WO2021056591A1 - 用于控制多轴运动机器的方法、装置和存储介质 - Google Patents

用于控制多轴运动机器的方法、装置和存储介质 Download PDF

Info

Publication number
WO2021056591A1
WO2021056591A1 PCT/CN2019/109244 CN2019109244W WO2021056591A1 WO 2021056591 A1 WO2021056591 A1 WO 2021056591A1 CN 2019109244 W CN2019109244 W CN 2019109244W WO 2021056591 A1 WO2021056591 A1 WO 2021056591A1
Authority
WO
WIPO (PCT)
Prior art keywords
motion control
target object
control mode
axis motion
task
Prior art date
Application number
PCT/CN2019/109244
Other languages
English (en)
French (fr)
Inventor
王子健
范顺杰
Original Assignee
西门子股份公司
西门子（中国）有限公司
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 西门子股份公司, 西门子（中国）有限公司 filed Critical 西门子股份公司
Priority to CN201980100024.5A priority Critical patent/CN114341823A/zh
Priority to PCT/CN2019/109244 priority patent/WO2021056591A1/zh
Publication of WO2021056591A1 publication Critical patent/WO2021056591A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F13/00—Interconnection of, or transfer of information or other signals between, memories, input/output devices or central processing units
    • G06F13/10—Program control for peripheral devices

