Method and device for the improvement of the pose accuracy of effectors on mechanisms and for the measurement of objects in a workspace
First Claim
1. A method for improving the pose accuracy of a mechanism in a workspace, wherein the mechanism is movable in at least one axis with tolerances and includes an effector, at least one effector object is mounted via a rigid connection to the effector eccentric to the at least one axis of the mechanism in an estimated pose with a tolerance in position and orientation, at least one reference object is arranged in the workspace with a tolerance in position and orientation, and a computer system is connected to the mechanism having a measurement control program, a parameter identification program, and a mechanism control program, the at least one effector object and the at least one reference object forming at least one trigger/detector pair comprising a signal trigger device and a signal detector for triggering and detecting a binary signal, wherein a totality of signal poses of the signal detector relative to the trigger device in which a signal is triggered on the signal detector is described by at least one non-trivial characteristic equation, said method comprising the following steps:
- (a) selecting a proximity sequence N of a finite number of proximity poses for the at least one trigger-detector pair, each of the proximity poses being located in the vicinity of a respective one of signal poses, wherein the proximity sequence N is selected such that the following criteria are fulfilled;
DG(N)≧
DG(AI)/15, whereby the distance on an arbitrary straight line G between two neighbouring points of the projection of the proximity sequence N onto G is at the most DG(N)/4, wherein DG is a function which maps each subset of the set AI to the distance between those two points of the projection of this subset on G which are at maximum distance from each other on G;
AI is the space of all those reachable effector poses of the given mechanism which result from elementary kinematic calculations on the basis of the known mechanism model which in turn is afflicted with tolerances;
G is an arbitrary straight line which contains at least two points of SI; and
SI is a subset of AI which is denotes the space of proximity poses and is defined by the totality of all those effector object poses where a signal would be expected according to a mathematical calculation based on the parameter values of the known tolerance-afflicted mechanism model, the estimated pose of the reference objects in space, and the estimated pose of the effector object on the mechanism;
(b) searching for a nearby signal pose for each of the proximity poses consecutively through movement of one of the at least one effector object and the mechanism using a simple search algorithm until a signal pose is detected;
(c) passing a momentary joint configuration of the mechanism onto the computer system after the detection thereof in said step (b) and storing the momentary joint configuration in the computer system as a data record;
(d) using a parameter identification program to identify the true values of one of the parameters influencing the pose accuracy of the mechanism and user-specific subsets of this parameter set, whereby a scaling factor is used for the identification of all length-parameters.
0 Assignments
0 Petitions
Accused Products
Abstract
A device and a method for measuring the pose of mechanisms includes at least one effector object fixedly joined to a mechanism (e.g., industrial robot, hexapod) in which the at least one effector object moves along one of several axes. The ideal, effective shape of the at least one effector object is a point, a straight line, a plane, an ellipsoid, a cylinder, a hyperboloid or a combination thereof. The movable effector objects interact with reference objects which are arranged in defined positions relative to the mechanism. The interactions are detected by a suitable sensor. The interactions are detected, only the pertaining joint configuration of the mechanism is transmitted to the information processing unit and evaluated and no further continuous values of measuring parameters are required for the evaluation.
