vector of two estimated values, for example ν (α, β) = P _n -Pi, and for the path parameter values λ, the distances between estimated positions λι = 0 and = | P, - P _M | for 2 <i <n and for the offset Q zero or the initial assignment of the detection means with the relative rotary 2.

It then stores the thus determined offset Q from. Thereby, the detecting means is adjusted with the relative rotary 2 and the robot 1 can be controlled and / or monitored in a determined by the detecting means with the relative rotary 2 based on or taking into account this offset joint positions in a step S60, after in step S50 metal ball 4th and metal profile 5 was removed. If the adjustment or the relationship between the relative mechanical position of the joints or members of the robot and the values determined by the detection means with the relative rotary 2 are lost, the robot 1 or its detection means with the relative rotary 2 in situ again simply, quickly and accurately be adjusted by the method S10 - S50 described above is performed again.

If the orientation of the metal ball 4 in the Cartesian working space of the robot 1 does not change when the web g is traversed, then instead of the (estimated values for) positions Pi,..., P _{n of} the center of the ball 4 (estimates for the)

Positions of the tool flange 11 itself used or the TCP are placed in this, so that the position of the ball relative to the tool flange can be disregarded. This can be advantageously used, for example, when adjusting the proximal axes 1-3 of a delta robot.

Although exemplary embodiments have been explained in the foregoing description, it should be understood that a variety of modifications are possible. Thus, in particular, the pose used in step S10 with the aid of the marking elements 6 for pre-adjustment and the pose from which the straight line g is traversed may be different poses. In particular, it is possible to bring the metal ball 4 to the metal profile 5 only after the coarse or pre-adjustment of step S10, in particular by manual methods, and then to begin its movement. It should also be noted that it is the exemplary

The explanations are merely examples of the scope of protection

Applications and the structure should in no way limit. Rather, the expert is by the preceding description a guide for the

Implementation of at least one exemplary embodiment given, wherein various changes, in particular with regard to the function and arrangement of the described components, can be made without departing from the scope, as it is apparent from the claims and these equivalent

Feature combinations results.

LIST OF REFERENCES

1 robot

2 relative rotary encoders (detection means)

3 (robot) control

4 metal ball (robot-proof reference)

5 leadership

6 mark

11 tool flange

12 base

g given course

joint position

Claims

claims

1. A method for adjusting a robot (1), comprising the steps of:

Determining (S30) a plurality of joint positions (q 1, q ₂ , i, q ₃ .i, q4j, qsj. Qe.) When a predetermined path (g) is traveled with a robot-fixed reference (4) by means of a detection means (2) ; and

- determining (S40) an offset of the detection means on the basis of the determined joint positions.

2. The method according to claim 1, characterized in that the predetermined path (g) force-controlled and / or hand-held and / or with constant or varying orientation of the robot-fixed reference (4) on and / or driven off.

3. The method according to any one of the preceding claims, characterized in that the web (g) by a guide (5) for the robot-fixed reference (4) is predetermined.

4. The method according to the preceding claim, characterized in that the guide (5) comprises metal and / or temporarily or stationary stationary to a member (12) of the robot is arranged opposite to the robot-fixed reference (4) by adjusting the to be adjusted Axes of the robot (1) is movable.

5. The method according to any one of the preceding claims, characterized in that the web has a straight line (g) in the working space of the robot and / or by a profile (5) is predetermined.

6. The method according to any one of the preceding claims, characterized in that the robot for adjusting, in particular traversing the predetermined path, in a, in particular by a robot-proof mark (6), predetermined output poses is positioned.

7. The method according to any one of the preceding claims, characterized in that the robot-fixed reference (4) is temporarily or stationarily attached to a member (1 1) of the robot and / or metal and / or has an at least partially spherical outer contour.

8. The method according to any one of the preceding claims, characterized in that at least two, in particular base-next and / or rotational and / or translational, axes of the robot, in particular together, are adjusted.

9. The method according to any one of the preceding claims, characterized in that the detection means comprises at least one relative encoder (2) for detecting relative positions of an axis to be adjusted of the robot.

10. The method according to any one of the preceding claims, characterized in that the number of determined joint positions equal to or greater than (D + k - 1) / (d - 1) with the minimum number of parameters D for geometrically defining the predetermined path (g), the number k of the joints of the robot to be adjusted and the dimension d of a position (P,) of the robot-fixed reference (4) predetermined by the web is and / or is at least 10.

11. Method according to one of the preceding claims, characterized in that for determining the offset by means of an optimization method a

Quality criterion is optimized (S40), preferably by deviations between positions (P,) of the robot-fixed reference (4), based on the determined

Joint positions (q ^, q _2, i, qs, i, q4, i, qs, i, qej) are determined, and positions on the given path (g) depends.

12. A method for controlling and / or monitoring a robot (1), with the

steps:

- Detecting the offset (S40) of the detection means (2) according to a method according to one of the preceding claims; and

Determining (S60) at least one desired or actual joint position and / or position of a robot-fixed reference on the basis of the determined offset.

13. Control device (3) for adjusting, in particular for controlling and / or monitoring, a robot (1) which is set up to carry out a method according to one of the preceding claims and / or comprises:

Means for determining a plurality of joint positions (qy, q _2, i, sensing amplifier QS j, q _4, q _5ii, q _6, i) when traveling along a predetermined path (g) with a robot fixed reference (4) by means of a detecting means (2); and Means for determining an offset of the detection means based on the determined joint positions.

14. Robot arrangement with a robot (1) and a controller (3) for adjusting, in particular for controlling and / or monitoring, the robot according to one of the preceding claims.

A computer program product having a program code stored on a computer readable medium for carrying out a method according to any one of the preceding claims.