METHOD FOR COMPENSATING FOR A DEVIATION IN AN OPERATING POINT

Abstract

The invention relates to a method for compensating for a deviation in an operating point (1) of a manipulator (2) during the processing of a workpiece (3) via an end effector (4) on the manipulator (2); wherein a command sequence is processed for controlling the manipulator (2) for the purpose of processing the workpiece (3), and a piece of setpoint position information (5) corresponding to a setpoint position (6) is generated based on the command sequence, the operating point (1) of the manipulator (2) being set based on said piece of setpoint position information (5); wherein the piece of setpoint position information (5) is processed using a compensation parameter set (7) which is related to the piece of setpoint position information (5), for ascertaining a compensation value (8), and the piece of setpoint position information (5) is adjusted according to the compensation value (8) for compensating for a deviation (9) between an actual position (10) of the operating point (1) and the setpoint position (6). The invention is characterized in that the actual position (10) is measured during the processing of the workpiece (3); in that a correction value (12) is ascertained, based on a comparison between the measured actual position (10) and the setpoint position (6); and in that the compensation parameter set (7) is adjusted during the processing of the workpiece (3) for reducing the deviation (9), based on the correction value (12).

Description

The present invention relates to a method for compensating for a deviation in an operating point of a manipulator according to the preamble of claim 1.

When processing workpieces by means of an end effector on a manipulator, it is frequently required to continually move and thus position the operating point, which may also be referred to as the tool center point (TCP), of the manipulator during the machining process. In particular, if the manipulator is a multiaxis robot, the control of the manipulator is based on calculations performed in an NC controller, which include trajectory planning and interpolation, according to which the corresponding controls, for example, of the axes of the multiaxis robot, take place.

The calculations performed by the NC controller are based on various physical parameters of the manipulator, for example, lengths, weights, and other variables which are relevant to the calculations. However, in practice, these parameters are not ideally constant, but may change, for example, as a function of the instantaneously prevailing temperature. If these changes are not taken into account when controlling the manipulator, a deviation results between the actual position of the operating point of the manipulator and the setpoint position of the operating point of the manipulator according to the calculation. Here and below, the term “NC controller” is to be understood in the broad sense, thus meaning a controller which comprises some or all functionalities of a PLC (programmable logic controller).

A method and a device for compensating for a temperature-dependent change in position on a machine tool is known from the related art and specifically from the published patent application DE 10 2010 003 303 A1. Here, an NC controller calculates setpoint values for various linear axes of the machine tool. These setpoint values take into account a temperature compensation which is based on instantaneous temperature measurements and which ascertains a compensation value for the individual setpoint values of the respective linear axes and applies it to them. However, the compensation values are not only temperature-dependent; they also depend on the position of the linear axes. Thus, it is also taken into account that for precise temperature compensation, a position-dependent component must also be taken into account. To be able to perform the compensation in a timely manner specifically in the case of rapid changes in position, the compensation is determined via a system of fixed coefficients, which are applied to the measured temperatures and the axis positions, according to the determination via the NC controller.

However, the related art has the disadvantage that, despite the attempt to compensate by taking into account the temperature, whether as a function of position or independently of position, a deviation may result between the setpoint position and the actual position. This may be because the prevailing boundary conditions, for example, during an initial calibration carried out simultaneously in the laboratory for ascertaining the above coefficients, no longer exist. Thus, for instance, a specific temperature distribution may exist along a route of the manipulator, which cannot be sufficiently differentiated even by means of multiple distributed temperature sensors; or process forces acting on the manipulator may generate a significant influence on the position of the manipulator. Deviations which have their cause herein cannot be compensated for under the approach from the related art.

Against this background, the object of the present invention is to provide an improved method for compensating for a deviation in an operating point of a manipulator during the processing of a workpiece via an end effector on the manipulator, said method being able to respond dynamically, via compensation, to deviation effects occurring during the processing.

The aforementioned problem is solved via a method for compensating for a deviation in an operating point of a manipulator according to the preamble of claim 1, via the features of the characterizing portion of claim 1.

