DE102014012670A1 - Method for calibrating an at least single kinematic redundant robot - Google Patents

Abstract

The invention relates to a method for calibrating at least one kinematic redundant robot (10), in particular a lightweight robot, comprising the steps of: fixing an end effector (14) of the robot (10); Setting at least as many different axis positions of the robot (10) with fixed end effector (14) as the robot (10) axes (12); Detecting respective axis parameters for the respective axis positions; Determining respective, the relative positioning of the different axes (12) mutually characteristic kinematics parameters (Qj, τj, dj, aj, αj) of the robot (10) taking into account the detected axis parameters.

Description

The invention relates to a method for calibrating an at least simply kinematically redundant robot, in particular a lightweight robot.

Lightweight robots used in industrial applications often have a lack of positioning accuracy. As a result, points approached by the controller deviate significantly from the actual approached points of the robot. One reason for this is inter alia in not exactly modeled deviations of the kinematic properties of the robot in question, for example with respect to tolerances of their axial lengths, faulty axle zero positions and the like. Over the lifetime of the robots, these kinematic properties may also change after collisions or wear. In the case of faulty kinematic parameters, robots do not move the spatial points as required, but with positional deviations, which means that the functioning of the associated process can not be ensured. This also applies to tools which are mounted on such robots, which are usually subject to a certain wear. In particular, the geometric properties of the tools used are also detected by such kinematic parameters. If these are not adjusted according to the changing geometry of the tools, this also leads to undesired positioning inaccuracies of the tools guided by the robot.

There are already surveying systems to counteract the above problem. For example, the shows US 5,402,582 A a measuring arm with an accuracy compensation, by means of which a robot could be calibrated. The system shown there is relatively expensive and also expensive to handle. In particular, such systems are difficult to use in serial systems of production systems and therefore rather little or not at all suitable for use in lightweight robots.

It is therefore the object of the present invention to provide an improved method for calibrating a robot.

This object is achieved by a method for calibrating an at least single kinematic redundant robot having the features of the independent patent claim. Advantageous embodiments with expedient and non-trivial developments of the invention are specified in the dependent claims.

• – Fixieren eines Endeffektors des Roboters;
• – Einstellen zumindest so vieler unterschiedlicher Achsstellungen des Roboters bei fixiertem Endeffektor, wie der Roboter Achsen aufweist;
• – Erfassen jeweiliger Achsparameter für die jeweiligen Achsstellungen;
• – Ermitteln jeweiliger, die Relativpositionierung der unterschiedlichen Achsen zueinander kennzeichnenden Kinematikparameter des Roboters unter Berücksichtigung der erfassten Achsparameter.

The method according to the invention for calibrating an at least single kinematic redundant robot, in particular a lightweight robot, comprises the following steps:

• - Fixing an end effector of the robot;
• - setting at least as many different axis positions of the robot with the end effector fixed, as the robot has axes;
• - Acquisition of respective axis parameters for the respective axis positions;
• Determining respective, the relative positioning of the different axes to each other characteristic kinematic parameters of the robot, taking into account the detected axis parameters.

Under a kinematic redundant robot or a robot with a redundant kinematics is to be understood that the robot in question can approach a particular point in space with its end effector in any number of poses. In other words, any number of different axis positions can be set with the robot fixed with the end effector, so that the end effector always remains at one and the same point in the arbitrarily different axis positions of the robot. This type of robot movement is also called zero-space motion.

In the method according to the invention, this null space movement is operated by setting at least as many different axis positions of the robot with the end effector fixed, as the robot has axes. After respective adjustment of the axes of the robot respective axis parameters for the respective Achsstellungen, z. B. their approached angular positions detected. Taking into account the detected axis parameters, respective kinematics parameters of the robot characterizing the relative positioning of the different axes are determined. Since the real position of the end effector does not change when setting the different axis positions, the positional difference of two points measured with respect to the end effector is always the same. This is inventively used to determine the kinematic parameters of the robot, so that they can be incorporated into an internal robot model. As a result, the end effector or the so-called tool center point moves again along expected or desired points and paths. As a result, the absolute accuracy of the thus calibrated robot is significantly improved.