Definitions

  • the present invention relates to a method, a device and a computer-readable storage medium for controlling a multi-axis motion machine, in particular to a method that can control a multi-axis motion machine to automatically switch between multiple motion control modes according to task-level instructions.
  • Device and computer readable storage medium are used to store instructions for controlling a multi-axis motion machine.
  • Multi-axis motion machines such as multi-axis manipulators and collaborative robots, are widely used in modern industrial production to take and deliver workpieces, process workpieces, or cooperate with operators to complete designated tasks.
  • a multi-axis motion controller for controlling the multi-axis motion of the multi-axis motion machine is the core component.
  • multi-axis motion machines When performing different types of tasks, due to safety considerations and the requirements of the application environment, multi-axis motion machines usually need to work in different motion control modes (such as position control mode, impedance control mode, etc.), for example, when required When quickly moving to a specified position, a multi-axis machine usually works in position control mode; when it needs to output an ideal force, a multi-axis machine usually works in impedance control mode.
  • Multi-axis motion machines usually can only switch between the two motion control modes through manual operation or manual programming.
  • manual operation and manual programming specific application scenarios need to be considered and programmers/operators are required to have professional knowledge in related fields.
  • this switching method of the motion control mode of multi-axis motion machines can no longer meet the requirements.
  • Multi-axis can autonomously switch in multiple motion control modes according to task-level instructions. Sports machines are undoubtedly more ideal.
  • the present invention provides a control method for a multi-axis motion machine that can autonomously switch in multiple motion control modes according to task-level instructions.
  • the present invention also provides a multi-axis motion control device for controlling a multi-axis motion machine, the multi-axis motion machine including an end operator capable of moving around a plurality of axes.
  • the present invention also provides a computer-readable storage medium containing the control method.
  • the present invention provides a multi-axis motion control method for controlling a multi-axis motion machine, the multi-axis motion machine including an end operator capable of moving around multiple axes, the method including:
  • Planning a motion trajectory planning the motion trajectory of the end manipulator according to the static environment information obtained in the environmental monitoring subtask;
  • the multiple subtasks also include:
  • a motion control subtask which determines in real time a motion control mode for controlling the end manipulator according to the dynamic environment information obtained in the environmental monitoring subtask;
  • the motion control mode switching subtask is to switch the current motion control mode to the motion control mode determined in the motion control subtask and adjust the corresponding control parameters.
  • the above-mentioned multi-axis motion control method can autonomously switch in multiple motion control modes according to task-level instructions.
  • the initial motion control mode for controlling the end effector is a position control mode.
  • the beneficial technical effect of this preferred embodiment is at least that the end manipulator can quickly move to the target position.
  • the task-level instructions include one or more of the following items: moving the target object, processing the target object, interacting with the target object and avoiding hitting the target object, avoiding hitting the non-target object .
  • the beneficial technical effect of this preferred embodiment is at least that the control method can receive and implement various task-level instructions.
  • the environment monitoring subtask when the task-level instruction includes moving the target object, includes identifying the target object and calculating the current position and the target position of the target object.
  • the beneficial technical effect of this preferred embodiment is at least that the control method can more accurately control the end operator according to the output of the environmental monitoring subtask.
  • the motion control subtask includes determining that the motion control mode for controlling the end manipulator is a position control mode.
  • the beneficial technical effect of this preferred embodiment is at least that the end manipulator can quickly move to the target position.
  • the environmental monitoring subtask when the task-level instruction includes processing the target object, includes identifying the target object, determining the part to be processed of the target object, calculating the position of the part to be processed, and Orientation, and calculate the current position and orientation of the end effector.
  • the beneficial technical effect of this preferred embodiment is at least that the control method can more accurately process the to-be-processed part of the target object according to the output of the environmental monitoring subtask.
  • the motion control subtask includes switching the motion control mode for controlling the end manipulator to the Cartesian space impedance control mode according to the current position of the end manipulator,
  • the motion control mode switching subtask includes adjusting the control parameters corresponding to the Cartesian spatial impedance control mode according to the following formula:
  • ⁇ is the output torque of the power system
  • g(q) is the gravity torque vector function
  • J(q) is the Jacobian matrix
  • ⁇ (x) is the inertia matrix
  • D d is the Cartesian space damping
  • K d is the Cartesian space stiffness
  • F c is the force exerted by the end operator
  • K f is the force selection used to determine the direction of F c Matrix
  • q is the joint position
  • x is the actual position of the end manipulator in Cartesian space
  • x d is the planned position of the end manipulator in Cartesian space
  • the beneficial technical effect of this preferred embodiment is at least that the control method can adjust the control
  • the environmental monitoring subtask when the task-level instructions include interacting with the target object and avoiding hitting the target object, includes identifying the target object, calculating the position of the target object, and calculating the terminal Calculate the current position of the end manipulator and the interaction position between the end manipulator and the target object.
  • the beneficial technical effect of this preferred embodiment is at least that the end manipulator can quickly move to the target position according to the output of the environmental monitoring subtask, and avoid hitting the target object.
  • the motion control subtask includes switching the motion control mode for controlling the end manipulator to the shafting space impedance control mode according to the current position of the end manipulator,
  • the motion control mode switching subtask includes adjusting the control parameters corresponding to the shafting space impedance control mode according to the following formula:
  • ⁇ is the output torque of the power system
  • M(q) is the inertia matrix
  • g(q) is the gravity torque vector function
  • k p and k d are the joint stiffness matrix and joint damping matrix respectively, which can be configured manually
  • q is the actual joint position
  • q * is the ideal joint position
  • Is the ideal joint acceleration is at least that the control method can adjust the control parameters according to the formula so as to better achieve the purpose of interacting with the target object and avoiding the target object from colliding with the target object.
  • the environmental monitoring subtask when the task-level instruction includes avoiding collisions with non-target objects, includes real-time monitoring on the multi-axis motion machine (or the end manipulator) Whether there are non-target objects in the predetermined range around.
  • the beneficial technical effect of this preferred embodiment is at least that the end operator can enter the required control mode in advance according to the output of the environmental monitoring subtask.
  • the motion control subtask includes switching a motion control mode for controlling the end manipulator to a gravity compensation mode, and the motion control mode switching subtask includes adjustment according to the following formula The control parameters corresponding to the gravity compensation mode:
  • ⁇ is the output torque of the power system
  • g(q) is the gravity torque vector function
  • q is the joint position
  • the present invention also provides a multi-axis motion control device for controlling a multi-axis motion machine.
  • the multi-axis motion machine includes an end operator capable of moving around multiple axes.
  • the device includes the following modules:
  • An environmental monitoring module for sensing and determining environmental information around the multi-axis motion machine
  • a task planning module configured to decompose the received task-level instructions into multiple subtasks and plan the movement trajectory of the end operator according to the static environment information obtained by the environment monitoring module;
  • the motion control module is used to determine the motion control mode used to control the end operator in real time according to the dynamic environmental information obtained in the environmental monitoring module, and used to execute the motion control used to control the end operator Control mode
  • the motion control mode switching module is used to switch the current motion control mode of the end manipulator to the determined motion control mode and adjust the corresponding control parameters.
  • the above-mentioned multi-axis motion control device can autonomously switch in multiple motion control modes according to task-level instructions.
  • the motion control module includes an impedance control sub-module for performing an impedance control mode, a gravity compensation sub-module for performing a gravity compensation mode, and a position control sub-module for performing a position control mode.
  • the beneficial technical effect of this preferred embodiment is at least that there are multiple control sub-modules corresponding to multiple control modes to choose from to deal with various situations more flexibly.
  • the motion control mode switching module switches the current motion control mode of the end manipulator to the motion control determined in the task planning module by switching the sub-modules mode.
  • the beneficial technical effect of this preferred embodiment is at least that the motion control mode switching can autonomously switch a suitable motion control mode according to the current situation.
  • the task-level instructions include one or more of the following items: moving the target object, processing the target object, interacting with the target object and avoiding hitting the target object, avoiding hitting the non-target object .
  • the beneficial technical effect of this preferred embodiment is at least that the control device can receive and implement various task-level instructions.
  • the environment monitoring module when the task-level instruction includes moving the target object, includes identifying the target object and calculating the current position and the target position of the target object.
  • the beneficial technical effect of this preferred embodiment is at least that the control device can more accurately control the end operator according to the output of the environmental monitoring module.
  • the environmental monitoring module is used to identify the target object, determine the part to be processed of the target object, calculate the position of the part to be processed, and Orientation, and calculate the current position and orientation of the end effector.
  • the beneficial technical effect of this preferred embodiment is at least that the control device can more accurately process the to-be-processed part of the target object according to the output of the environmental monitoring module.
  • the environmental monitoring module is used to identify the target object, calculate the position of the target object, and calculate the terminal Calculate the current position of the end manipulator and the interaction position between the end manipulator and the target object.
  • the beneficial technical effect of this preferred embodiment is at least that the control device can quickly move the end manipulator to the target position according to the output of the environmental monitoring module and avoid hitting the target object.
  • the environmental monitoring module is used for real-time monitoring of the multi-axis motion machine (or the end manipulator) Whether there are non-target objects in the predetermined range around.
  • the beneficial technical effect of this preferred embodiment is at least that the control device can enter the required control mode in advance according to the output of the environmental monitoring module.
  • the present invention also provides a multi-axis motion control device for controlling a multi-axis motion machine.
  • the multi-axis motion control device includes a memory and a controller, wherein the memory is used to store a computer program, the The controller is configured to be able to implement the control method as described in the above embodiment when the computer program is executed.
  • the multi-axis motion control device can autonomously switch in multiple motion control modes according to task-level instructions.
  • the present invention also provides a computer-readable storage medium in which a computer program is stored.
  • a computer program is stored.
  • the control method as described in the above embodiment can be implemented.
  • the computer-readable storage medium can realize a control method of autonomously switching in multiple motion control modes according to task-level instructions.
  • Fig. 1 schematically shows a flowchart of a multi-axis motion control method for controlling a multi-axis motion machine according to a preferred embodiment of the present invention.
  • Figure 2 schematically shows the first application scenario of the control method of the present invention.
  • Fig. 3 schematically shows a flow chart when the control method of the present invention is applied to the scene shown in Fig. 2.
  • Fig. 4 schematically shows a second application scenario of the control method of the present invention.
  • Fig. 5 schematically shows a flow chart when the control method of the present invention is applied to the scene shown in Fig. 4.
  • Fig. 6 schematically shows a third application scenario of the control method of the present invention.
  • Fig. 7 schematically shows a flow chart when the control method of the present invention is applied to the scene shown in Fig. 6.
  • Fig. 8 schematically shows an architecture diagram of a multi-axis controller for controlling a multi-axis motion machine according to a preferred embodiment of the present invention.
  • the method of the present invention is mainly aimed at multi-axis motion machines or multi-axis motion robots, such as multi-axis manipulators and collaborative robots, which have end-manipulators (ie, have multiple joints) capable of moving around multiple axes, for example Different task types automatically switch the motion control mode of the manipulator.
  • Figure 1 shows a general flow diagram of the multi-axis motion control method according to the present invention, which can be applied to the above-mentioned multi-axis motion machine or multi-axis motion robot to determine and switch the motion control mode of its end manipulator .
  • the control method of the present invention includes a main task S1 and three subtasks S2, S3, and S4.
  • the main task S1 first receive task-level instructions in step 110, for example, move or process the target object (workpiece) through the manipulator, avoid hitting the non-target object (intruder), and the target object (workpiece, operator). ) To interact and avoid colliding with the target object to cause injury, etc.
  • the "target object” in this article generally refers to the task-related planned objects, such as workpieces, pallets, operators, etc.
  • the “non-target object” generally refers to Unplanned objects that have nothing to do with the task, such as intruders.
  • the task-level instructions can be to move the workpiece from one pallet (current position) to another pallet (target position), process the workpiece (for example, drill a hole or screw a screw on a surface of the workpiece, etc.), and avoid collisions.
  • the intruder who enters a certain range around the multi-axis motion machine or the end operator, deliver the workpiece to the operator or pick up the workpiece from the operator and avoid hurting the operator, and any combination of the above instruction types .
  • step 120 the received instruction is interpreted and decomposed into three subtasks: the environmental monitoring subtask S2, the motion control subtask S3, and the motion control mode switching subtask S4.
  • the control in the motion control subtask S3 and the motion control mode switching subtask S4 is corrected in real time through the dynamic environment information obtained in the environmental monitoring subtask S2.
  • step 130 the motion trajectory of the end manipulator is planned according to the static environmental information obtained in the environmental monitoring subtask (for example, the current position and posture of the workpiece to be moved and the target position and posture, the position of the operator, etc.).
  • step 210 one or more sensors are used to obtain environmental data around the multi-axis motion machine.
  • the environmental data may be, for example, colors or 3D images.
  • step 220 identify target objects (such as workpieces or operators) and/or monitor whether there are non-target objects (such as intruders) within a safe range.
  • Machine learning methods such as deep learning, pattern-based object recognition can be used. ) And so on.
  • step 230 the position of the above-mentioned object is calculated, and the position of the target object is provided as static environment information to step 130 to plan the movement trajectory of the end manipulator, and at the same time, whether there is a non-target object and the position of the non-target object (if there is If the end manipulator has approached or left the operator, whether it has reached the processing start position and the processing end position, etc., are provided as dynamic environment information to steps 320 and 340 to determine whether the switching conditions are met.
  • step 310 the initial motion control mode of the end manipulator is determined.
  • the initial motion control mode is generally a position control mode, so that the end manipulator quickly moves to a predetermined position in response to an instruction.
