DE102018112370A1 - Direction-dependent collision detection for a robot manipulator - Google Patents

Direction-dependent collision detection for a robot manipulator

Info

Publication number
DE102018112370A1
DE102018112370A1 DE102018112370.1A DE102018112370A DE102018112370A1 DE 102018112370 A1 DE102018112370 A1 DE 102018112370A1 DE 102018112370 A DE102018112370 A DE 102018112370A DE 102018112370 A1 DE102018112370 A1 DE 102018112370A1
Authority
DE
Germany
Prior art keywords
ext
axes
task
execution
robot manipulator
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
DE102018112370.1A
Other languages
German (de)
Inventor
Sven Parusel
Saskia Golz
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Franka Emika GmbH
Original Assignee
Franka Emika GmbH
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Franka Emika GmbH filed Critical Franka Emika GmbH
Priority to DE102018112370.1A priority Critical patent/DE102018112370A1/en
Publication of DE102018112370A1 publication Critical patent/DE102018112370A1/en
Pending legal-status Critical Current

Links

Images

Classifications

    • BPERFORMING OPERATIONS; TRANSPORTING
    • B25HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
    • B25JMANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
    • B25J9/00Programme-controlled manipulators
    • B25J9/16Programme controls
    • B25J9/1628Programme controls characterised by the control loop
    • B25J9/1633Programme controls characterised by the control loop compliant, force, torque control, e.g. combined with position control
    • BPERFORMING OPERATIONS; TRANSPORTING
    • B25HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
    • B25JMANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
    • B25J9/00Programme-controlled manipulators
    • B25J9/16Programme controls
    • B25J9/1674Programme controls characterised by safety, monitoring, diagnostic
    • GPHYSICS
    • G05CONTROLLING; REGULATING
    • G05BCONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B2219/00Program-control systems
    • G05B2219/30Nc systems
    • G05B2219/39Robotics, robotics to robotics hand
    • G05B2219/39319Force control, force as reference, active compliance
    • GPHYSICS
    • G05CONTROLLING; REGULATING
    • G05BCONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B2219/00Program-control systems
    • G05B2219/30Nc systems
    • G05B2219/39Robotics, robotics to robotics hand
    • G05B2219/39343Force based impedance control
    • GPHYSICS
    • G05CONTROLLING; REGULATING
    • G05BCONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B2219/00Program-control systems
    • G05B2219/30Nc systems
    • G05B2219/39Robotics, robotics to robotics hand
    • G05B2219/39346Workspace impedance control
    • GPHYSICS
    • G05CONTROLLING; REGULATING
    • G05BCONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B2219/00Program-control systems
    • G05B2219/30Nc systems
    • G05B2219/39Robotics, robotics to robotics hand
    • G05B2219/39347Joint space impedance control
    • GPHYSICS
    • G05CONTROLLING; REGULATING
    • G05BCONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B2219/00Program-control systems
    • G05B2219/30Nc systems
    • G05B2219/39Robotics, robotics to robotics hand
    • G05B2219/39348Generalized impedance control
    • GPHYSICS
    • G05CONTROLLING; REGULATING
    • G05BCONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B2219/00Program-control systems
    • G05B2219/30Nc systems
    • G05B2219/40Robotics, robotics mapping to robotics vision
    • G05B2219/40201Detect contact, collision with human
    • GPHYSICS
    • G05CONTROLLING; REGULATING
    • G05BCONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B2219/00Program-control systems
    • G05B2219/30Nc systems
    • G05B2219/40Robotics, robotics mapping to robotics vision
    • G05B2219/40202Human robot coexistence
    • GPHYSICS
    • G05CONTROLLING; REGULATING
    • G05BCONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B2219/00Program-control systems
    • G05B2219/30Nc systems
    • G05B2219/40Robotics, robotics mapping to robotics vision
    • G05B2219/40483Find possible contacts

Abstract

The invention relates to a method for controlling an actuator-driven robotic manipulator (1) having an end effector (3) in which the end effector (3) executes a predetermined set movement and executes a task during the execution of the set movement, comprising the steps: - during execution determining movement (S1) of an external force windler K introduced into the robot manipulator (1), wherein there is no vector F for at least one external force and / or vector Mzat least one external moment, - detecting (S2) an undesired collision of the robot manipulator (1), if Kin or n axes of a predefined coordinate system with N axes exceeds a predefined first limit, where n∈N, - detecting (S3) a faulty execution of the task if Kin or by m axes of the coordinate system exceeds a predefined second limit or K <Kist, where none depends on the task and is expected and / or desired force winders occurring in or about the m axes, and where m∈N and m∉n, where the first limit is less than the second limit, and - driving (S4) the robot manipulator (1) in FIG an error mode when an undesired collision of the robot manipulator (1) and / or a faulty execution of the task is detected.