As a result of the calibration according to the invention by means of the internal sensors present in the robot, the process accuracy of the robot is considerably increased. Furthermore, the interchangeability of the robot is significantly improved, wherein At the same time loss or downtime of corresponding systems can be reduced when exchanging robots.

In an advantageous embodiment of the invention, it is provided that in addition at least three more axis positions are set at a fixed end effector. This is particularly advantageous if a tool is still attached to the end effector. Because at least three further axis positions are set with a fixed end effector and taken into account in the determination of the kinematic parameters, a changing geometry of the tool, in particular due to tool wear, can be taken into account in the calibration of the robot. In addition, it can also be taken into account from the outset dimensional inaccuracies with respect to corresponding tolerances of the tool, so that the attached to the robot tool can be positioned very accurately.

According to a further advantageous embodiment of the invention, it is provided that the robot independently performs the calibration after a predetermined number of work steps or a predetermined duration. In other words, it can be provided that the robot independently performs a so-called Nachteachen. During the use of the robot, the latter can thus automatically carry out the calibration process according to the invention automatically, so that tool wear or even wear on the robot can be compensated for independently. As a result, the robot can travel particularly accurately according to desired trajectories or points even with prolonged use.

In a further advantageous embodiment of the invention, it is provided that the kinematics model according to Denavit-Hartenberg is used to determine the kinematic parameters. In this context, the kinematic parameters are preferably determined by numerical solution of a system of equations in which the same coordinates are always used for the position of the end effector. For example, a Newton-Euler method can be used as a numerical solution method.

Further advantages, features and details of the invention will become apparent from the following description of preferred embodiments and from the drawings. The features and feature combinations mentioned above in the description as well as the features and feature combinations mentioned below in the description of the figures and / or in the figures alone can be used not only in the respectively specified combination but also in other combinations or in isolation, without the scope of To leave invention.

Embodiments of the invention are explained in more detail below with reference to schematic drawings.

The drawing shows in:

1 a perspective view of a lightweight robot, which has seven adjustable axes; and in

2 three different axis settings of the robot, whereby an end effector of the robot always remains in one and the same position and only the axis positions of the robot are varied.

A lightweight robot 10 is in a perspective view in 1 shown. The robot 10 is presently designed as a seven-axis robot, thus has seven axes 12 on, by means of which an end effector 14 of the robot 10 can be positioned. To approach a certain point in space, six axes would be sufficient so that the predetermined point could be approached with respect to its three translational and three rotational coordinates. Because the robot 10 in the present case seven axes 12 Thus, this is simply kinematically over-determined or simply kinematically redundant. As a result, the different axes can 12 occupy substantially arbitrarily different positions, to one and the same given space point with the end effector 14 to approach.

In 2 is the robot 10 shown in three different poses, with the end effector 14 always arranged at one and the same spatial position. In other words, the robot leads 10 each a zero space movement through to take the different poses. The following is a procedure for calibrating the robot 10 explains what makes use of that the robot 10 is simply kinematically redundant or simply kinematically over-determined, so that the robot 10 the said zero space movements can perform.

First, the end effector 14 of the robot 10 fixed or clamped. Subsequently, at least as many different axis positions of the robot at the fixed end effector 14 set, like the robot 10 axes 12 having. In the present case, therefore, at least seven different poses by a corresponding adjustment of the respective axes 12 of the robot 10 ingested. Preferably, however, more than seven different poses are made with the robot 10 ingested.

For example, it may be that at the end effector 14 a tool not shown here is arranged. In this case, in addition to the seven poses, there are at least three more axis positions with the end effector fixed 14 or fixed tool set. For each pose taken, the respective axis parameters for the respective axis positions are recorded. Subsequently, respective, the relative positioning of the different axes 12 kinematic parameters Q _j , τ _j , d _j , a _j , α _j , which characterize one another. The kinematics model according to Denavit-Hartenberg is used to determine the kinematic parameters Q _j , τ _j , d _j , a _j , α _j .