  • step 320 based on the dynamic environment data obtained in step 230, it is determined whether the switching condition is satisfied. If the switching condition is not satisfied, step 310 is returned to continue to execute the position control mode. If the switching conditions are met, it enters the motion control mode switching subtask S4.
  • step 410 a suitable motion control mode is selected based on the instruction type.
  • the position control mode can be selected; for processing (such as drilling or screwing) the target object (such as a workpiece)
  • the target object such as a workpiece
  • Cartesian space impedance control mode For instructions, you can select Cartesian space impedance control mode; for instructions to avoid hitting non-target objects (such as intruders), you can select gravity compensation control mode; for interacting with target objects (such as workpieces, operators, etc.) and avoiding collisions
  • the control parameters can be adjusted according to the formulas corresponding to different motion control modes, and in step 430, the motion control mode with appropriate control parameters is provided to step 330 in subtask S3.
  • step 340 of the motion control subtask S3 it is further judged whether the above switching conditions are still met. If it is still met, the motion control mode just switched before is still adopted; if it is no longer met, then go to step 350 to resume the initial motion Control mode.
  • priority can be set for the above-mentioned multiple motion control modes to determine the control mode that is preferentially switched.
  • the motion control mode is preferentially switched from the position control mode to the gravity compensation control mode (to avoid hitting the intruder), and when the intruder leaves the safe range, then switch from the gravity compensation control mode to the shafting space impedance control Mode (interact with the operator and avoid hitting the operator).
  • FIG. 2 schematically show the first application scenario and the corresponding control method flow chart.
  • the motion control of the end manipulator 10 is realized by the multi-axis motion control device 30 according to the present invention.
  • the task instruction is to move the workpieces 41, 42 from the tray 40 to the tray 50 through the end operator 10, and place them in a place that does not interfere with the existing workpieces 51 and 52 in the tray 50.
  • step 111 the instruction is received in step 111, and then the instruction is interpreted in step 121 and decomposed into three subtasks: environmental monitoring subtask S21, motion control subtask S31, and motion control mode Switch subtask S41.
  • step 131 the movement trajectory of the end operator 10 is planned according to the static environmental information obtained in the environmental monitoring subtask, that is, the positions of the tray 40, the workpieces 41 and 42, the tray 50, and the workpieces 51 and 52.
  • step 211 one or more sensors 20 are used to obtain environmental data around the multi-axis motion machine, such as three-dimensional images of pallets, workpieces, and intruders.
  • step 221 the type of the workpiece to be moved and the type of the existing workpiece in the target pallet are identified to determine the size, shape, etc. of the workpiece.
  • step 231 the take-in position of the workpiece is calculated and the result is fed back to step 131 to plan the movement trajectory of the end manipulator.
  • step 241 it is monitored whether there is an intruder 60 within the safety range of the end manipulator 10, and the information is provided to the motion control subtask S31.
  • the initial motion control mode of the end manipulator is determined in step 311.
  • the initial motion control mode is generally a position control mode, so that the end manipulator quickly moves to a predetermined position in response to an instruction.
  • step 321 it is determined whether there is an intruder based on the monitoring result of step 241. If there is no intruder, the end operator 10 is continuously controlled in the position control mode; if there is an intruder, the end operator is activated in step 331 10Stop the exercise and enter the exercise control mode switching subtask S41.
  • the gravity compensation control mode is selected.
  • step 421 the corresponding control parameters are adjusted according to the following formula (1), and in step 431, the gravity compensation control mode with appropriate control parameters is provided to step 341 in subtask S31. So far, the movement of the end operator The control mode has been switched to the gravity compensation control mode. In the gravity compensation control mode, if the intruder comes into contact with the end manipulator, the end manipulator will not apply (feedback) any force to the intruder or only apply (feedback) ) Very little force.
  • step 351 of the motion control subtask S31 it is further judged whether the above switching conditions are still met, if it is still met, the gravity compensation control mode is still adopted; if it is no longer met, then go to step 361 to restore the initial motion control mode , That is, position control mode, the above steps will be repeated.
  • ⁇ is the output torque of the power system
  • g(q) is the gravity torque vector function
  • q is the joint position
  • the output of the motion control device 30 will change from the motor speed setting value plus the torque feed forward to the motor torque, so that no additional force will be applied in the event of a collision with an intruder to avoid injury Intruder.
  • Figures 4 and 5 schematically show the second application scenario and the corresponding control method flow chart.
  • the motion control of the end manipulator 10 is realized by the multi-axis motion control device 30 according to the present invention.
  • the task instruction is to deliver the workpieces 41, 42 from the pallet 40 to the operator 70 through the end manipulator 10, and the operator 70 places the workpieces 41, 42 on the pallet 50.
  • the difference from the first scenario described above is that the end manipulator 10 needs to actively transfer the workpiece to the operator 70 and cannot harm the operator during the transfer process, while the operator 10 must maintain efficient operation (not Stop exercising).
  • step 112 the instruction is received in step 112, and then the instruction is interpreted in step 122 and decomposed into three subtasks: environmental monitoring subtask S22, motion control subtask S32, and motion control mode Switch subtask S42.
  • step 132 according to the static environmental information obtained in the environmental monitoring subtask, that is, the position of the tray 40, the workpieces 41 and 42, the tray 50, the workpieces 51 and 52, the position of the operator 70, etc., the movement of the end operator 10 is planned Trajectory.
  • step 212 one or more sensors 20 are used to obtain environmental data around the multi-axis motion machine, such as three-dimensional images of pallets, workpieces, and operators.
  • step 222 the types of obstacles and workpieces to be moved are identified to determine the size, shape, etc. of the workpieces.
  • step 232 the operator 70 is identified, and the operator can be identified from beginning to end using an image processing method or a machine learning method, and the information is provided to the motion control subtask S32.
  • step 242 the delivery position of the workpiece and the interaction position with the operator 70 are calculated, and the result is fed back to step 132 to plan the movement trajectory of the end operator.
  • the initial motion control mode of the end manipulator is determined in step 312.
  • the initial motion control mode is generally a position control mode, so that the end manipulator quickly moves to a predetermined position in response to an instruction.
  • step 322 based on the recognition result of step 232, it is determined whether the operator 70 has been approached. If it is still not approached, the end operator 10 will continue to be controlled in the position control mode; if approached, it will enter the motion control mode switching subtask. S42.
  • step 412 the shafting space impedance control mode is selected, that is, the end manipulator 10 has a certain degree of elasticity like a spring when it moves around the shaft (joint), and can give a certain feedback to the operator when it comes into contact with the operator 70 The force is large and small, but the operator 70 will not be injured.
  • the corresponding control parameters are adjusted according to the following formula (2), and in step 432, the shaft system space impedance control mode with appropriate control parameters is provided to step 332 in subtask S32. So far, the movement of the operator The control mode has been switched to the shafting space impedance control mode.
  • step 342 of the motion control subtask S32 it is further determined whether the end manipulator 10 has been far away from the operator 70. If it is not far away, the shafting space impedance control mode is still adopted; if it is far away, then go to step 352, To restore the initial motion control mode, that is, the position control mode, the above steps will be repeated.
  • ⁇ is the output torque of the power system
  • M(q) is the inertia matrix
  • g(q) is the gravity torque vector function
  • k p and k d are the joint stiffness matrix and joint damping matrix respectively, which can be configured manually
  • q is the actual joint position
  • q * is the ideal joint position
  • Is the ideal joint acceleration is the ideal joint acceleration.
  • the output of the motion control device 30 will change from the motor speed setting value plus the torque feed forward to the motor torque, so that when it comes into contact with the operator, it will yield in the direction of the axis (joint) movement. , To avoid injury to the operator.
  • Fig. 6 and Fig. 7 schematically show a third application scenario and a flow chart of the corresponding control method.
  • the motion control of the end manipulator 10 is realized by the multi-axis motion control device 30 according to the present invention.
  • the task instruction is to use an end operator (drill or screwdriver) 10 to drill holes or screw screws on the inclined surface 81 in a direction perpendicular to the inclined surface 81 of the part 80 to be processed with a predetermined force.
  • the coordinate systems represented by x0 and z0 are the coordinate systems of the base of the collaborative robot having the end manipulator 10
  • the coordinate systems represented by x and z are the coordinate systems of the end manipulator.
  • step 113 the instruction is received in step 113, and then the instruction is interpreted in step 123 and decomposed into three subtasks: environmental monitoring subtask S23, motion control subtask S33, and motion control mode Switch subtask S43.
  • step 133 according to the static environmental information obtained in the environmental monitoring subtask, that is, the position of the part to be processed 80, the position and orientation of the part to be processed 81, etc., the movement trajectory of the drill or screwdriver 10 is planned.
  • step 213 one or more sensors 20 obtain environmental data around the multi-axis motion machine.
  • image processing and machine learning methods may be used to identify the surface to be processed and the inclination angle of the surface.
  • step 233 calculate the entry point, that is, the position at a certain distance vertically above the surface to be processed; and calculate the target point (ie, the ideal hole depth) or determine the feedback torque of a predetermined magnitude, and feed back the calculation result to step 133 To plan the movement trajectory of the end manipulator.
  • step 243 the coordinate system of the end manipulator (drill or screwdriver) is converted into the base coordinate system and the result is provided to the motion control subtask S33.
  • the initial motion control mode of the end manipulator 10 is determined in step 313.
  • the initial motion control mode is generally a position control mode, so that the end manipulator quickly moves to a predetermined position in response to an instruction, that is, Entry point.
  • step 323 based on the calculation result of step 243, it is determined whether the end operator has reached the entry point (ie, the position to start drilling or screwing), if not, continue to control the end operator in the position control mode 10; If it has arrived, enter the motion control mode switching subtask S43.
  • the Cartesian spatial impedance control mode is selected, so that the end effector 10 has a certain elasticity like a spring in its advancing direction.
  • step 423 the corresponding control parameters are adjusted according to the following formula (3), and in step 433, the Cartesian spatial impedance control mode with appropriate control parameters is provided to step 333 in subtask S33, so far, the end operator The motion control mode of 10 has been switched to Cartesian spatial impedance control mode.
  • step 343 of the motion control subtask S33 it is further determined whether the end manipulator 10 has reached the target point, for example, whether the drill bit has reached a predetermined drilling depth or whether the screwed screw feeds back a predetermined size to the end manipulator If the torque is not reached, the Cartesian space impedance control mode is still used to continue the operation; if it has been reached, the process goes to step 353 to restore the initial motion control mode, that is, the position control mode withdraws the end operator 10.
  • the above steps can be repeated to drill multiple holes or screw multiple screws.
  • ⁇ is the output torque of the power system; g(q) is the gravity torque vector function; Is the joint friction function; J(q) is the Jacobian matrix; ⁇ (x) is the inertia matrix; Is the nonlinear term in Cartesian space; D d is the Cartesian space damping; K d is the Cartesian space stiffness; F c is the force exerted by the end operator; K f is the force selection used to determine the direction of F c Matrix; q is the joint position; Is the joint speed; x is the actual position of the end manipulator in Cartesian space; x d is the planned position of the end manipulator in Cartesian space; Is the planned speed of the end manipulator in Cartesian space; Is the planned acceleration of the end manipulator in Cartesian space; Is the position error of the end effector in Cartesian space.
  • the present invention also provides a multi-axis motion control device or controller for controlling a multi-axis motion machine based on the above control method, and its schematic structure is shown in the dashed box in FIG. 8.
  • the device includes an environment monitoring module, a task planning module, a motion control module and a motion control mode switching module.
  • the environmental monitoring module 2000 is used to collect environmental data and identify targets (for example, workpieces, operators, intruders, obstacles, etc.), and the collected environmental data is one of the basic information used by the task planning module to plan tasks .
  • the environmental monitoring module 2000 sends a sensor control instruction 2010 to the sensor 20, and the sensor 20 feeds back the sensing data 21 to the environmental monitoring module 2000 according to the instruction 2010.
  • the task planning module is used to decompose the received task-level instructions into multiple subtasks and perform autonomous task planning according to the static environmental information collected by the environmental monitoring module 2000 (for example, planning the movement trajectory of the end operator, etc.).
  • the task planning module includes an autonomous task planning module 1100 for interpreting instructions and splitting subtasks, and an autonomous path planning module 1200 for planning the movement trajectory of the end manipulator.
  • the motion control module is used to determine in real time the motion control mode for controlling the end manipulator according to the dynamic environment information obtained in the environmental monitoring module and is used to execute the motion control mode for controlling the end manipulator, It includes a forward & reverse kinematics module 3100, a kinematics interpolation module 3200, a dynamics model module 3300, and a position feedback module 3400.
  • the dynamics model module 3300 includes an impedance control sub-module 3310, A gravity compensation sub-module 3320 for performing gravity compensation control and a torque feedforward sub-module or position control sub-module 3330 for performing position control.
  • the motion control mode switching module is used to switch the current motion control mode of the end manipulator to the determined motion control mode and adjust the corresponding control parameters.
  • the motion control mode switching module includes a motion control mode selection module 4100 for selecting a motion control mode and a control parameter adjustment module 4200 for adjusting control parameters.
  • the motion control mode switching module further includes a motion control mode switching switch 4300 for switching the motion control mode and a communication module 4400 for communicating with the servo drive device 90.
  • the servo drive device 90 is connected to the motor 100 for controlling the motor 100.
  • the servo drive device 90 is also used to feed back the actual position, speed and torque data 91 of the motor to the communication module 4400.
  • the forward & reverse kinematics module 3100 and the kinematic interpolation module 3200 are used to interpolate and calculate the motion trajectory of the end manipulator and generate the theoretical position, velocity and acceleration of the joint on a regular basis (for example, every 2ms) 3210 Provided to the dynamic model module 3300.
  • the impedance control sub-module 3310 and the gravity compensation sub-module 3320 calculate the theoretical joint torque according to the theoretical joint position, velocity and acceleration 3210 and respectively output the motor torque 3311 and 3321 for controlling the motor 100.
  • the position control sub-module 3330 calculates the theoretical joint torque according to the theoretical joint position, velocity and acceleration 3210 and outputs the motor feedforward torque 3331 for controlling the motor 100 to improve the dynamic performance of the system.
  • the position feedback module 3400 is used to monitor the position error between the theoretical position and the actual position of the end operator, calculate the theoretical speed adjustment value 3410 and add it to the motor speed setting value 3420.
  • the present invention also provides a multi-axis motion control device for controlling a multi-axis motion machine.
  • the multi-axis motion control device includes a memory and a controller, wherein the memory is used to store a computer program, the The controller is configured to be able to implement the control method as described above when the computer program is executed.
  • the present invention also provides a computer-readable storage medium in which a computer program is stored, and when the computer program is executed by a controller, the above-mentioned control method can be implemented.