Description

  • The invention relates to a method for controlling an actuator-driven robot manipulator with an end effector, and to a device for controlling an actuator-driven robot manipulator with an end effector and a robot with such an apparatus.
  • The object of the invention is to better detect collisions of a robot manipulator, in particular during the execution of a task by the robot manipulator, and thereby perform the task better.
  • The invention results from the features of the independent claims. Advantageous developments and refinements are the subject of the dependent claims.
  • A first aspect of the invention relates to a method for controlling an actuator-driven robotic manipulator having an end effector, in which the end effector executes a predetermined target movement and performs a task during execution of the target movement, comprising the steps of:
    • during the execution of the desired movement, determination of an external power wincher K ext introduced into the robot manipulator, wherein K ext has a vector F ext of at least one external force and / or a vector M ext of at least one external moment,
    • Detecting an unwanted collision of the robot manipulator, if K ext in or around n axes of a predefined coordinate system with N axes exceeds a predefined first limit, where n ∈ N,
    • - detecting a defective execution of the task when K ext exceeds in or near m axes of the coordinate system a predefined second threshold value or if K ext <K of, where K of a dependent of the task and occur exclusively in, or to the m axis of expected and / or desired force Winder, where K of a vector F of the at least one expected and / or desired force and / or a vector m of the expected of at least one and / or has the desired torque, and wherein m ∈ N and m ∉ n, wherein the first threshold is less than the second threshold, and
    • Driving the robot manipulator in an error mode when an undesired collision of the robot manipulator and / or an erroneous execution of the task is detected.
  • In particular, an external force occurs as such in an axis, and an external moment as such about an axis. Thus, K DES is a dependent on the task and occur exclusively in, or to the m axis of expected and / or desired force Winder, in particular the pre-defined coordinate system and the m axis, and especially its orientation, defined on the basis of prior knowledge of K of the appropriately. K des results advantageously by knowing the task and the expected during the execution of the task by interaction with an environment, acting on the robot manipulator mechanical forces and / or moments. Since the execution of the task required by the robot manipulator a period of K is also advantageous time-dependent: K = K of the (t).
  • The end effector is also referred to as such when it is not aktuiert itself, that is, even has no movable or otherwise controllable actuators. The end effector can thus also simply be the distal end of the robot manipulator.
  • Under a desired movement is advantageously understood a kinematic data set for specifying the movement of the end effector of the robot manipulator. Advantageously, the desired movement is defined via a desired movement path, that is to say by means of a sequence of desired positions of a specific reference point on the robot manipulator, in particular a reference point on the end effector of the robot manipulator or on the distal end of the robot manipulator, and furthermore advantageously by means of a desired sequence for this purpose of positions associated speed and an acceleration of the reference point. Accordingly, the term "desired movement" also includes a desired standstill of this reference point considered. Then, in this case of the desired standstill, a setpoint position that is constant over time and a setpoint speed of zero and a setpoint acceleration of zero of the reference point apply.
  • External forces and moments are understood here as meaning those which do not arise through the drives which are connected to the robot manipulator. Preferably, the joints of the robot manipulator as drives on electric motors for generating moments between the robot members, which are each rotatably connected to each other by the joint. These moments, which can be equivalently converted into a force by dividing a corresponding radius to be considered, are, in contrast to the external forces and moments mentioned above, understood as internal forces and moments.
  • The comparison of K ext with the K and K ext of, preferably, with the first and second component-wise limit of vector generally K ext and of the vectorial in general K, so scalars are compared with each of the scalars. Alternatively, scalar vector norms || K ext || are preferred and || K of || formed and compared with each other and with the first and the second limit. The vector standard is preferably the 2-norm, that is to say || K ext || 2 and || K of || 2 , which as a time integral of the squared vector components forms an energy-equal standard, or alternatively preferably as a vector standard the ∞-norm, that is to say || K ext || and || K of || , which gives the largest value of all vector components over time. In the case of using the ∞ standard and / or the 2 standard, the comparisons are advantageously made individually for the m axes and the n axes after separation of the vector variables into the components of the m axes and into the components of the n axes.
  • Preferably, the determination of the external power winch K ext by means of sensors, in particular by means of force sensors and / or torque sensors. In particular, the moment sensors are preferably already installed in joints of the robot manipulator together with the motor located in the respective joint. Furthermore, force sensors are preferably arranged on a structural component of a robot member, wherein this type of force sensors determines a tension or a force in the respective robot member via the material expansion of the robot member and the known material constants, in particular of the modulus of elasticity. Furthermore, electrical current sensors are preferably used which measure the electric current through an electric motor of the robot manipulator in order to conclude from changes in the electric current to an abnormal change in the torque of the motor and thus to a collision which manifests itself in an external force or in an external moment. In this case, the motors themselves serve as torque sensors.
  • Detecting an undesired collision of the robotic manipulator consists, in particular, in detecting an undesired collision of the robotic manipulator with an object of the environment, with an object picked up at the end effector, or by a collision of the robotic manipulator with itself, for example between two links of the robotic manipulator, or with one Robot base to which the robot manipulator is connected and which serves as a support for the robot manipulator.