What is essential to the invention is the recognition that by measuring the actual position of the operating point during the processing of the workpiece, and thus after compensation has been performed, the compensation parameters may be dynamically adjusted in the case of a deviation which still exists, so that both disturbance variables which are not known beforehand and changing relationships of the known disturbance variables may be taken into account and compensated for. In this respect, feedback takes place in real time for improving the compensation parameters.

In this case, the preferred embodiment of subclaim 2 relates to the possibility of carrying out this adjustment of the compensation parameters in the interval in which the NC controller carries out the interpolation. This enables a very rapid response to deviations which occur.

NC controllers exist which provide special programmer interfaces via which proprietary software modules which are defined, for example, by operators of the

NC controller or integrators, can intervene into the processing of the NC controllers. Subclaims 3 to 5 provide for allowing steps of the provided method to run in an extension software module which interacts with the NC controller via such an interface. In this case, this extension software module may be called or initiated via this interface and may for its part access state information about the NC controller or trigger actions of the NC controller.

A preferred option for measuring the actual position in the ongoing processing is provided via the provision of an optical sensor as a laser tracker, as described by subclaim 6.

Subclaim 7 provides for an additional refinement, according to which the parameters used for compensation are subdivided to the extent that only certain parameters are dynamically adjusted, and the remaining parameters remain constant. In this way, a more precise depiction of the physical effects underlying the deviation is possible.

Subclaim 8 relates to the measurement of an environmental variable as the basis for the compensation, wherein subclaim 9 specifically provides for taking into account a pressure as an environmental variable for the compensation, and this pressure may act specifically on a weight compensation system and may be measured there. Subclaims 10 and 11 relate to the measurement of an actual temperature as the environmental variable and compensation for temperature-related deviations. Subclaim 12 in turn provides for detection of a process force acting on the end effector for compensating for corresponding effects.

Finally, subclaims 13 to 15 relate to the case in which the manipulator is a multiaxis robot, and specifically to embodiments tailored to this variant.

Additional details, features, objectives, and advantages of the present invention will be described below, based on the drawing of only one preferred exemplary embodiment. The following are shown:

FIG. 1 shows a schematic representation of an arrangement including an end effector on a manipulator for processing a workpiece for carrying out the method according to the proposal, and

FIG. 2 shows a schematic diagram of the interface between an NC controller and an extension software module for carrying out the method according to the proposal.

The method according to the proposal is used for compensating for a deviation in an operating point 1, also referred to as a tool center point (TCP), of a manipulator 2 during the processing of a workpiece 3 via an end effector 4 on the manipulator 2. The corresponding arrangement is depicted in FIG. 1. In the depicted exemplary embodiment, the end effector 4 is a welding device 4a, and the workpiece 3 is an aircraft structural component 3a. In the method according to the proposal, a command sequence is processed for controlling the manipulator 2 for the purpose of processing the workpiece 3. On the one hand, this command sequence may be an NC program which is made up of individual NC commands. However, it may be a command sequence which was first created from the NC commands of the NC program. This creation may comprise various calculation steps, in particular kinematics transformation, interpolation, or trajectory planning. The processing takes place periodically in an NC controller 11a which is further described below.

According to the proposal, a piece of setpoint position information 5 corresponding to a setpoint position 6 is generated based on this command sequence. The setpoint position 6, which is a setpoint position 6 of the operating point 1, thus results from the command sequence, either directly or via one or multiple calculation steps as described above by way of example. The piece of setpoint position information 5 corresponds to the control data which were generated from the processing of the command sequence and via which the manipulator 2 is to be controlled, so that the operating point 1 is at the setpoint position 6. In this case, in addition to a pure position specification in three dimensions, both the operating point 1 and the setpoint position 6 may also comprise an orientation specification, for example, also in three dimensions. A differentiation will be not made below between these two possible components, and possibly additional components, of the operating point 1 or the setpoint position 6. The piece of setpoint position information 5 may accordingly also be a time sequence or a profile of control data via which the manipulator 2 is controlled.