Abstract

一种用于控制多轴运动机器的多轴运动控制方法，所述方法包括：接收任务级指令；将任务级指令分解为多个子任务，所述多个子任务包括环境监测子任务，在环境监测子任务中，感测并确定多轴运动机器周围的环境信息；规划运动轨迹，根据在环境监测子任务中得到的环境信息规划多轴运动机器的端部操作器的运动轨迹；所述多个子任务还包括：运动控制子任务，根据在环境监测子任务中得到的环境信息确定用于控制端部操作器的运动控制模式；运动控制模式切换子任务，将当前的运动控制模式切换成在运动控制子任务中确定的运动控制模式并调整对应的控制参数。

Description

用于控制多轴运动机器的方法、装置和存储介质 技术领域
本发明涉及一种用于控制多轴运动机器的方法、装置和计算机可读存储介质，尤其是一种能够根据任务级指令控制多轴运动机器在多种运动控制模式之间自主切换的方法、装置和计算机可读存储介质。
背景技术
多轴运动机器，如多轴机械手和协作机器人等，广泛应用于现代工业生产中，用于取送工件、加工工件或与操作员协作完成指定任务。在多轴运动机器中，用于控制多轴运动机器的多轴运动的多轴运动控制器是核心部件。在执行不同类型的任务时，出于安全性的考虑以及应用环境的要求，多轴运动机器通常需要在不同的运动控制模式(例如位置控制模式、阻抗控制模式等)下工作，例如，当需要快速移动到指定位置时，多轴机器通常要在位置控制模式下工作；当需要输出理想大小的力时，多轴机器通常要在阻抗控制模式下工作。
现有的多轴运动机器通常只能通过手动操作或人工编程在两种运动控制模式之间进行切换。而对于手动操作和人工编程这两种切换方式，需要考虑具体的应用场景并且要求程序员/操作员具有相关领域的专业知识。随着柔性制造的发展，多轴运动机器的运动控制模式的这种切换方法已不能满足要求，无需手动操作和程序改造、能够根据任务级指令在多种运动控制模式下自主进行切换的多轴运动机器无疑是更加理想的。
发明内容
本发明针对以上现有技术中存在的问题，提供一种能够根据任务级指令在多种运动控制模式下自主进行切换的用于多轴运动机器的控制方法。本发明还提供一种用于控制多轴运动机器的多轴运动控制装置，所述多轴运动机器包括能够围绕多个轴线运动的端部操作器。本发明还提供包含所述控制方法的计算机可读存储介质。
一方面，本发明提供一种用于控制多轴运动机器的多轴运动控制方法，所述多轴运动机器包括能够围绕多个轴线运动的端部操作器，所述方法包括：
接收任务级指令；
将所述任务级指令分解为多个子任务，所述多个子任务包括环境监测子任务，在所述环境监测子任务中，感测并确定所述多轴运动机器周围的环境信息；
规划运动轨迹，根据在所述环境监测子任务中得到的静态环境信息规划所述端部操作器的运动轨迹；
所述多个子任务还包括：
运动控制子任务，根据在所述环境监测子任务中得到的动态环境信息实时确定用于控制所述端部操作器的运动控制模式；
运动控制模式切换子任务，将当前的运动控制模式切换成在所述运动控制子任务中确定的运动控制模式并调整对应的控制参数。
上述多轴运动控制方法能够根据任务级指令在多种运动控制模式下自主进行切换。
根据本发明的一种优选实施方式，用于控制所述端部操作器的初始运动控制模式为位置控制模式。本优选实施方式的有益技术效果至少在于，端部操作器能够快速运动到目标位置。
根据本发明的一种优选实施方式，所述任务级指令包括以下项目中的一个或多个：移动目标对象、加工目标对象、与目标对象交互并避免撞伤目标对象、避免撞伤非目标对象。本优选实施方式的有益技术效果至少在于，所述控制方法能够接收并实施多种任务级指令。
根据本发明的一种优选实施方式，当所述任务级指令包括移动目标对象时，所述环境监测子任务包括识别目标对象、计算目标对象的当前位置和目标位置。本优选实施方式的有益技术效果至少在于，所述控制方法能够根据环境监测子任务的输出更准确地对端部操作器进行控制。
根据本发明的一种优选实施方式，所述运动控制子任务包括确定用于控制所述端部操作器的运动控制模式为位置控制模式。本优选实施方式的有益技术效果至少在于，端部操作器能够快速运动到目标位置。
根据本发明的一种优选实施方式，当所述任务级指令包括加工目标对象时，所述环境监测子任务包括识别目标对象、确定目标对象的待加工部位、计算所述待加工部位的位置和朝向、以及计算所述端部操作器的当前位置和朝向。本优选实施方式的有益技术效果至少在于，所述控制方法能够根据环境监测子任务的输出更准确地对目标对象的待加工部位进行加工。
根据本发明的一种优选实施方式，所述运动控制子任务包括根据所述端部操作器的当前位置将用于控制所述端部操作器的运动控制模式切换为笛卡尔空间阻抗控制模式，所述运动控制模式切换子任务包括根据以下公式调整对应于笛卡尔空间阻抗控制模式的控制参数：
Figure PCTCN2019109244-appb-000001
其中，τ为动力系统的输出扭矩；g(q)为重力扭矩向量函数；
Figure PCTCN2019109244-appb-000002
为关节摩擦力函数；J(q)为雅可比矩阵；Λ(x)为惯性矩阵；
Figure PCTCN2019109244-appb-000003
为笛卡尔空间中的非线性项；D d为笛卡尔空间阻尼；K d为笛卡尔空间刚度；F c为端部操作器施加的力；K f为用于确定F c的方向的力选择矩 阵；q为关节位置；
Figure PCTCN2019109244-appb-000004
为关节速度；x为端部操作器在笛卡尔空间中的实际位置；x d为端部操作器在笛卡尔空间中的规划位置；
Figure PCTCN2019109244-appb-000005
为端部操作器在笛卡尔空间中的规划速度；
Figure PCTCN2019109244-appb-000006
为端部操作器在笛卡尔空间中的规划加速度；
Figure PCTCN2019109244-appb-000007
为端部操作器在笛卡尔空间中的位置误差。本优选实施方式的有益技术效果至少在于，所述控制方法能够根据公式调整控制参数以更有效率地对目标对象的待加工部位进行加工。
根据本发明的一种优选实施方式，当所述任务级指令包括与目标对象交互并避免撞伤目标对象时，所述环境监测子任务包括识别目标对象、计算目标对象的位置、计算所述端部操作器的当前位置、计算所述端部操作器与目标对象的交互位置。本优选实施方式的有益技术效果至少在于，端部操作器能够根据环境监测子任务的输出在快速运动到目标位置的同时避免撞伤目标对象。
根据本发明的一种优选实施方式，所述运动控制子任务包括根据所述端部操作器的当前位置将用于控制所述端部操作器的运动控制模式切换为轴系空间阻抗控制模式，所述运动控制模式切换子任务包括根据以下公式调整对应于轴系空间阻抗控制模式的控制参数：
Figure PCTCN2019109244-appb-000008
其中，τ为动力系统的输出扭矩；M(q)为惯性矩阵；
Figure PCTCN2019109244-appb-000009
为科里奥利/离心矩阵；g(q)为重力扭矩向量函数；
Figure PCTCN2019109244-appb-000010
为关节摩擦力函数；k p和k d分别为关节刚度矩阵和关节阻尼矩阵，可以人工配置；q为实际关节位置；
Figure PCTCN2019109244-appb-000011
为实际关节速度；q *为理想关节位置；
Figure PCTCN2019109244-appb-000012
为理想关节速度；
Figure PCTCN2019109244-appb-000013
为理想关节加速度。本优选实施方式的有益技术效果至少在于，所述控制方法能够根据公式调整控制参数以更好地实现与目标对象进行交互且避免撞伤目标对象的目的。
根据本发明的一种优选实施方式，当所述任务级指令包括避免撞伤非目标对象时，所述环境监测子任务包括实时监测在所述多轴运动机器(或所述端部操作器)周围的预定范围内是否存在非目标对象。本优选实施方式的有益技术效果至少在于，端部操作器能够根据环境监测子任务的输出提前进入所需的控制模式。
根据本发明的一种优选实施方式，所述运动控制子任务包括将用于控制所述端部操作器的运动控制模式切换为重力补偿模式，所述运动控制模式切换子任务包括根据以下公式调整对应于重力补偿模式的控制参数：
Figure PCTCN2019109244-appb-000014
其中，τ为动力系统的输出扭矩；g(q)为重力扭矩向量函数；
Figure PCTCN2019109244-appb-000015
为关节摩擦力函数；q为关节位置；
Figure PCTCN2019109244-appb-000016
为关节速度。本优选实施方式的有益技术效果至少在于，所述控制方法能够根据公式调整控制参数以更好地实现避免撞伤非目标对象的目的。
另一方面，本发明还提供一种用于控制多轴运动机器的多轴运动控制装置，所述多轴运 动机器包括能够围绕多个轴线运动的端部操作器，所述装置包括如下模块：
环境监测模块，用于感测并确定所述多轴运动机器周围的环境信息；
任务规划模块，用于将接收到的任务级指令分解为多个子任务并根据所述环境监测模块得到的静态环境信息规划所述端部操作器的运动轨迹；
运动控制模块，用于根据在所述环境监测模块中得到的动态环境信息实时确定用于控制所述端部操作器的运动控制模式，以及用于执行用于控制所述端部操作器的运动控制模式；
运动控制模式切换模块，用于将所述端部操作器的当前的运动控制模式切换成所确定的运动控制模式并调整对应的控制参数。
上述多轴运动控制装置能够根据任务级指令在多种运动控制模式下自主进行切换。
根据本发明的一种优选实施方式，所述运动控制模块包括用于执行阻抗控制模式的阻抗控制子模块、用于执行重力补偿模式的重力补偿子模块、用于执行位置控制模式的位置控制子模块。本优选实施方式的有益技术效果至少在于，有对应于多种控制模式的多个控制子模块可供选择以更灵活地应对各种情形。
根据本发明的一种优选实施方式，所述运动控制模式切换模块通过切换所述子模块来将所述端部操作器的当前的运动控制模式切换成在所述任务规划模块中确定的运动控制模式。本优选实施方式的有益技术效果至少在于，运动控制模式切换能够根据当前的情形自主切换适合的运动控制模式。
根据本发明的一种优选实施方式，所述任务级指令包括以下项目中的一个或多个：移动目标对象、加工目标对象、与目标对象交互并避免撞伤目标对象、避免撞伤非目标对象。本优选实施方式的有益技术效果至少在于，所述控制装置能够接收并实施多种任务级指令。
根据本发明的一种优选实施方式，当所述任务级指令包括移动目标对象时，所述环境监测模块包括识别目标对象、计算目标对象的当前位置和目标位置。本优选实施方式的有益技术效果至少在于，所述控制装置能够根据环境监测模块的输出更准确地对端部操作器进行控制。