68 Citations
31 Claims
-
1. A method for improving the pose accuracy of a mechanism in a workspace, wherein the mechanism is movable in at least one axis with tolerances and includes an effector, at least one effector object is mounted via a rigid connection to the effector eccentric to the at least one axis of the mechanism in an estimated pose with a tolerance in position and orientation, at least one reference object is arranged in the workspace with a tolerance in position and orientation, and a computer system is connected to the mechanism having a measurement control program, a parameter identification program, and a mechanism control program, the at least one effector object and the at least one reference object forming at least one trigger/detector pair comprising a signal trigger device and a signal detector for triggering and detecting a binary signal, wherein a totality of signal poses of the signal detector relative to the trigger device in which a signal is triggered on the signal detector is described by at least one non-trivial characteristic equation, said method comprising the following steps:
-
(a) selecting a proximity sequence N of a finite number of proximity poses for the at least one trigger-detector pair, each of the proximity poses being located in the vicinity of a respective one of signal poses, wherein the proximity sequence N is selected such that the following criteria are fulfilled;
DG(N)≧
DG(AI)/15, whereby the distance on an arbitrary straight line G between two neighbouring points of the projection of the proximity sequence N onto G is at the most DG(N)/4, whereinDG is a function which maps each subset of the set AI to the distance between those two points of the projection of this subset on G which are at maximum distance from each other on G;
AI is the space of all those reachable effector poses of the given mechanism which result from elementary kinematic calculations on the basis of the known mechanism model which in turn is afflicted with tolerances;
G is an arbitrary straight line which contains at least two points of SI; and
SI is a subset of AI which is denotes the space of proximity poses and is defined by the totality of all those effector object poses where a signal would be expected according to a mathematical calculation based on the parameter values of the known tolerance-afflicted mechanism model, the estimated pose of the reference objects in space, and the estimated pose of the effector object on the mechanism;
(b) searching for a nearby signal pose for each of the proximity poses consecutively through movement of one of the at least one effector object and the mechanism using a simple search algorithm until a signal pose is detected;
(c) passing a momentary joint configuration of the mechanism onto the computer system after the detection thereof in said step (b) and storing the momentary joint configuration in the computer system as a data record;
(d) using a parameter identification program to identify the true values of one of the parameters influencing the pose accuracy of the mechanism and user-specific subsets of this parameter set, whereby a scaling factor is used for the identification of all length-parameters. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 24)
measuring a distance between two arbitrary points on the calibration objects;
selecting one signal pose one each of the calibration objects;
measuring distances between the signal poses on each of the calibration objects and the arbitrary points;
feeding the measured distances to the computer system; and
calculating the geometrical relationship between the calibration objects and the scaling factor.
-
-
6. The method of claim 4, wherein said step of identifying the scaling factor is performed only during a first calibration and the scaling factor is determined during a re-calibration from a comparison of already known selected lengths of the mechanism which are unaltered during customary industrial use.
-
7. The method of claim 1, wherein the at least one effector object comprises at least two effector objects with a known relative pose to each other, the method further comprising the step of identifying the scaling factor by using at least three calibration objects comprising the at least one reference object and the at least two effector objects, such that at least two effector objects are moved into signal poses with at least one reference object, whereby the distance between the poses of the at least two effector objects is more than ⅙
- of a diameter of the workspace (Δ
), preferably ¾
Δ
, whereby the diameter of the workspace is defined by the maximum of DG(AI) where G ranges over all straight lines G and a mean value calculation is carried out for the determination of distances.
- of a diameter of the workspace (Δ
-
8. The method of claim 7, wherein said step of identifying the scaling factor comprises using elongated, stretched out calibration objects which are non-parallel and have a known orientation from a preceding calibration and includes the substeps of:
-
measuring a distance between two arbitrary points on the calibration objects;
selecting one signal pose one each of the calibration objects;
measuring distances between the signal poses on each of the calibration objects and the arbitrary points;
feeding the measured distances to the computer system; and
calculating the geometrical relationship between the calibration objects and the scaling factor.
-
-
9. The method of claim 7, wherein said step of identifying the scaling factor is performed only during a first calibration and the scaling factor is determined during a re-calibration from a comparison of already known selected lengths of the mechanism which are unaltered during customary industrial use.
-
10. The method of claim 1, wherein one of two effector objects and two reference objects are used.
-
11. The method of claim 1, further comprising the step of arranging the effector objects and the reference objects such that during execution of arbitrary movements, the effector objects attain signal poses with respect to the reference objects.
-
12. The method of claim 1, wherein the mechanism includes a pluraity of physically connected subsections and said steps (a) to (d) are completed for a selected subsection of the plural physically connected subsections, whereby only the joints of the selected subsection are moved during said steps (a)-(d) and the remaining ones of the physically connected subsections are considered rigid bodies.