  • In particular, the external force winder K ext is a vector which combines the detected forces F ext as a column vector and the detected moments M ext as a column vector and records them together in a preferably six-dimensional column vector, since both F ext and M ext, in particular in a Cartesian coordinate system recorded three entries for three mutually orthogonal spatial directions. In particular, if only external forces are determined, the components of this vector contain a zero in all entries for the external moments. Preferably, and in the most general case, therefore, for the external force winder K ext = [F ext T , M ext T ] T , where the superscript "T" in the form of "(·) T " indicates the transposed operator through which a row vector becomes the column vector and vice versa.
  • The predefined coordinate system has in particular N axes. These N axes are preferably such that an indication of a coordinate by means of the predefined coordinate system requires the specification of only one value on each individual axis of the coordinate system and also allows what is the case in particular for orthogonal axes. Especially for complex load cases with several force directions of an expected and / or desired force Winders K of at execution of the task is alternatively preferably an overdetermined predefined coordinate system used, so that the m axis, occurs in or near that of the expected and / or desired force Winder K of , not necessarily orthogonal to each other and preferably form an angle of <90 ° to each other and thus more than six spatial directions (that is, more than three spatial directions in ignoring signs) by the m axes can be considered.
  • The external force winder is determined in a Winder coordinate system, which is generally independent of the predefined coordinate system in which the n axes and the m axes are defined. If both coordinate systems differ in particular in the location of their origin or also in their orientation, or even in their type, for example if the Winder coordinate system defines rotation coordinates with radius r and vectorial angle a and the predefined coordinate system is a Cartesian coordinate system, in particular the external Kraftwinder transformed by coordinate system transformation from the Winder coordinate system in the predefined coordinate system to specify in the predefined coordinate system, a respective component of the external power winders in a respective axis of the predefined coordinate system. The coordinate system transformation consists in particular of an image, preferably comprising a rotation matrix and a displacement vector for taking into account different origins of the respective coordinate systems.
  • As a result, the m axes differ from the n axes in that a force winder K des dependent on the task is expected and / or desired in or around the m axes. This is expressed by the terms m∈N m∉ and n, which state that the pre-defined coordinate system, in particular the N axes, whose orientation in space and in particular the selection of the m axis according to the direction and orientation of the force K the Winders are defined. Furthermore, the terms m∈N and m∉ n mean that an axis can fall below either the m axes or below the n axes, but can not simultaneously be an axis of the m and the n axes. In other words, for a certain axis, a component of the power wind K of is expected or not.
  • This expected and / or desired force Winder K des , in particular by a desired contact with the environment of the robot manipulator with an object from the environment to, for example, when editing an object by the end effector, when gripping an object by the end effector, when editing a workpiece through the end effector, for example, in welding, drilling, milling, painting, or the like. By such activities arises according to the Newton's law "Actio is equal to Reactio" a force on in particular the end effector, which should not be regarded as an undesirable collision. Therefore, this expected and / or desired force Winder K is the taken into account as described above, in particular a desired force Winder K of preferably from a force control of the robot manipulator is determined, and an expected force Winder K of the preferred estimation a contact force of a manipulation object is determined. In turn applies to the expected and / or desired force Winder K of the vector property analogous to the external force Winder K ext, where K of the expected and / or desired forces F of as a column vector and the expected and / or desired moments M of the combined as a column vector and together is noted in a six-dimensional column vector, since both F des and M of the three, in particular in a Cartesian coordinate system, each capture three entries for three mutually orthogonal spatial directions. If, in particular, only expected and / or desired forces are determined, then the components of this vector contain a zero in all entries for the external moments. Preferably, and in the most general case, therefore applies to the expected and / or desired force Winder K of = [F of T , M of T ] T , wherein the superscript "T" in the form of "(·) T " the transposed operator indicates that a row vector becomes the column vector and vice versa.
  • The axes are preferably to be considered separately in their respective direction for the comparison of the determined force winders with the respective limit value, that is to say that in a Cartesian coordinate system, a distinction is made in particular between a positive and a negative axis, for example a direction "+ x" and an order 180 ° rotated on the other hand running direction "-x". Advantageously, a first limit value can thus be defined in a first direction of a Cartesian coordinate system, and a further first limit value can be defined with differing values for the negative direction of the first direction. Particularly preferably, this distinction is made according to the sign of the direction only in the m axes, whereas in the n axes of the force Winder K ext is compared in terms of amount only with the first and second limit. For all directions of each axis, that is to say for both directions "+ x" and "-x", the absolute value component of the external power wind K ext is hereby compared with a respective positive limit value. The statement that the first limit value is smaller than the second limit value is therefore always to be understood as positive limit values in comparison to absolute components of the external power wind K ext and analogously thereto in comparison with the magnitude expected and / or desired force winder K des .
  • The term "fault mode" generally indicates a signal, command or control program that defines the actuation of the robotic manipulator in response to detecting an undesired collision of the robotic manipulator and / or an erroneous execution of the task. Preferred embodiments of the failure mode are those listed below.