根据本发明的一种优选实施方式，当所述任务级指令包括加工目标对象时，所述环境监测模块用于识别目标对象、确定目标对象的待加工部位、计算所述待加工部位的位置和朝向、以及计算所述端部操作器的当前位置和朝向。本优选实施方式的有益技术效果至少在于，所述控制装置能够根据环境监测模块的输出更准确地对目标对象的待加工部位进行加工。
根据本发明的一种优选实施方式，当所述任务级指令包括与目标对象交互并避免撞伤目标对象时，所述环境监测模块用于识别目标对象、计算目标对象的位置、计算所述端部操作器的当前位置、计算所述端部操作器与目标对象的交互位置。本优选实施方式的有益技术效 果至少在于，所述控制装置能够根据环境监测模块的输出使端部操作器在快速运动到目标位置的同时避免撞伤目标对象。
根据本发明的一种优选实施方式，当所述任务级指令包括避免撞伤非目标对象时，所述环境监测模块用于实时监测在所述多轴运动机器(或所述端部操作器)周围的预定范围内是否存在非目标对象。本优选实施方式的有益技术效果至少在于，所述控制装置能够根据环境监测模块的输出提前进入所需的控制模式。
又一方面，本发明还提供一种用于控制多轴运动机器的多轴运动控制装置，所述多轴运动控制装置包括存储器和控制器，其中，所述存储器用于存储计算机程序，所述控制器被配置成能够在执行所述计算机程序时实施如以上实施方式所述的控制方法。该多轴运动控制装置能够根据任务级指令在多种运动控制模式下自主进行切换。
进一步地，本发明还提供一种计算机可读存储介质，所述计算机可读存储介质中存储有计算机程序，当所述计算机程序被控制器执行时，能够实施如以上实施方式所述的控制方法。该计算机可读存储介质能够实现根据任务级指令在多种运动控制模式下自主进行切换的控制方法。
附图说明
以下附图仅旨在于对本发明做示意性说明和解释，并不限定本发明的范围。为使图面简洁，各图中只示意性地标示出了与本发明相关的部分，它们并不代表其作为产品的实际结构或架构。为使图面便于理解，在有些图中具有相同结构或功能的部分，用相同或相似的附图标记标示。另外，为使图面简洁，在有些图中具有相同结构或功能的部分，仅示意性地绘示了其中的一个，或仅标出了其中的一个。在附图中，
图1示意性地示出了根据本发明的一种优选实施方式的用于控制多轴运动机器的多轴运动控制方法的流程图。
图2示意性地示出了本发明控制方法的第一种应用场景。
图3示意性地示出了本发明控制方法应用于图2所示场景时的流程图。
图4示意性地示出了本发明控制方法的第二种应用场景。
图5示意性地示出了本发明控制方法应用于图4所示场景时的流程图。
图6示意性地示出了本发明控制方法的第三种应用场景。
图7示意性地示出了本发明控制方法应用于图6所示场景时的流程图。
图8示意性地示出了根据本发明的一种优选实施方式的用于控制多轴运动机器的多轴控制器的架构图。
附图标记：10-端部操作器；20-传感器；30-多轴运动控制装置；40、50-托盘；41、42、51、52-工件；60-闯入者；70-操作员；80-待加工部件；81-待加工部位
具体实施方式
为了对本发明的技术特征、目的和效果有更加清楚的理解，现对照附图说明本发明的具体实施方式，在各图中相同的标号表示结构相同或结构相似但功能相同的部件。
在本文中，“示意性”表示“充当实例、例子或说明”，不应将在本文中被描述为“示意性”的任何图示、实施方式解释为一种更优选的或更具优点的技术方案。
本发明的方法主要针对具有能够围绕多个轴进行运动的端部操作器(即，具有多个关节)的多轴运动机器或多轴运动机器人，例如多轴机械手和协作机器人等，用于针对不同的任务类型自主切换操作器的运动控制模式。
图1所示为根据本发明的多轴运动控制方法的一般性流程示意图，其可以应用于如上所述的多轴运动机器或多轴运动机器人，确定并切换其端部操作器的运动控制方式。如图所示，本发明的控制方法包括主任务S1和三个子任务S2、S3、S4。在主任务S1中，首先在步骤110中接收任务级指令，例如，通过操作器移动或加工目标对象(工件)、避免撞伤非目标对象(闯入者)、与目标对象(工件、操作员)进行交互并且避免碰撞该目标对象使其受伤等，本文中的“目标对象”一般是指与任务相关的计划内对象，如工件、托盘、操作员等，而“非目标对象”一般是指与任务无关的非计划内对象，如闯入者等。具体地，任务级指令可以是将工件从一个托盘(当前位置)移动到另一个托盘(目标位置)、对工件进行加工(例如在工件的一个表面上钻孔或旋拧螺丝等)、避免撞伤进入到多轴运动机器或端部操作器周围一定范围内的闯入者、将工件递送到操作员手中或从操作员手中接取工件并且避免碰伤操作员、以及上述指令类型的任意组合。在步骤120中，解读接收到的指令并将其分解为三个子任务：环境监测子任务S2、运动控制子任务S3和运动控制模式切换子任务S4。在后续操作步骤中，通过环境监测子任务S2中得到的动态环境信息对运动控制子任务S3和运动控制模式切换子任务S4中的控制进行实时修正。在步骤130中，根据在环境监测子任务中得到的静态环境信息(例如待移动工件的当前位置及姿态和目标位置及姿态、操作员的位置等)规划端部操作器的运动轨迹。
在环境监测子任务S2中，在步骤210通过一个或多个传感器获取多轴运动机器周围的环境数据，所述环境数据可以是例如颜色或3D图像等。在步骤220中，识别目标对象(例如工件或操作员)并且/或者监测在安全范围内是否存在非目标对象(例如闯入者)，可以利用机器学习方法(例如深度学习、基于模式的对象识别)等实现。在步骤230中，计算上述对象 的位置，将目标对象的位置作为静态环境信息提供至步骤130以规划端部操作器的运动轨迹，同时将是否存在非目标对象以及非目标对象的位置(如果存在的话)、端部操作器是否已接近或离开操作员、是否已到达加工开始位置和加工结束位置等作为动态环境信息提供至步骤320和340以用于判断是否满足切换条件。
在运动控制子任务S3中，在步骤310中，确定端部操作器的初始运动控制模式，初始运动控制模式一般为位置控制模式，以使端部操作器响应于指令快速运动到预定位置。在步骤320中，基于步骤230得到的动态环境数据判断是否满足切换条件，如不满足切换条件，则重新回到步骤310，继续执行位置控制模式。如满足切换条件，则进入到运动控制模式切换子任务S4。在步骤410中，基于指令类型选择合适的运动控制模式，例如，对于移动目标对象(如工件)指令，可以选择位置控制模式；对于加工(例如钻孔或旋拧螺丝)目标对象(如工件)指令，可以选择笛卡尔空间阻抗控制模式；对于避免撞伤非目标对象(如闯入者)指令，可以选择重力补偿控制模式；对于与目标对象(例如工件、操作员等)进行交互并且避免撞伤该目标对象的指令，可以选择轴系空间阻抗控制模式。在步骤420中，可以根据不同运动控制模式所对应的公式调整控制参数，并在步骤430中将具有合适控制参数的运动控制模式提供至子任务S3中的步骤330，至此，操作器的运动控制模式已切换成所需的控制模式。在运动控制子任务S3的步骤340中，进一步判断上述切换条件是否仍然得到满足，如果仍然满足，则仍然采取此前刚刚切换的运动控制模式；如果不再满足，则来到步骤350，恢复初始运动控制模式。当上述类型的指令以组合形式存在时，可以对上述多种运动控制模式设置优先级，以确定优先切换的控制模式。例如，当指令为将工件从托盘递送到操作员手中并且避免撞伤在此期间可能出现的闯入者时，如端部操作器已经接近与操作员的交互位置并且同时在安全范围内出现了闯入者，则运动控制模式从位置控制模式优先切换为重力补偿控制模式(避免撞伤闯入者)，当闯入者离开安全范围之后，再从重力补偿控制模式切换为轴系空间阻抗控制模式(与操作员交互并且避免撞伤操作员)。
以下将结合几种不同的应用场景对本发明的控制方法进行举例介绍。图2和图3示意性地示出了第一种应用场景和对应的控制方法流程图。如图2所示，通过根据本发明的多轴运动控制装置30来实现对端部操作器10的运动控制。在图2所示的应用场景中，任务指令为通过端部操作器10将工件41、42从托盘40中移动到托盘50中，放置在不与托盘50中已存在的工件51和52相干涉的位置，并且在有闯入者进入多轴运动机器或其端部操作器10周围的安全范围内或与端部操作器10发生接触时使端部操作器10停住并且避免撞伤闯入者。结合图3进行进一步说明，在主任务S11中，在步骤111接收所述指令，然后在步骤121解读该指令并分解成三个子任务：环境监测子任务S21、运动控制子任务S31和运动控制模式切 换子任务S41。在步骤131中，根据在环境监测子任务中得到的静态环境信息，即托盘40、工件41和42、托盘50、工件51和52的位置等，规划端部操作器10的运动轨迹。
在环境监测子任务S21中，在步骤211通过一个或多个传感器20获取多轴运动机器周围的环境数据，例如托盘、工件以及闯入者的三维图像。在步骤221中，识别待移动工件的类型和目标托盘中既有工件的类型以确定工件的尺寸、形状等。在步骤231中，计算工件的取送位置并将结果反馈至步骤131以规划端部操作器的运动轨迹。在步骤241中，监测端部操作器10的安全范围内是否存在闯入者60，并将信息提供给运动控制子任务S31。
在运动控制子任务S31中，在步骤311确定端部操作器的初始运动控制模式，初始运动控制模式一般为位置控制模式，以使端部操作器响应于指令快速运动到预定位置。在步骤321中，基于步骤241的监测结果判断是否存在闯入者，如果不存在，则继续以位置控制模式控制端部操作器10；如果存在闯入者，则在步骤331使端部操作器10停止运动并进入到运动控制模式切换子任务S41。在步骤411中，选择重力补偿控制模式。在步骤421中，根据以下公式(1)调整对应的控制参数，并在步骤431中将具有合适控制参数的重力补偿控制模式提供至子任务S31中的步骤341，至此，端部操作器的运动控制模式已切换成重力补偿控制模式，在重力补偿控制模式下，如闯入者与端部操作器发生接触，端部操作器不会向闯入者施加(反馈)任何力或仅施加(反馈)极小的力。在运动控制子任务S31的步骤351中，进一步判断上述切换条件是否仍然得到满足，如果仍然满足，则仍然采取重力补偿控制模式,；如果不再满足，则来到步骤361，恢复初始运动控制模式，即，位置控制模式，上述步骤将会重复进行。
Figure PCTCN2019109244-appb-000017
在对应于重力补偿控制模式的公式(1)中，τ为动力系统的输出扭矩；g(q)为重力扭矩向量函数；
Figure PCTCN2019109244-appb-000018
为关节摩擦力函数；q为关节位置；
Figure PCTCN2019109244-appb-000019
为关节速度。
在经过以上控制模式切换后，运动控制装置30的输出将会从电机速度设定值加扭矩前馈变成电机扭矩，从而在与闯入者发生碰撞时不会进一步施加额外的力，避免伤害闯入者。
