Skip to main content

LUP Student Papers

LUND UNIVERSITY LIBRARIES

Dual-Motor Control for Backlash Reduction in Parallel-Kinematic Robot Joints

Artursson, Mathias (2019)
Department of Automatic Control
Abstract
A new high-performance robot called the Gantry-Tau robot was developed by ABB Robotics, the Robotics Lab at Lund University and Güdel AG. This robot seemed promising in terms of speed, accuracy, stiffness and bandwidth of the motion control. However, the robot joints were based on the rack-and-pinion principle, which introduced significant backlash into the system. To solve this problem, it was proposed to use two motors to control each joint, where the motors would go in opposite directions to ensure that the gears and motors were in contact at all times. How this should be implemented is still under development.
This master thesis attempted to implement backlash compensation together with conventional uses of the robot joints, such as... (More)
A new high-performance robot called the Gantry-Tau robot was developed by ABB Robotics, the Robotics Lab at Lund University and Güdel AG. This robot seemed promising in terms of speed, accuracy, stiffness and bandwidth of the motion control. However, the robot joints were based on the rack-and-pinion principle, which introduced significant backlash into the system. To solve this problem, it was proposed to use two motors to control each joint, where the motors would go in opposite directions to ensure that the gears and motors were in contact at all times. How this should be implemented is still under development.
This master thesis attempted to implement backlash compensation together with conventional uses of the robot joints, such as position-, velocity- and accelerationtrajectories as well as torque feed forward. The goal for the cart was to have regular control for both motors when following trajectories, except for stationary points
where the motors would work in different directions. A test rack was provided for this purpose by Lund University and Cognibotics.
First, trajectories were generated for the robot joint to follow. A control structure was then implemented for the robot joint, consisting of a cascade structure for position control, as well as torque feed forward from a friction and dynamic model. Different methods for backlash-compensation were also presented and implemented.
These methods for backlash-compensation together with the control structure were then evaluated in terms of position accuracy without external disturbances, and overshoot, settling time and backlash traversal with external disturbances.
It was concluded that some implemented backlash compensation methods resulted in a more dampened response when no external disturbances were present. Experiments with external disturbances showed that performance could be improved in some situation, but became worse in others. The reason for this decreased performance may have been because the controlling motor went through the backlash gap. (Less)
Please use this url to cite or link to this publication:
author
Artursson, Mathias
supervisor
organization
year
type
H3 - Professional qualifications (4 Years - )
subject
report number
TFRT-6086
ISSN
0280-5316
language
English
id
8995379
date added to LUP
2019-09-19 09:28:28
date last changed
2019-09-19 09:28:28
@misc{8995379,
  abstract     = {{A new high-performance robot called the Gantry-Tau robot was developed by ABB Robotics, the Robotics Lab at Lund University and Güdel AG. This robot seemed promising in terms of speed, accuracy, stiffness and bandwidth of the motion control. However, the robot joints were based on the rack-and-pinion principle, which introduced significant backlash into the system. To solve this problem, it was proposed to use two motors to control each joint, where the motors would go in opposite directions to ensure that the gears and motors were in contact at all times. How this should be implemented is still under development.
 This master thesis attempted to implement backlash compensation together with conventional uses of the robot joints, such as position-, velocity- and accelerationtrajectories as well as torque feed forward. The goal for the cart was to have regular control for both motors when following trajectories, except for stationary points
where the motors would work in different directions. A test rack was provided for this purpose by Lund University and Cognibotics.
 First, trajectories were generated for the robot joint to follow. A control structure was then implemented for the robot joint, consisting of a cascade structure for position control, as well as torque feed forward from a friction and dynamic model. Different methods for backlash-compensation were also presented and implemented.
These methods for backlash-compensation together with the control structure were then evaluated in terms of position accuracy without external disturbances, and overshoot, settling time and backlash traversal with external disturbances.
 It was concluded that some implemented backlash compensation methods resulted in a more dampened response when no external disturbances were present. Experiments with external disturbances showed that performance could be improved in some situation, but became worse in others. The reason for this decreased performance may have been because the controlling motor went through the backlash gap.}},
  author       = {{Artursson, Mathias}},
  issn         = {{0280-5316}},
  language     = {{eng}},
  note         = {{Student Paper}},
  title        = {{Dual-Motor Control for Backlash Reduction in Parallel-Kinematic Robot Joints}},
  year         = {{2019}},
}