  • The failure mode is preferably that the execution of the task is aborted first and then a new attempt to execute the task is repeated with changed parameters. Alternatively preferably, the error mode consists in a termination of the instantaneous movement until the robot manipulator is at a standstill, that is to say in a so-called "safe stop". As an alternative alternative, the execution of the error mode is a change in the parameters of a compliance control. Such a compliance control generates an artificial spring-mass-damper model of the robotic manipulator and defines this model as desired behavior which the controller of the robotic manipulator is to produce. Preferably, this change in the compliance control in the failure mode results in a lower spring force constant with correspondingly reduced damping of the spring-mass-damper model. Another alternatively preferred failure mode is an active avoidance of the external force Winder K ext . In this case, a controller of the robot manipulator becomes active controlled accordingly and acted upon by corresponding setpoint values of position and / or speed and / or acceleration, so that the extant external force Winder K ext is actively avoided. For this purpose, a method for determining an acting location, that is localization, of at least one external force or external moments is advantageously used. Alternatively, the execution of the failure mode is to issue a warning to a user or to another person. Generally formulated are advantageous for the fault mode three general ways to consider, namely an active reaction, a passive reaction, or stopping the movement of the robot manipulator on the respective limit exceeding external force Winder K ext .
  • It is an advantageous effect of the invention that a collision of a robot manipulator can be more accurately detected. This advantageous effect occurs in particular in that in or around the m axes of the predefined coordinate system in which a force is expected and / or desired, the second limit value is set higher than the first limit value for the n axes in which no contact force expected and / or desired. As a result, unwanted collisions are recognized rather quickly by the relatively low first limit in the n axes, whereas the forces and / or moments on the axes do not exert any deleterious influence on the axes that could lead to misdetections of unwanted collisions on the robot manipulator.
  • According to an advantageous embodiment of a faulty execution of the task is the power Winder K ext determined by the expected and / or desired force Winder K of compensated and the compensated force Winder with the second threshold value is compared, at least for the detection.
  • This embodiment is carried out in particular with known or predetermined force Winder K of . Accordingly, only the difference between the force Winder K des and the force Winder K ext is advantageously compared to the second limit value. Thereby an uncertainty, which share in the power actually Winder K ext of the force of the blast K occupies taken out, which provides a comparison with the second threshold value is advantageously more reliable results.
  • According to a further advantageous embodiment, the external force Winder K ext is determined by means of a pulse observer.
  • In particular, the pulse observer records the angular accelerations of the joints, estimated on the basis of the engine torques, between the robot members of the robot manipulator. These can be compared with the actual angular accelerations of the joints and thus advantageously closed in case of deviations to external forces and moments.
  • According to a further advantageous embodiment, the predefined coordinate system and / or the m axes and / or the n axes are time-variant and dependent on the desired movement and / or the task.
  • Advantageously, and particularly if the expected force Winder K of the changes in orientation of the course of execution of the task, are here advantageously about the axes m and / or n axes adjusted.
  • According to a further advantageous embodiment, the first limit value and / or the second limit value are time-variant and dependent on the progress in the course of the execution of the task.
  • According to a further advantageous embodiment, the first and / or the second limit value are predetermined by a user and are adaptable by the user.
  • According to a further advantageous embodiment, the first and / or the second limit value are defined together with the task.
  • According to a further advantageous embodiment, the first and / or the second limit value are determined and adapted by machine learning.
  • By "machine learning" is preferably understood a parametric adaptation of the first limit value and / or the second limit value. The parametric adaptation preferably takes place gradient-based or based on a general cost or energy function, wherein the respective parameter or limit itself enters the energy function or the cost function at least quadratically, so that when forming a time derivative of the cost or energy function the value of the respective function decreases over time and thus has a convergence of the respective limit value or parameter to lower values of the cost function. Alternatively, preferably one or more statistical functions are used for machine learning, in particular to form an expected value for the first or second limit value depending on the past first or second limit values. Further, machine learning relies on the use of neural networks or related trainable constructs which, in particular, provide an input / output function via superimposed and adapted sigmoid functions. Depict output behavior depending on environmental parameters as input values, and thus determine a first and / or second limit value as the respective output value of the neural network depending on the particular situation, which in particular detects environmental conditions and parameters of the respective task. Further preferably, machine learning relies on linear regression to statistically adjust the respective linear factors of a linear equation system such that the result of the linear system of equations provides the first or second threshold.
  • Another aspect of the invention relates to an apparatus for controlling an actuator driven robotic manipulator having an end effector, wherein the end effector is configured to perform a predetermined target movement and to perform a task during the execution of the target movement, comprising: a force determination unit adapted to determine one in the Robot external manipulator K ext is executed during the execution of the desired movement, wherein K ext has a vector F ext at least one external force and / or a vector M ext at least one external moment, a computing unit, which is designed to an undesirable collision of the Robot manipulator to detect when K ext in or around n axes of a predefined coordinate system with N axes exceeds a predefined first limit, where n∈N, and is designed to detect an erroneous execution of the task when K ext i n or m axis of the coordinate system exceeds a predefined second threshold value or if K ext <K of, where K is the one dependent on the object and occur exclusively in, or to the m axis of expected and / or desired force Winder, where K is the has a vector F of the at least one expected and / or desired force and / or a vector M of the at least one expected and / or desired torque, and wherein m∈N and m∉ n, wherein the first threshold is less than the second threshold and a control unit configured to control the robotic manipulator in an error mode when the computing unit detects an undesired collision of the robotic manipulator and / or an erroneous execution of the task.