According to the proposal, the operating point 1 of the manipulator 2 is thus set based on the piece of setpoint position information 5. Relationships which are ascertained in advance via measurement, and which establish a deviation of the operating point 1 from the setpoint position 6 as a function of various boundary conditions, are now taken into account for compensation. According to the proposal, for this purpose, it is provided that the piece of setpoint position information 5 is processed using a compensation parameter set 7 which is related to the piece of setpoint position information 5, for ascertaining a compensation value 8. This processing may in principle be any arbitrary calculating operation. According to the proposal, the piece of setpoint position information 5 is adjusted according to the compensation value 8 for compensating for a deviation 9 between an actual position 10 of the operating point 1 and the setpoint position 6. Thus, the compensation value 8 is to reduce this deviation 9 with respect to the situation in which the compensation value 8 is not taken into account, if possible. In this case, the compensation value 8, as well as the piece of setpoint position information 5, may be provided to the manipulator 2, so that this processing takes place in the manipulator 2. The above determinations with respect to the dimensionality of the setpoint position also apply analogously to the actual position 10. Alternatively or in addition, the adjustment of the piece of setpoint position information may also take place in a control device 11 which is separate from the manipulator 2, so that the manipulator 2 is controlled using only the piece of setpoint position information 5 which is adjusted in this sense. This control device 11 may in principle be any arbitrary data processing device and in particular a computer such as an industrial PC or the like.

However, as already described above, here, it may be that this adjustment of the piece of setpoint position information 5 according to the compensation value 8 is not sufficient to completely compensate for the deviation 9. In fact, it regularly happens that not all effects which occur during the processing of the workpiece 3 and which may result in the deviation 9 may be completely detected via a previously ascertained compensation parameter set 7.

The method according to the proposal is therefore characterized in that the actual position 10 is measured during the processing of the workpiece 3; in that a correction value 12 is ascertained, based on a comparison between the measured actual position 10 and the setpoint position 6; and in that the compensation parameter set 7 is adjusted during the processing of the workpiece 3 for reducing the deviation 9, based on the correction value 12. In this way, the mechanism for compensating for the deviation 9 may still be further optimized during the processing of the workpiece 3 in such a way that it may even completely disappear. By adjusting the compensation parameter set 7, the compensation value 8 also changes correspondingly, resulting in a reduction in the deviation 9.

It is preferred that the piece of setpoint position information 5 is generated by an NC controller 11a, based on the command sequence, in particular after an interpolation and/or a transformation. This NC controller 11a may be a separate device or, as depicted in FIG. 2, software which possibly runs with other software on a computing device, here, the control device 11. To be able to carry out the above-described ongoing adjustment during the processing in a particularly responsive manner, it is preferred that the piece of setpoint position information 5 is generated by the NC controller 11a at an interpolation rate and that the adjustment of the compensation parameter set 7 for reducing the deviation 9 takes place at the interpolation rate. The term “interpolation rate” may be understood to mean the clock period at which the NC controller 11a carries out the interpolation.

It is particularly advantageous if the method according to the proposal may be applied on an NC controller 11a which is already known. This is particularly simple if the NC controller 11a, which is frequently obtained from a third-party vendor, has a connection interface 14 for incorporating proprietary software. Therefore, it is preferred that the processing of the piece of setpoint position information 5 using the compensation parameter set 7 related to the piece of setpoint position information 5 for ascertaining the compensation value 8, and the adjustment of the compensation parameter set 7 for reducing the deviation 9, take place in a running extension software module 13, said extension software module 13 communicating in terms of data technology with the NC controller 11a via the connection interface 14 of said NC controller. This connection interface 14 may be a pure programming interface.

In this case, the execution of this extension software module 13 may be under the control of the NC controller 11a, which, for instance, may be provided in that corresponding access points of the extension software module 13 are passed to the NC controller 11a as a recall function. It is thus preferred that the connection interface 14 calls the extension software module 13 in terms of data technology based on execution events of the NC controller 11a. However, the NC controller 11a may also trigger actions in the extension software module 13 by means of another mechanism in terms of data technology by, for example, enabling semaphores. Reciprocally, it is also preferably provided that the extension software module 13 may call functions of the NC controller 11a in terms of data technology via the connection interface 14.

Via this connection interface 14, it may also be made possible for the extension software module 13 to access the required or helpful information for ascertaining the compensation value 8. Thus, the extension software module 13 preferably receives state information about the manipulator 2, which may in particular comprise the actual position 10 and/or the piece of setpoint position information 5 and/or the setpoint position 6, via the connection interface 14. Alternatively, and as depicted in FIG. 2, the extension software module may also receive the actual position 10 independently of the NC controller 11a and thus independently of the connection interface 14.