Because the real position or the absolute spatial position of the end effector 14 or a tool arranged thereon does not change, is the positional difference of two successively measured points with respect to the end effector 14 or a tool arranged on it always the same. The relationship thus arises for a respective position P _{j Pj} - _{Pj + 1} = 0.

The position P _{j of} the end effector 14 or a tool attached to this is a function of the kinematic parameters Q _j , τ _j , d _j , a _j , α _j , so that applies P _j = f (Q _j, τ _j, d _j, a _j, α _j).

Thus, the following numerical approach can be chosen: 0 = f (Q _j , τ _j , d _j , a _j , α _j ,) - f (Q _{j + 1} , τ _{j + 1} , d _{j + 1} , a _{j + 1} , α _{j + 1} ).

By using a numerical mathematical method to solve a non-linear equation system, all kinematic parameters Q _j , τ _j , d _j , a _j , α _j can thus be obtained. For example, a Newton-Euler method can be used as a numerical solution method.

The determined kinematics parameters Q _j , τ _j , d _j , a _j and α _j are then entered into an internal robot model. This is done analogously to the procedure of already known calibration systems.

As a result, the tool center point or the end effector moves 14 again on correspondingly predetermined real path points, or the absolute accuracy of the robot 10 is significantly improved over the state before calibration.

Preferably, more than the ten Achsstellungen with fixed end effector 14 approached so that as many different axis parameters as input variables for the determination of kinematic parameters are available. Preferably, the axes 12 Also adjusted as much as possible between the different poses during the Nullraumbewegung, which has a positive effect on the determination of the kinematic parameters.

Preferably, the robot performs 10 the above-explained calibration automatically after a predetermined number of steps or after a predetermined duration automatically. This can cause signs of wear on tools used or on the robot 10 even be compensated by a recurring automatic self-calibration.

The described method has the potential to increase the interchangeability of the robot 10 significantly increased, so that corresponding downtime and personnel costs can be significantly reduced in robot changes. Furthermore, tool deformations and age-related wear on the robot can be particularly simple 10 which also reduces downtime on the robot itself and on any equipment over the life of the robot 10 as a result of which a corresponding added value can be increased.

QUOTES INCLUDE IN THE DESCRIPTION

This list of the documents listed by the applicant has been generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Cited patent literature

• US 5402582 A [0003]

Claims (6)

Method for calibrating an at least single kinematically redundant robot ( 10 ), in particular a lightweight robot, comprising the steps of: - fixing an end effector ( 14 ) of the robot ( 10 ); - setting at least as many different axis positions of the robot ( 10 ) with fixed end effector ( 14 ) like the robot ( 10 ) Axes ( 12 ) having; - Acquisition of respective axis parameters for the respective axis positions; Determining the respective, the relative positioning of the different axes ( 12 ) characteristic kinematics parameters (Q _j , τ _j , d _j , a _j , α _j ) of the robot ( 10 ) taking into account the recorded axis parameters. A method according to claim 1, characterized in that in addition at least three more Achsstellungen with fixed end effector ( 14 ). Method according to claim 1 or 2, characterized in that the robot ( 10 ) Performs the calibration after a predetermined number of steps or a predetermined duration independently. Method according to one of the preceding claims, characterized in that the kinematics model according to Denavit-Hartenberg is used to determine the kinematic parameters (Q _j , τ _j , d _j , a _j , α _j ). Method according to one of the preceding claims, characterized in that the kinematics parameters (Q _j , τ _j , d _j , a _j , α _j ) are determined by numerical solution of a system of equations in which the position of the end effector ( 14 ) always use the same coordinates. A method according to claim 5, characterized in that a Newton-Euler method is used as a numerical solution method.

Images