  • According to a further advantageous embodiment, the device furthermore has a user interface for specifying the first limit value and / or the second limit value.
  • Advantages and preferred developments of the proposed device result from an analogous and analogous transmission of the statements made above in connection with the proposed method.
  • Another aspect of the invention relates to a robot with a device as described above and below.
  • Further advantages, features and details will become apparent from the following description in which - where appropriate, with reference to the drawings - at least one embodiment is described in detail. The same, similar and / or functionally identical parts are provided with the same reference numerals.
  • Show it:
    • 1 a method for controlling an actuator-driven robot manipulator according to an embodiment of the invention, and
    • 2 a robot with a device for controlling an actuator-driven robot manipulator according to another embodiment of the invention.
  • The illustrations in the figures are schematic and not to scale.
  • 1 shows a method for controlling an actuator-driven robot manipulator 1 with an end effector 3 in which the end effector 3 executes a predetermined desired movement and executes a task during the execution of the desired movement. In a first step, a determination is made during execution of the desired movement S1 one in the robot manipulator 1 introduced external power winders K ext instead, where K ext has a vector F ext of at least one external force and / or a vector M ext of at least one external moment. The external force Winder K ext is determined by means of a force and torque sensor. In a further step, a detection takes place S2 an undesired collision of the robot manipulator 1 if K ext in or around n axes of a predefined coordinate system with N axes exceeds a predefined first limit, where n∈N. Furthermore, a detection takes place S3 a faulty execution of the task, if K ext in or near m axes of the coordinate system exceeds a predefined second threshold value or if K ext <K of, where K of a dependent of the task and occur exclusively in, or to the m axis of expected and / or desired forcewinder, and where m∈N and m∉ n. The first limit value is smaller than the second limit value. In the final step, a drive takes place S4 of the robot manipulator 1 in an error mode when an unwanted collision of the robot manipulator 1 and / or a faulty execution of the task is detected. In addition, the predefined coordinate system, along with the m axes and the n axes, is time-variant and dependent upon an execution progress task. The first and / or the second limit value are further specified by a user and can be adapted by the user during the execution of the task.
  • 2 shows a robot 200 with a device 100 for controlling an actuator-driven robot manipulator 1 with an end effector 3 , wherein the end effector 3 to execute a predetermined target movement and to execute a task during execution of the target movement. The end effector 3 here has a drill chuck and a drill clamped therein. The device 100 has a force determination unit 5 on getting one into the robot manipulator 1 introduced external power winders K ext is executed during the execution of the desired movement, wherein K ext has a vector F ext at least one external force and / or a vector M ext at least one external moment. Furthermore, the device 100 an arithmetic unit 7 an executed unwanted collision of the robot manipulator 1 to detect, if K ext in n axes of a predefined coordinate system with N axes exceeds a predefined first limit, where n∈N, and to detect an erroneous execution of the task if K ext in or around m axes of the coordinate system exceeds a predefined second limit, or if K ext <K of, where K is the one dependent on the task and exclusively occurring in or around the axes m expected and / or desired force Winder, and wherein m∈N m∉ and n, wherein the first threshold value smaller than the second Limit is. In this case, the following N axes are considered: "x", "y", "z", "-z", where | N | = 4. As in the z-direction, a feed as a target movement for drilling by means of the end effector 3 arranged drill, and from the component to be drilled, a contact force in the form of an expected and / or desired power wind K of the takes place, the axis "-z" is defined as the single m axis. The other axes "x", "y", "z" are consequently n axes. However, remains from the force of the blast K, i.e. K ext <K of enters, so a faulty execution of the task is detected as discussed above, since the drilling not carried out as planned, for example, if the slip is to be drilled component. Furthermore, the device 100 a control unit 9 on which is carried out, the robot manipulator 1 in a fault mode, if the arithmetic unit 7 an unwanted collision of the robot manipulator 1 and / or a faulty execution of the task detected. The failure mode consists here in a termination of the execution of the task and a standstill of the end effector.
  • Although the invention has been further illustrated and explained in detail by way of preferred embodiments, the invention is not limited by the disclosed examples, and other variations can be derived therefrom by those skilled in the art without departing from the scope of the invention. It is therefore clear that a multitude of possible variations exists. It is also to be understood that exemplified embodiments are really only examples that are not to be construed in any way as limiting the scope, applicability, or configuration of the invention. Rather, the foregoing description and description enable the skilled artisan to practice the exemplary embodiments, and those of skill in the knowledge of the disclosed inventive concept may make various changes, for example, to the function or arrangement of particular elements recited in an exemplary embodiment. without departing from the scope defined by the claims and their legal equivalents, such as further explanation in the specification.
  • LIST OF REFERENCE NUMBERS
  • 1
    robot manipulator
    3
    end effector
    5
    Force determining unit
    7
    computer unit
    9
    control unit
    100
    device
    200
    robot
    S1
    Determine
    S2
    detect
    S3
    detect
    S4
    head for