Optical detection is particularly suitable for reliably determining the actual position 10 of the operating point 1. As shown in FIG. 1, it is therefore preferred that the actual position 10 is measured via an optical position sensor 15 during the processing of the workpiece 3. In this case, according to the depiction in figure 1, a laser tracker 15a is particularly suitable, wherein a plurality of laser trackers 15a may also be used for measuring the actual position 10. The laser tracker 15a targets a reflector 16 on the manipulator 2. The precision of the adjustment of the compensation parameter set 7 may be may be increased by increasing the number of different measurements carried out by the optical position sensor 15. As also depicted in FIG. 1, it is therefore preferred that the optical position sensor 15 performs a measurement of the actual position 10 at a plurality of processing positions 17 on the workpiece 3, said processing positions 17 being successively controlled by the manipulator 2 for processing via the end effector 4. In FIG. 1, these processing positions 17 are indicated by the corresponding position of the reflector 16 at the respective processing position 17.

In such a case, the compensation parameter set 7 may be adjusted based on the correction value 12 from multiple comparisons between the measured actual position 10 and the setpoint position 6. Here, a reference position 17a for the optical position sensor 15 is also defined on the workpiece 3. It is furthermore preferred that the optical position sensor 15 is connected in terms of data technology to the NC controller 11a and/or to the extension software module 13 via running measuring software 18.

During the determination of the compensation parameter set 7, a differentiation may already be made between, on the one hand, those parameters which, being type-related or design-related, may be viewed as constant, and on the other hand, those parameters which may differ from one manipulator to the other. The former parameters may also be viewed as fixed model parameters which fundamentally define the computational model in general, and the latter parameters may be viewed as parameters which this model then populates with specific values. Thus, it is preferred that the compensation parameter set 7 comprises an in particular manipulator type-related model parameter set 19, here, corresponding to the former parameters, and a physical data set 20, corresponding to the latter parameters. The physical data set 20 may comprise mass information (i.e., in the sense of a weight), and additionally or alternatively, center of mass information, relating to the manipulator 2. Preferably, only the model parameter set 19 is adjusted based on the correction value 12 for reducing the deviation 9.

According to a preferred embodiment, the compensation parameter set 7 is also used to compensate for the influence of a measurable environmental variable on the actual position 10 of the operating point 1. Thus, in addition to the piece of setpoint position information 5, this measured environmental variable is also taken as a basis for the compensation. In this case, the term “environment” is presently to be interpreted broadly and comprises any arbitrary physical variables, in particular also variables occurring and being measurable on or in the manipulator.

Therefore, it is preferred that during the processing of the workpiece 3, at least one environmental variable is measured, and that the compensation parameter set 7 comprises an environmental compensation parameter which is related to the respective measured environmental variable, the respective measured environmental variable being processed using said environment compensation parameter for ascertaining the compensation value 8. The measurement of the at least one environmental variable preferably takes place via an environmental sensor arrangement; according to the example of FIG. 1, via a temperature sensor arrangement 21a.

One preferred variant provides that the at least one environmental variable comprises a pressure 22a. This pressure 22a may be a pressure 22a acting on the manipulator 2, and here in particular acting on a weight compensation system of the manipulator 2.

According to another preferred variant, the at least one environmental variable comprises an actual temperature 22b. This actual temperature 22b may in particular be an actual temperature 22b at the manipulator 2. In this case, this measured actual temperature 22b may be due both to warming from the environment and to warming from the operation of the manipulator 2. Preferably, the actual temperature 22b is thus influenced by ambient heat and by self-heating of the manipulator 2.

One preferred embodiment provides here that the environmental compensation parameter comprises a material-dependent temperature expansion coefficient and/or an in particular axis-related temperature-deflection coefficient and/or an in particular axis-related length parameter. Furthermore, it is preferred that the environmental compensation parameter is related to the measured actual temperature 22b.

Thus, the aforementioned length parameter may comprise a measured, estimated, or calculated length specification, which may relate in particular to a distance between two axes of the manipulator 2. The temperature expansion coefficient may be an expansion coefficient which is related to a material of the manipulator 2. The temperature-deflection coefficient may also relate to a deflection of the path between two axes, analogously to the aforementioned length parameter. The environmental compensation parameter may also comprise a weighting of these individual parameters and coefficients.