图4和图5示意性地示出了第二种应用场景和对应的控制方法流程图。如图4所示，通过根据本发明的多轴运动控制装置30来实现对端部操作器10的运动控制。在图4所示的应用场景中，任务指令为通过端部操作器10将工件41、42从托盘40递送给操作员70，并由操作员70将工件41、42放置到托盘50中。与如上所述的第一种场景的不同之处在于，端部操作器10需要主动将工件传递给操作员70并且不能在传递过程中伤害操作员，同时操作器10还要保持高效运行(不停止运动)。结合图5进行进一步说明，在主任务S12中，在步骤112接收所述指令，然后在步骤122解读该指令并分解成三个子任务：环境监测子任务S22、 运动控制子任务S32和运动控制模式切换子任务S42。在步骤132中，根据在环境监测子任务中得到的静态环境信息，即托盘40、工件41和42、托盘50、工件51和52、操作员70的位置等，规划端部操作器10的运动轨迹。
在环境监测子任务S22中，在步骤212通过一个或多个传感器20获取多轴运动机器周围的环境数据，例如托盘、工件以及操作员的三维图像。在步骤222中，识别障碍物、待移动工件的类型以确定工件的尺寸、形状等。在步骤232中，识别操作员70，可以利用图像处理方法或机器学习方法等自始至终识别操作员，并将信息提供给运动控制子任务S32。在步骤242中，计算工件的取送位置和与操作员70的交互位置并将结果反馈至步骤132以规划端部操作器的运动轨迹。
在运动控制子任务S32中，在步骤312确定端部操作器的初始运动控制模式，初始运动控制模式一般为位置控制模式，以使端部操作器响应于指令快速运动到预定位置。在步骤322中，基于步骤232的识别结果判断是否已接近操作员70，如果仍未接近，则继续以位置控制模式控制端部操作器10；如果已接近，则进入到运动控制模式切换子任务S42。在步骤412中，选择轴系空间阻抗控制模式，即使得端部操作器10在围绕轴(关节)进行运动时像弹簧一样具有一定弹性，可以在与操作员70发生接触时向操作员反馈一定大小的力但不会使操作员70受伤。在步骤422中，根据以下公式(2)调整对应的控制参数，并在步骤432中将具有合适控制参数的轴系空间阻抗控制模式提供至子任务S32中的步骤332，至此，操作器的运动控制模式已切换成轴系空间阻抗控制模式。在运动控制子任务S32的步骤342中，进一步判断端部操作器10是否已远离操作员70，如果仍未远离，则仍然采取轴系空间阻抗控制模式；如果已经远离，则来到步骤352，恢复初始运动控制模式，即，位置控制模式，上述步骤将会重复进行。
Figure PCTCN2019109244-appb-000020
在对应于轴系空间阻抗控制模式的公式(2)中，τ为动力系统的输出扭矩；M(q)为惯性矩阵；
Figure PCTCN2019109244-appb-000021
为科里奥利/离心矩阵；g(q)为重力扭矩向量函数；
Figure PCTCN2019109244-appb-000022
为关节摩擦力函数；k p和k d分别为关节刚度矩阵和关节阻尼矩阵，可以人工配置；q为实际关节位置；
Figure PCTCN2019109244-appb-000023
为实际关节速度；q *为理想关节位置；
Figure PCTCN2019109244-appb-000024
为理想关节速度；
Figure PCTCN2019109244-appb-000025
为理想关节加速度。
在经过以上控制模式切换后，运动控制装置30的输出将会从电机速度设定值加扭矩前馈变成电机扭矩，从而在与操作员接触时会在围绕轴(关节)运动的方向发生屈服，避免伤害操作员。
图6和图7示意性地示出了第三种应用场景和对应的控制方法流程图。如图6所示，通过根据本发明的多轴运动控制装置30来实现对端部操作器10的运动控制。在图6所示的应 用场景中，任务指令为利用端部操作器(钻头或螺丝刀)10以预定力沿垂直于待加工部件80的斜面81的方向在斜面81上钻孔或旋拧螺丝。如图6所示，x0和z0所表示的坐标系为具有端部操作器10的协作机器人的基座的坐标系，x和z所表示的坐标系为端部操作器的坐标系。
结合图7进行进一步说明，在主任务S13中，在步骤113接收所述指令，然后在步骤123解读该指令并分解成三个子任务：环境监测子任务S23、运动控制子任务S33和运动控制模式切换子任务S43。在步骤133中，根据在环境监测子任务中得到的静态环境信息，即待加工部件80的位置，待加工部位81的位置和朝向等，规划钻头或螺丝刀10的运动轨迹。
在环境监测子任务S23中，在步骤213通过一个或多个传感器20获取多轴运动机器周围的环境数据。在步骤223中，可以利用图像处理和机器学习方法等识别待加工表面以及该表面的倾斜角度。在步骤233中，计算进入点，即，待加工表面垂直上方一定距离处的位置；并计算目标点(即，理想的孔深度)或确定预定大小的反馈扭矩，并将计算结果反馈至步骤133以规划端部操作器的运动轨迹。在步骤243中，将端部操作器(钻头或螺丝刀)坐标系换算成基座坐标系并将结果提供给运动控制子任务S33。
在运动控制子任务S33中，在步骤313确定端部操作器10的初始运动控制模式，初始运动控制模式一般为位置控制模式，以使端部操作器响应于指令快速运动到预定位置，即，进入点。在步骤323中，基于步骤243的计算结果判断端部操作器是否已到达进入点(即，开始钻孔或旋拧螺丝的位置)，如果未到达，则继续以位置控制模式控制端部操作器10；如果已到达，则进入到运动控制模式切换子任务S43。在步骤413中，选择笛卡尔空间阻抗控制模式，使得端部操作器10在其前进方向像弹簧一样具有一定弹性。在步骤423中，根据以下公式(3)调整对应的控制参数，并在步骤433中将具有合适控制参数的笛卡尔空间阻抗控制模式提供至子任务S33中的步骤333，至此，端部操作器10的运动控制模式已切换成笛卡尔空间阻抗控制模式。在运动控制子任务S33的步骤343中，进一步判断端部操作器10是否已到达目标点，例如，钻头是否已达到预定钻孔深度或被旋拧的螺丝是否向端部操作器反馈了预定大小的扭矩，如果未到达，则仍然采取笛卡尔空间阻抗控制模式继续操作；如果已到达，则来到步骤353，恢复初始运动控制模式，即，位置控制模式撤回端部操作器10。上述步骤可以重复进行以钻取多个孔或旋拧多个螺丝。
Figure PCTCN2019109244-appb-000026
在对应于笛卡尔空间阻抗控制模式的公式(3)中，τ为动力系统的输出扭矩；g(q)为重力扭矩向量函数；
Figure PCTCN2019109244-appb-000027
为关节摩擦力函数；J(q)为雅可比矩阵；Λ(x)为惯性矩阵；
Figure PCTCN2019109244-appb-000028
为笛卡尔空间中的非线性项；D d为笛卡尔空间阻尼；K d为笛卡尔空间刚度；F c为端部操作器施加的力；K f为用于确定F c的方向的力选择矩阵；q为关节位置；
Figure PCTCN2019109244-appb-000029
为关节速度；x为端 部操作器在笛卡尔空间中的实际位置；x d为端部操作器在笛卡尔空间中的规划位置；
Figure PCTCN2019109244-appb-000030
为端部操作器在笛卡尔空间中的规划速度；
Figure PCTCN2019109244-appb-000031
为端部操作器在笛卡尔空间中的规划加速度；
Figure PCTCN2019109244-appb-000032
为端部操作器在笛卡尔空间中的位置误差。
另一方面，本发明还提供一种基于上述控制方法的用于控制多轴运动机器的多轴运动控制装置或控制器，其示意性架构如图8中虚线框内所示。所述装置包括环境监测模块、任务规划模块、运动控制模块和运动控制模式切换模块。其中，环境监测模块2000用于收集环境数据并识别目标(例如，工件、操作员、闯入者和障碍物等)，其所收集的环境数据是任务规划模块用于规划任务的基本信息之一。环境监测模块2000向传感器20发送传感器控制指令2010，传感器20根据指令2010将感测数据21反馈至环境监测模块2000。任务规划模块用于将接收到的任务级指令分解为多个子任务并根据环境监测模块2000收集的静态环境信息进行自主任务规划(例如规划端部操作器的运动轨迹等)。任务规划模块包括用于解读指令和拆分子任务的自主任务规划模块1100和用于规划端部操作器的运动轨迹的自主路径规划模块1200。运动控制模块用于根据在所述环境监测模块中得到的动态环境信息实时确定用于控制所述端部操作器的运动控制模式并用于执行用于控制所述端部操作器的运动控制模式，其包括正向&反向运动学模块3100、运动学插补模块3200、动力学模型模块3300和位置反馈模块3400，其中，动力学模型模块3300包括用于执行阻抗控制的阻抗控制子模块3310、用于执行重力补偿控制的重力补偿子模块3320和用于执行位置控制的扭矩前馈子模块或位置控制子模块3330。运动控制模式切换模块用于将所述端部操作器的当前的运动控制模式切换成所确定的运动控制模式并调整对应的控制参数。运动控制模式切换模块包括用于选择运动控制模式的运动控制模式选择模块4100和用于调整控制参数的控制参数调整模块4200。运动控制模式切换模块还包括用于切换运动控制模式的运动控制模式切换开关4300和用于与伺服驱动装置90通信的通信模块4400。伺服驱动装置90与电机100连接用于控制电机100。伺服驱动装置90还用于向通信模块4400反馈电机的实际位置、速度和扭矩数据91。
如图8所示，正向&反向运动学模块3100和运动学插补模块3200用于插补计算端部操作器的运动轨迹并定期(例如每2ms)生成关节理论位置、速度和加速度3210提供给动力学模型模块3300。阻抗控制子模块3310和重力补偿子模块3320根据关节理论位置、速度和加速度3210计算关节理论扭矩并分别输出用于控制电机100的电机扭矩3311和3321。位置控制子模块3330根据关节理论位置、速度和加速度3210计算关节理论扭矩并输出用于控制电机100的电机前馈扭矩3331以改善系统的动态性能。位置反馈模块3400用于监测端部操作器的理论位置和实际位置之间的位置误差、计算理论速度调整值3410并将其加入到电机速度设定值3420中。
又一方面，本发明还提供一种用于控制多轴运动机器的多轴运动控制装置，所述多轴运动控制装置包括存储器和控制器，其中，所述存储器用于存储计算机程序，所述控制器被配置成能够在执行所述计算机程序时实施如上所述的控制方法。
进一步地，本发明还提供一种计算机可读存储介质，所述计算机可读存储介质中存储有计算机程序，当所述计算机程序被控制器执行时，能够实施如上所述的控制方法。
应当理解，虽然本说明书是按照各个实施例描述的，但并非每个实施例仅包含一个独立的技术方案，说明书的这种叙述方式仅仅是为清楚起见，本领域技术人员应当将说明书作为一个整体，各实施例中的技术方案也可以经适当组合，形成本领域技术人员可以理解的其他实施方式。
以上所述仅为本发明示意性的具体实施方式，并非用以限定本发明的范围。任何本领域的技术人员，在不脱离本发明的构思和原则的前提下所作的等同变化、修改与结合，均应属于本发明保护的范围。