Claims (10)

  1. Method for controlling an actuator-driven robotic manipulator (1) having an end effector (3) in which the end effector (3) performs a predetermined set movement and executes a task during the execution of the set movement, comprising the steps of: - determining during the execution of the set movement ( S1) of an external power wincher K ext introduced into the robot manipulator (1), wherein K ext has a vector F ext of at least one external force and / or a vector M ext of at least one external moment, - detecting (S2) an undesired collision of the robot manipulator ( 1), if K ext in or around n axes of a predefined coordinate system with N Axes exceeds a predefined first limit value, wherein, n ∈ N, - detecting (S3) a faulty execution of the task, if K ext exceeds a predefined second limit value in or near m axes of the coordinate system or if K ext <K of, where K is an expected and / or desired force-winder dependent on the task and occurring exclusively in or about the m axes, and where m ∈ N and m ∉ n, where the first limit value is less than the second limit value, and - driving (S4 ) of the robot manipulator (1) in an error mode when an undesired collision of the robot manipulator (1) and / or an erroneous execution of the task is detected.
  2. Method according to Claim 1 in which, at least for detecting an erroneous execution of the task, the determined force winder K ext is compensated for by the expected and / or desired force winder K des and the compensated force winder is compared with the second limit value.
  3. Method according to one of the preceding claims, wherein the external force Winder K ext is determined by means of a pulse observer.
  4. Method according to one of the preceding claims, wherein the predefined coordinate system and / or the m axes and / or the n axes are time-variant and dependent on the desired movement and / or the task.
  5. Method according to one of Claims 1 to 4 wherein the first and / or the second limit value are predetermined by a user and are user-adaptable.
  6. Method according to one of Claims 1 to 4 wherein the first and / or the second threshold are defined together with the task.
  7. Method according to one of Claims 1 to 4 wherein the first and / or second thresholds are determined and adjusted by machine learning.
  8. Device (100) for controlling an actuator-driven robotic manipulator (1) having an end effector (3), wherein the end effector (3) is designed to execute a predetermined set movement and to execute a task during the execution of the set movement, comprising: - a force determination unit ( 5), which is designed to determine an external power winch K ext introduced into the robot manipulator (1) during the execution of the desired movement, K ext having a vector F ext of at least one external force and / or a vector M ext of at least one external moment, a computing unit (7) designed to detect an undesired collision of the robot manipulator (1) when K ext in or around n axes of a predefined coordinate system with N axes exceeds a predefined first limit, where n∈N, and to detect an erroneous execution of the task, if K ext in or around m axes of the coordinate ensystems exceeds a predefined second threshold value or if K ext <K of, where K is the one dependent on the task and exclusively occurring in or around the axes m expected and / or desired force Winder, and wherein m∈N and applies m∉n wherein the first threshold is less than the second threshold, and a control unit (9) adapted to drive the robotic manipulator (1) in an error mode when the arithmetic unit (7) detects an undesired collision of the robotic manipulator (1) and / or detected an erroneous execution of the task.
  9. Device (100) according to Claim 8 , further comprising: a user interface for setting the first and / or the second limit value.
  10. Robot (200) with a device (100) according to one of Claims 8 to 9 ,
DE102018112370.1A 2018-05-23 2018-05-23 Direction-dependent collision detection for a robot manipulator Pending DE102018112370A1 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
DE102018112370.1A DE102018112370A1 (en) 2018-05-23 2018-05-23 Direction-dependent collision detection for a robot manipulator