-
13. The method of claim 1, further comprising the step of eliminating the influence of threshold values of the signal detectors, asymmetry in the arrangement of measurement, braking distances of the joints on the joint configuration relayed to the computer system and other signal delays by recording a systematic measurement sequence for each signal pose which includes the approach to and traverse of the signal pose on selected paths, whereby only individual axes of the mechanism are moved, and using the computer system to determine a correct data record using one of elementary mathematical operations and median calculations from detector reactions and associated joint configurations and store the correct data record on the computer system.
-
14. The method of claim 1 further comprising the steps of measuring and recording the residue errors of a reached target position which remain after calibration and passing the registered values onto one of a learning/adaptive system of an artificial neural network and a rule interpreter for processing and correcting the effector pose.
-
15. The method of claim 1, further comprising the steps of collecting further data from sensors suitable for the measurement of joint properties, passing the data onto the computer system, and evaluating the data in said step (d).
-
16. The method of claim 1, wherein one effector object of the at least one effector object and one reference object of the at least one reference object comprises a trigger/detector pair from the pairs comprising:
-
i) signal detector and straight electromagnetic wave/cylindrical interrupter rod;
ii) signal detector and electrically conductive wire/contact rod;
iii) laser beam/light-sensitive matrix area;
iv) electrically conductive plane/conductive contact rod;
v) multi-axis suspended linear guidance with contact threshold/multi-axis suspended, conducted rod;
vi) point-shaped, line-shaped or planar signal detectors/plane of electromagnetic waves; and
vii) wedge-shaped electromagnetic wave/several signal detectors.
-
-
17. The method of claim 1, wherein at least one calibration object is integrated in the mechanism to be calibrated.
-
18. The method of claim 1, further comprising the steps of receiving at least one calibration object at the mechanism connected firmly to it and discarding the calibration object after said step (d).
-
19. The method of claim 1, wherein said step (b) comprises searching for a signal pose using a higher-dimensional sensor areas.
-
20. The method of claim 1, wherein at least two reference objects are arranged in already known relative pose to each other, said method further comprising the step of determining a measurement of the relative pose of several mechanisms with respect to each other, the poses of the several mechanisms being different in at least one of location and time, by determining the pose of the mechanisms with respect to the reference objects using the parameters identified in the computer system in said step (d) and determining therefrom the poses of the several mechanisms with respect to each other.
-
21. The method of claim 1, wherein at least two reference objects are arranged in known relative pose, said method further comprising the step of determining the absolute pose of a mechanism with respect to an arbitrary coordinate system having a pose that is known with respect to at least one of the reference objects by determining the pose of the mechanism to the reference objects using the parameters identified in the computer system in said step (d) and determining therefrom the pose of the mechanism to the coordinate system.
-
22. The method of claim 1, wherein arbitrary objects in the workspace are equipped with reference objects which are at a known distance to each other, said method further comprising the step of measuring the pose of one of the reference objects relative to the mechanism and the objects with respect to each other and determining signal poses successively by the motion of the effector object via motion of a single axis of the mechanism only and calculating therefrom the relative pose of the mechanism and the objects to each other by the computer system.
-
24. The method of claim 1, wherein after performing said steps (a)-(d) once, the following steps are performed:
-
(e) selecting a proximity sequence N such that it contains at least one proximity pose arranged at a reference object neighbouring a target pose;
(f) successively searching for detection of a signal pose nearby to the at least one proximity pose through the motion of one of the effector object and mechanism via a simple search algorithm;
(g) passing the momentary joint configuration of the mechanism onto the computer system and storing the momentary joint configuration as a data record upon detection of the signal in said step (f);
(h) using the computer system to calculate, for each data record, the incorrect pose in the workspace resulting mathematically on the basis of the mechanism parameters currently known to the controller; and
(i) calculating a correction movement from the difference between the signal poses and the associated incorrect poses, through elementary interpolation procedures and elementary error compensation algorithms, the correction movement compensating the deviation of the pose actually steered for by the mechanism control from the desired pose, whereby the scalar factor, which was determined from the exactly known pose of the reference object, is used for the calculation of the correction movement from incorrect poses and signal poses.