Claims (21)

  1. 一种用于控制多轴运动机器的多轴运动控制方法，所述多轴运动机器包括能够围绕多个轴线运动的端部操作器(10)，所述方法包括：
    接收任务级指令；
    将所述任务级指令分解为多个子任务，所述多个子任务包括环境监测子任务，在所述环境监测子任务中，感测并确定所述多轴运动机器周围的环境信息；
    规划运动轨迹，根据在所述环境监测子任务中得到的静态环境信息规划所述端部操作器(10)的运动轨迹；
    所述多个子任务还包括：
    运动控制子任务，根据在所述环境监测子任务中得到的动态环境信息实时确定用于控制所述端部操作器(10)的运动控制模式；
    运动控制模式切换子任务，将当前的运动控制模式切换成在所述运动控制子任务中确定的运动控制模式并调整对应的控制参数。
  2. 根据权利要求1所述的多轴运动控制方法，其中，用于控制所述端部操作器(10)的初始运动控制模式为位置控制模式。
  3. 根据权利要求2所述的多轴运动控制方法，其中，所述任务级指令包括以下项目中的一个或多个：移动目标对象、加工目标对象、与目标对象交互并避免撞伤目标对象、避免撞伤非目标对象。
  4. 根据权利要求3所述的多轴运动控制方法，其中，当所述任务级指令包括移动目标对象时，所述环境监测子任务包括识别目标对象、计算目标对象的当前位置和目标位置。
  5. 根据权利要求4所述的多轴运动控制方法，其中，所述运动控制子任务包括确定用于控制所述端部操作器(10)的运动控制模式为位置控制模式。
  6. 根据权利要求3所述的多轴运动控制方法，其中，当所述任务级指令包括加工目标对象时，所述环境监测子任务包括识别目标对象、确定目标对象的待加工部位、计算所述待加工部位的位置和朝向、以及计算所述端部操作器(10)的当前位置和朝向。
  7. 根据权利要求6所述的多轴运动控制方法，其中，所述运动控制子任务包括根据所述端部操作器(10)的当前位置将用于控制所述端部操作器(10) 的运动控制模式切换为笛卡尔空间阻抗控制模式，所述运动控制模式切换子任务包括根据以下公式调整对应于笛卡尔空间阻抗控制模式的控制参数：
    Figure PCTCN2019109244-appb-100001
    其中：τ为动力系统的输出扭矩；
    g(q)为重力扭矩向量函数；
    Figure PCTCN2019109244-appb-100002
    为关节摩擦力函数；
    J(q)为雅可比矩阵；
    Λ(x)为惯性矩阵；
    Figure PCTCN2019109244-appb-100003
    为笛卡尔空间中的非线性项；
    D d为笛卡尔空间阻尼；
    K d为笛卡尔空间刚度；
    F c为所述端部操作器施加的力；
    K f为用于确定F c的方向的力选择矩阵；
    q为关节位置；
    Figure PCTCN2019109244-appb-100004
    为关节速度；
    x为所述端部操作器在笛卡尔空间中的实际位置；
    x d为所述端部操作器在笛卡尔空间中的规划位置；
    Figure PCTCN2019109244-appb-100005
    为所述端部操作器在笛卡尔空间中的规划速度；
    Figure PCTCN2019109244-appb-100006
    为所述端部操作器在笛卡尔空间中的规划加速度；
    Figure PCTCN2019109244-appb-100007
    为所述端部操作器在笛卡尔空间中的位置误差。
  8. 根据权利要求3所述的多轴运动控制方法，其中，当所述任务级指令包括与目标对象交互并避免撞伤目标对象时，所述环境监测子任务包括识别目标对象、计算目标对象的位置、计算所述端部操作器(10)的当前位置、计算所述端部操作器(10)与目标对象的交互位置。
  9. 根据权利要求8所述的多轴运动控制方法，其中，所述运动控制子任务包括根据所述端部操作器(10)的当前位置将用于控制所述端部操作器(10)的运动控制模式切换为轴系空间阻抗控制模式，所述运动控制模式切换子任务包括根据以下公式调整对应于轴系空间阻抗控制模式的控制参数：
    Figure PCTCN2019109244-appb-100008
    其中：τ为动力系统的输出扭矩；
    M(q)为惯性矩阵；
    Figure PCTCN2019109244-appb-100009
    为科里奥利/离心矩阵；
    g(q)为重力扭矩向量函数；
    Figure PCTCN2019109244-appb-100010
    为关节摩擦力函数；
    k p为关节刚度矩阵；
    k d为关节阻尼矩阵；
    q为实际关节位置；
    Figure PCTCN2019109244-appb-100011
    为实际关节速度；
    q *为理想关节位置；
    Figure PCTCN2019109244-appb-100012
    为理想关节速度；
    Figure PCTCN2019109244-appb-100013
    为理想关节加速度。
  10. 根据权利要求3所述的多轴运动控制方法，其中，当所述任务级指令包括避免撞伤非目标对象时，所述环境监测子任务包括实时监测在所述多轴运动机器周围的预定范围内是否存在非目标对象。
  11. 根据权利要求10所述的多轴运动控制方法，其中，所述运动控制子任务包括将用于控制所述端部操作器(10)的运动控制模式切换为重力补偿模式，所述运动控制模式切换子任务包括根据以下公式调整对应于重力补偿模式的控制参数：
    Figure PCTCN2019109244-appb-100014
    其中：τ为动力系统的输出扭矩；
    g(q)为重力扭矩向量函数；
    Figure PCTCN2019109244-appb-100015
    为关节摩擦力函数；
    q为关节位置；
    Figure PCTCN2019109244-appb-100016
    为关节速度。
  12. 一种用于控制多轴运动机器的多轴运动控制装置，所述多轴运动机器包括能够围绕多个轴线运动的端部操作器(10)，所述多轴运动控制装置(30)包括如下模块：
    环境监测模块，用于感测并确定所述多轴运动机器周围的环境信息；
    任务规划模块，用于将接收到的任务级指令分解为多个子任务并根据所述环境监测模块得到的静态环境信息规划所述端部操作器(10)的运动轨迹；
    运动控制模块，用于根据在所述环境监测模块中得到的动态环境信息实时确定用于控制所述端部操作器(10)的运动控制模式，以及用于执行用于 控制所述端部操作器(10)的运动控制模式；
    运动控制模式切换模块，用于将所述端部操作器(10)的当前的运动控制模式切换成所确定的运动控制模式并调整对应的控制参数。
  13. 根据权利要求12所述的多轴运动控制装置，其中，所述运动控制模块包括用于执行阻抗控制模式的阻抗控制子模块、用于执行重力补偿模式的重力补偿子模块、用于执行位置控制模式的位置控制子模块。
  14. 根据权利要求13所述的多轴运动控制装置，其中，所述运动控制模式切换模块通过切换所述子模块来将所述端部操作器(10)的当前的运动控制模式切换成在所述任务规划模块中确定的运动控制模式。
  15. 根据权利要求12所述的多轴运动控制装置，其中，所述任务级指令包括以下项目中的一个或多个：移动目标对象、加工目标对象、与目标对象交互并避免撞伤目标对象、避免撞伤非目标对象。
  16. 根据权利要求15所述的多轴运动控制装置，其中，当所述任务级指令包括移动目标对象时，所述环境监测模块包括识别目标对象、计算目标对象的当前位置和目标位置。
  17. 根据权利要求15所述的多轴运动控制装置，其中，当所述任务级指令包括加工目标对象时，所述环境监测模块用于识别目标对象、确定目标对象的待加工部位、计算所述待加工部位的位置和朝向、以及计算所述端部操作器(10)的当前位置和朝向。
  18. 根据权利要求15所述的多轴运动控制装置，其中，当所述任务级指令包括与目标对象交互并避免撞伤目标对象时，所述环境监测模块用于识别目标对象、计算目标对象的位置、计算所述端部操作器(10)的当前位置、计算所述端部操作器(10)与目标对象的交互位置。
  19. 根据权利要求15所述的多轴运动控制装置，其中，当所述任务级指令包括避免撞伤非目标对象时，所述环境监测模块用于实时监测在所述多轴运动机器周围的预定范围内是否存在非目标对象。
  20. 一种用于控制多轴运动机器的多轴运动控制装置，所述多轴运动控制装置(30)包括存储器和控制器，其中，所述存储器用于存储计算机程序，所述控制器被配置成能够在执行所述计算机程序时实施如权利要求1-11中任一项所述的方法。
  21. 一种计算机可读存储介质，所述计算机可读存储介质中存储有计算机程序， 当所述计算机程序被控制器执行时，能够实施如权利要求1-11中任一项所述的方法。
PCT/CN2019/109244 2019-09-29 2019-09-29 用于控制多轴运动机器的方法、装置和存储介质 WO2021056591A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
CN201980100024.5A CN114341823A (zh) 2019-09-29 2019-09-29 用于控制多轴运动机器的方法、装置和存储介质
PCT/CN2019/109244 WO2021056591A1 (zh) 2019-09-29 2019-09-29 用于控制多轴运动机器的方法、装置和存储介质