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
DE102018112370.1A DE102018112370A1 (en) 2018-05-23 2018-05-23 Direction-dependent collision detection for a robot manipulator
PCT/EP2019/063298 WO2019224288A1 (en) 2018-05-23 2019-05-23 Direction-dependent collision detection for a robot manipulator

Publications (1)

Publication Number Publication Date
DE102018112370A1 true DE102018112370A1 (en) 2019-11-28

Family

ID=66690319

Family Applications (1)

Application Number Title Priority Date Filing Date
DE102018112370.1A Pending DE102018112370A1 (en) 2018-05-23 2018-05-23 Direction-dependent collision detection for a robot manipulator

Country Status (2)

Country Link
DE (1) DE102018112370A1 (en)
WO (1) WO2019224288A1 (en)

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE102007060680A1 (en) * 2007-12-17 2009-06-18 Kuka Roboter Gmbh Method and device for controlling a manipulator
DE102008010983A1 (en) * 2008-02-25 2009-08-27 Mtu Aero Engines Gmbh Method for optimized near-net shape milling
US20110270443A1 (en) * 2010-04-28 2011-11-03 Kabushiki Kaisha Yaskawa Denki Apparatus and method for detecting contact position of robot
EP1403746B1 (en) * 2002-09-30 2013-01-02 Dürr Systems GmbH Collision detection method
DE102011111758A1 (en) * 2011-08-24 2013-02-28 Dürr Systems GmbH Control method for a robot
DE102014103240A1 (en) * 2014-03-11 2015-10-01 Pro-Micron Gmbh & Co. Kg Method for setting up and / or monitoring operating parameters of a workpiece processing machine
EP2954986A1 (en) * 2014-06-10 2015-12-16 Siemens Aktiengesellschaft Apparatus and method for managing and controlling motion of a multiple body system