An additional preferred variant provides that the at least one environmental variable comprises a process force 22c acting on the manipulator 2 and/or on the end effector 4. Such a process force 22c is to be understood to mean a force which is exerted by the end effector 4 in particular on the workpiece 3 during the processing, said force then also being correspondingly absorbed by the manipulator 2 and, for example, possibly resulting in deflections there.

The manipulator 2 is preferably a multiaxis robot 23, as depicted in FIG. 1. Correspondingly, it is preferred that the piece of setpoint position information 5 comprises a piece of axis information 5a about the multiaxis robot 23. Such a piece of axis information 5a is thus to be understood to be a piece of information which relates to one or multiple axes of the multiaxis robot 23. In particular, it may be a piece of information about a setpoint position of an axis of the multiaxis robot 23. Specifically, the piece of axis information 5a may comprise a piece of setpoint information about a horizontal main axis of the multiaxis robot 23. The piece of axis information 5a may also specify a setpoint position of all axes of the multiaxis robot 23, so that in this respect, the setpoint position 6 may also be unambiguously defined via such a piece of axis information 5a. Preferably, the piece of axis information 5a specifies the setpoint position 6 in the joint coordinates of the multiaxis robot 23.

Furthermore, it is preferably provided that the compensation parameter set 7 comprises an axis compensation parameter which is related to the piece of axis information 5a, via which the piece of axis information 5a is processed for ascertaining the correction value 12. In other words, the correction value 12 is a function of the piece of axis information 5a. Here, it may in particular be that the piece of axis information 5a, and preferably only the piece of axis information 5a, is adjusted via a component of the correction value 12 which derives from the axis compensation parameter. In this respect, the piece of axis information 5a, and specifically the aforementioned piece of setpoint information about the horizontal main axis of the multiaxis robot 23, may be compensated for in a targeted manner via the axis compensation parameter, so that the complexity of the calculations for compensating for the deviation 9 is reduced overall.

According to one preferred embodiment, the piece of setpoint position information 5 comprises a piece of Cartesian coordinate information 5b. In this case, it may be a position specification in Cartesian coordinates which corresponds to the setpoint position 6. Here, it is furthermore provided that the piece of axis information 5a was generated based on the piece of Cartesian coordinate information 5b. This action is also referred to as a backward transformation, whereas the transformation from the joint coordinates to the Cartesian coordinates is referred to as a forward transformation. It is preferably provided that the compensation parameter set 7 comprises an absolute compensation parameter which is related to the piece of Cartesian coordinate information 5b, via which the piece of Cartesian coordinate information 5b is processed for ascertaining the correction value 12. In this way, a mechanism for compensating for the deviation may thus be provided, which also at least directly takes into account the Cartesian coordinates of the setpoint position 6 or a corresponding portion of the piece of setpoint position information 5. This may possibly mean a reduction in computing cost or complexity versus an approximately exclusive consideration of the joint coordinates in terms of the piece of axis information 5a.

Claims

1. A method for compensating for a deviation in an operating point (1) of a manipulator (2) during the processing of a workpiece (3) via an end effector (4) on the manipulator (2); wherein a command sequence is processed for controlling the manipulator (2) for the purpose of processing the workpiece (3), and a piece of setpoint position information (5) corresponding to a setpoint position (6) is generated based on the command sequence, the operating point (1) of the manipulator (2) being set based on said piece of setpoint position information (5); wherein the piece of setpoint position information (5) is processed using a compensation parameter set (7) which is related to the piece of setpoint position information (5), for ascertaining a compensation value (8), and the piece of setpoint position information (5) is adjusted according to the compensation value (8) for compensating for a deviation (9) between an actual position (10) of the operating point (1) and the setpoint position (6);whereinthe actual position (10) is measured during the processing of the workpiece (3);in that a correction value (12) is ascertained, based on a comparison between the measured actual position (10) and the setpoint position (6); and in that the compensation parameter set (7) is adjusted during the processing of the workpiece (3) for reducing the deviation (9), based on the correction value (12).