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2019/109244 WO2021056591A1 (zh) 2019-09-29 2019-09-29 用于控制多轴运动机器的方法、装置和存储介质

Publications (1)

Publication Number Publication Date
WO2021056591A1 true WO2021056591A1 (zh) 2021-04-01

Family

ID=75165437

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/109244 WO2021056591A1 (zh) 2019-09-29 2019-09-29 用于控制多轴运动机器的方法、装置和存储介质

Country Status (2)

Country Link
CN (1) CN114341823A (zh)
WO (1) WO2021056591A1 (zh)

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101870107A (zh) * 2010-06-26 2010-10-27 上海交通大学 骨科手术辅助机器人的控制系统
CN102292194A (zh) * 2009-08-21 2011-12-21 松下电器产业株式会社 机器人手臂的控制装置及控制方法、装配机器人、机器人手臂的控制程序及机器人手臂的控制用集成电路
CN102300679A (zh) * 2009-09-28 2011-12-28 松下电器产业株式会社 机器人手臂的控制装置及控制方法、机器人、机器人手臂的控制程序及机器人手臂控制用集成电路
CN104731642A (zh) * 2013-12-23 2015-06-24 国际商业机器公司 用于高度并行作业的回填调度方法和系统
CN105599821A (zh) * 2016-01-06 2016-05-25 山东优宝特智能机器人有限公司 具有环境感知能力的电驱动仿生四足机器人及控制方法
CN108536011A (zh) * 2018-03-19 2018-09-14 中山大学 一种基于深度强化学习的六足机器人复杂地形自适应运动控制方法

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102292194A (zh) * 2009-08-21 2011-12-21 松下电器产业株式会社 机器人手臂的控制装置及控制方法、装配机器人、机器人手臂的控制程序及机器人手臂的控制用集成电路
CN102300679A (zh) * 2009-09-28 2011-12-28 松下电器产业株式会社 机器人手臂的控制装置及控制方法、机器人、机器人手臂的控制程序及机器人手臂控制用集成电路
CN101870107A (zh) * 2010-06-26 2010-10-27 上海交通大学 骨科手术辅助机器人的控制系统
CN104731642A (zh) * 2013-12-23 2015-06-24 国际商业机器公司 用于高度并行作业的回填调度方法和系统
CN105599821A (zh) * 2016-01-06 2016-05-25 山东优宝特智能机器人有限公司 具有环境感知能力的电驱动仿生四足机器人及控制方法
CN108536011A (zh) * 2018-03-19 2018-09-14 中山大学 一种基于深度强化学习的六足机器人复杂地形自适应运动控制方法

Also Published As

Publication number Publication date
CN114341823A (zh) 2022-04-12

Similar Documents

Publication Publication Date Title
US11446820B2 (en) Robot path generating device and robot system
EP3222393B1 (en) Automated guidance system and method for a coordinated movement machine
TWI630081B (zh) Direct teaching method of robot
US20120059515A1 (en) Workspace safe operation of a force- or impedance-controlled robot
JP2009039852A (ja) 衝突回避機能を有するロボット
US11433537B2 (en) Automatic path generation device
JPH08227320A (ja) 直交座標系上で柔らかさが調節可能なサーボ系
Lee et al. Posture optimization in robotic drilling using a deformation energy model
US20210402596A1 (en) Control Method for Robot and Robot System
WO2021056591A1 (zh) 用于控制多轴运动机器的方法、装置和存储介质
Carelli et al. Combined force and visual control of an industrial robot
Ostanin et al. Programming by Demonstration Using Two-Step Optimization for Industrial Robot.
Long et al. Robotic cutting of soft materials using force control & image moments
Miyabe et al. An approach toward an automated object retrieval operation with a two-arm flexible manipulator
US20220297293A1 (en) Dynamic torque saturation limits for robot actuator(s)
Prakash et al. Adaptive backstepping neural tracking control of an uncertain robot manipulator with dynamic disturbances
Wu et al. Collaboration of multiple mobile manipulators with compliance based leader/follower approach
Yovchev Finding the optimal parameters for robotic manipulator applications of the bounded error algorithm for iterative learning control
JP7060700B2 (ja) コーディネートシステム、ハンドリング装置及び方法
Wang et al. Design of kinematic controller for real-time vision guided robot manipulators
Oh et al. Remote control for redundant humanoid arm using optimized arm angle
Ng et al. A dual joystick-trackball interface for accurate and time-efficient teleoperation of cable-driven parallel robots within large workspaces
Benbelkacem et al. Position-based visual servoing through Cartesian path-planning for a grasping task
Timmermann et al. AI4Assembly A Human-Robot Collaboration Assembly Application with AI Support
WO2024045091A1 (zh) 执行机构的运动规划方法、装置、系统及存储介质

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19947152

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 19947152

Country of ref document: EP

Kind code of ref document: A1