-
-
23. A method for improving the pose accuracy of a mechanism in a workspace, wherein the mechanism is movable in at least one axis with tolerances and includes an effector, at least one effector object is mounted via a rigid connection to the effector eccentric to the at least one axis of the mechanism in an estimated pose with a tolerance in position and orientation, at least one immaterial reference object is arranged in the workspace with an exact known pose, and a computer system is connected to the mechanism having a measurement control program, a parameter identification program, and a mechanism control program, the at least one effector object and the at least one immaterial reference object forming a trigger/detector pair comprising a signal trigger device and a signal detector for triggering and detecting a binary signal, wherein a totality of signal poses of the signal detector relative to the trigger device in which a signal is triggered on the signal detector is described by at least one non-trivial characteristic equation, said method comprising the following steps:
-
(a) selecting a proximity sequence N such that it contains at least one proximity pose arranged at a reference object neighbouring a target pose;
(b) successively searching for detection of a signal pose nearby to the at least one proximity pose through the motion of one of the effector object and mechanism via a simple search algorithm;
(c) passing the momentary joint configuration of the mechanism onto the computer system and storing the momentary joint configuration as a data record upon detection of the signal in said step (b);
(d) using the computer system to calculate, for each data record, the incorrect pose in the workspace resulting mathematically on the basis of the mechanism parameters currently known to the controller; and
(e) calculating a correction movement from the difference between the signal poses and the associated incorrect poses, through elementary interpolation procedures and elementary error compensation algorithms, the correction movement compensating the deviation of the pose actually steered for by the mechanism control from the desired pose, whereby the scalar factor, which was determined from the exactly known pose of the reference object, is used for the calculation of the correction movement from incorrect poses and signal poses.
-
-
25. A device for improving the pose accuracy of a mechanism and for pose measurement of objects in the work space, comprising:
-
a computer system comprising measurement control program, a parameter identification program, and a mechanism control program;
a mechanism moveable in at least one axis which has an effector, said mechanism connected to said computer system;
at least one pair of calibration objects, each pair of said at least one pair comprises an effector object rigidly connected with said effector and mounted eccentrically to the at least one axis of said mechanism and a reference object fixedly arranged relative to said mechanism in said workspace, each pair comprising a signal trigger device and a signal detector for binary signals; and
an installation for determining a scaling factor. - View Dependent Claims (26, 27, 28, 29, 30)
a CCD matrix-camera, a camera with diffuser which can be placed in a radiation beam, a laser or maser radiation source with a photo detector, a light sensor, a position-sensing device (PSD), a light barrier device, an electrically conductive contact rod, and electrically conductive prolonged wire, a radiation source creating light planes, a radiation source creating light wedges with mirrors or lenses which can be activated permanently or periodically, a two-dimensional sensor array made up of point-shaped light detectors, and rigid bodies of geometrically regular design which may be designed in the form of two or three orthogonal rods with exactly cylindrical rods.
-
-
28. The device of claim 25, wherein said effector object device comprises an integral part of said mechanism.
-
29. The device of claim 25, wherein said effector object is selectively held by said effector on said mechanism.
-
30. The device of claim 25, wherein said reference object comprises one of a large container known in some dimensions of length with accessible surface of a liquid contained therein and a large plate known in some length dimensions with traceable edges.
-
31. The device for improving the pose accuracy of a mechanism and for pose measurement of objects in a work space, comprising:
-
a computer system comprising measurement control program, a parameter identification program, and a mechanism control program;
a mechanism having tolerances and moveable in at least one axis with an effector;
at least one effector object in estimated pose with tolerance in position and orientation is rigidly connected with said effector and mounted eccentrically to said at least one axis of said mechanism;
at least one immaterial reference object arranged fixedly in the workspace and the pose of which is known exactly, wherein each one of said at least one effector object forms a trigger/detector pair with each one of said at least one reference object, said trigger/detector pair comprising a signal trigger device and a signal detector such that each pair is suitable for effecting a triggering and detection of at least binary signals, wherein a totality of signal poses of said signal detector relative to said signal trigger device in which a signal is triggered on the detector are described by at least one non-trivial characteristic equation.
-
Specification