2. The method as claimed in claim 1, wherein the piece of setpoint position information (5) is generated by an NC controller (11a), based on the command sequence, in particular after an interpolation and/or a transformation; preferably in that the piece of setpoint position information (5) is generated by the NC controller (11a) at an interpolation rate; and in that the adjustment of the compensation parameter set (7) for reducing the deviation (9) takes place at the interpolation rate.

3. The method as claimed in claim 2, wherein the processing of the piece of setpoint position information (5) using the compensation parameter set (7) related to the piece of setpoint position information (5) for ascertaining a compensation value (8), and the adjustment of the compensation parameter set (7) for reducing the deviation (9), take place in a running extension software module (13), said extension software module (13) communicating in terms of data technology with the NC controller (11a) via a connection interface (14) of said NC controller.

4. The method as claimed in claim 3, wherein the connection interface (14) calls the extension software module (13) in terms of data technology based on execution events of the NC controller (11a).

5. The method as claimed in claim 3 or 4, wherein the extension software module (13) receives state information about the manipulator (2), in particular comprising the actual position (10) and/or the piece of setpoint position information (5) and/or the setpoint position (6), via the connection interface (14).

6. The method as claimed in one of claims 1 to 5, wherein during the processing of the workpiece (3), the actual position (10) is measured via an optical position sensor (15), in particular a laser tracker (15a); preferably in that the optical position sensor (15) is connected in terms of data technology to the NC controller (11a) and/or to the extension software module (13) via running measuring software (18).

7. The method as claimed in one of claims 1 to 6, wherein the compensation parameter set (7) comprises an in particular manipulator type-related model parameter set (19) and a physical data set (20), preferably comprising mass information and/or center of mass information relating to the manipulator (2); and in that furthermore, in particular only the model parameter set (19) is adjusted based on the correction value (12) for reducing the deviation (9).

8. The method as claimed in one of claims 1 to 7, wherein during the processing of the workpiece (3), at least one environmental variable is measured; and in that the compensation parameter set (7) comprises an environmental compensation parameter which is related to the respective measured environmental variable, the respective measured environmental variable being processed using said environmental compensation parameter for ascertaining the compensation value (8).

9. The method as claimed in claim 8, wherein the at least one environmental variable comprises a pressure (22a), preferably, a pressure (22a) acting on the manipulator (2), in particular acting on a weight compensation system of the manipulator (2).

10. The method as claimed in claim 8 or 9, wherein the at least one environmental variable comprises an actual temperature (22b), preferably, an actual temperature (22b) at the manipulator (2); in particular in that the actual temperature (22b) is influenced by ambient heat and by self-heating of the manipulator (2).

11. The method as claimed in claim 10, wherein the environmental compensation parameter comprises a material-dependent temperature expansion coefficient and/or an in particular axis-related temperature-deflection coefficient and/or an in particular axis-related length parameter, said environmental compensation parameter being related to the measured actual temperature (22b).

12. The method as claimed in one of claims 1 to 11, wherein the at least one environmental variable comprises a process force (22c) acting on the manipulator (2) and/or on the end effector (4).

13. The method as claimed in one of claims 1 to 12, wherein the manipulator (2) is a multiaxis robot (23), preferably in that the piece of setpoint position information (5) comprises a piece of axis information (5a) about the multiaxis robot (23), in particular in that the piece of axis information (5a) comprises a piece of setpoint information about a horizontal main axis.

14. The method as claimed in claim 13, wherein the compensation parameter set (7) comprises an axis compensation parameter which is related to the piece of axis information (5a), via which the piece of axis information (5a) is processed for ascertaining the correction value (12); preferably in that in particular only the piece of axis information (5a) is adjusted via a component of the correction value (12) which derives from the axis compensation parameter.

15. The method as claimed in one of claims 1 to 14, wherein the piece of setpoint position information (5) comprises a Cartesian coordinate system (5b), preferably in that the piece of axis information (5a) was generated based on the piece of Cartesian coordinate information (5b); in particular in that the compensation parameter set (7) comprises an absolute compensation parameter which is related to the piece of Cartesian coordinate information (5b), via which the piece of Cartesian coordinate information (5b) is processed for ascertaining the correction value (12).