Safe rendering of high impedance on a series-elastic actuator with disturbance observer-based torque control
Kevin Haninger, Abner Asignacion, Sehoon Oh

TL;DR
This paper investigates how disturbance observer-based torque control can enhance the safe impedance range of series-elastic actuators, validated through experiments in high-stiffness environments, showing improved performance over traditional high-gain PD control.
Contribution
It introduces passivity-based stability conditions for DOB torque control variants and proposes a dynamic feedforward compensator to increase safe impedance range.
Findings
DOB reduces the need for high-gain PD feedback
DOB allows higher rendered impedance than traditional methods
The proposed compensator increases maximum safe impedance
Abstract
An important performance metric for series-elastic actuators is the range of impedance which they can safely render. Advanced torque control, using techniques such as the disturbance observer, improve torque tracking bandwidth and accuracy, but their impact on safe impedance range is not established. However, to define a safe impedance range requires a practical coupled stability condition. Here, passivity-based conditions are proposed for two variants of DOB torque control, and validated experimentally in a high-stiffness environment. While high-gain PD torque control has been shown to reduce Z-width, it is here shown that a DOB reduces the need for high-gain PD feedback and allows a higher rendered impedance. A dynamic feedforward compensator is proposed which increases the maximum safe impedance of the DOB, validated in experimentally in collision with high-stiffness environments and…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsProsthetics and Rehabilitation Robotics · Teleoperation and Haptic Systems · Robot Manipulation and Learning