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP6100727B2 (en) * 2014-04-09 2017-03-22 ファナック株式会社 Human cooperative industrial robot with read-through function
JP5927284B1 (en) * 2014-12-22 2016-06-01 ファナック株式会社 A robot controller that detects contact force with a person and stops the robot
JP6055002B2 (en) * 2015-02-20 2016-12-27 ファナック株式会社 Human cooperative robot system that retracts the robot
US10228669B2 (en) * 2015-05-27 2019-03-12 Rolls-Royce Corporation Machine tool monitoring
KR20180084841A (en) * 2015-12-01 2018-07-25 카와사키 주코교 카부시키 카이샤 Monitoring system of robot system
DE102016107842B4 (en) * 2016-04-27 2018-09-06 Franka Emika Gmbh Method for inserting an object into an article receptacle by means of a robot manipulator

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP1403746B1 (en) * 2002-09-30 2013-01-02 Dürr Systems GmbH Collision detection method
DE102007060680A1 (en) * 2007-12-17 2009-06-18 Kuka Roboter Gmbh Method and device for controlling a manipulator
DE102008010983A1 (en) * 2008-02-25 2009-08-27 Mtu Aero Engines Gmbh Method for optimized near-net shape milling
US20110270443A1 (en) * 2010-04-28 2011-11-03 Kabushiki Kaisha Yaskawa Denki Apparatus and method for detecting contact position of robot
DE102011111758A1 (en) * 2011-08-24 2013-02-28 Dürr Systems GmbH Control method for a robot
DE102014103240A1 (en) * 2014-03-11 2015-10-01 Pro-Micron Gmbh & Co. Kg Method for setting up and / or monitoring operating parameters of a workpiece processing machine
EP2954986A1 (en) * 2014-06-10 2015-12-16 Siemens Aktiengesellschaft Apparatus and method for managing and controlling motion of a multiple body system

Also Published As

Publication number Publication date
WO2019224288A1 (en) 2019-11-28

Similar Documents

Publication Publication Date Title
Cherubini et al. Collaborative manufacturing with physical human–robot interaction
JP5686775B2 (en) Method for dynamic optimization of robot control interface
Geravand et al. Human-robot physical interaction and collaboration using an industrial robot with a closed control architecture
KR101661810B1 (en) Method for manually guided displacement of the pose of a manupulator arm of an industrial robot and related industrial robot
DE102013113044A1 (en) Electric gripping hand with force sensor
EP2692488B1 (en) Control device and control method for robot and the robot
EP2883665B1 (en) Method and device for controlling a manipulator
EP1267234B1 (en) Method for supervising a technical installation, in particular a manipulator, as well as a supervision and control apparatus
JP3923053B2 (en) Robot teaching device
KR101795847B1 (en) Method for programming an industrial robot and related industrial robot
US20160023355A1 (en) Robot system control method and robot system
CN105034025B (en) The security monitoring device of robot
DE112013003209B4 (en) Robot control device and robot control method
Colomé et al. External force estimation during compliant robot manipulation
US8271134B2 (en) Robot having learning control function
DK2321623T3 (en) Procedure for controling a robot&#39;s brake
DE102012104194B4 (en) Robot and spot welding robot with learning control function
DE102015002348B4 (en) Robot controller with the function to simplify the teaching process and to improve the mobility of the robot
US7212886B2 (en) Robot control apparatus and method
DE102015002764B4 (en) Robot control with detection of contact in external environment
DE102017003000A1 (en) Robotic system cooperating with humans
EP2243602B1 (en) Method and device for controlling a manipulator
US8594847B2 (en) Manipulator, particularly industrial robot, having a redundant sensor arrangement, and method for the control thereof
KR102015664B1 (en) Method and device for executing a manipulator process
JP4202365B2 (en) Force control device

Legal Events

Date Code Title Description
R012 Request for examination validly filed
R016 Response to examination communication