This paper presents a novel controller for soft arms to be employed in dynamic tracking tasks. Creating dynamic controllers for continuum soft robots has been among the most important objectives for this field. This is because relying on the steady-state assumption of static controllers greatly limits the capabilities of these robotic platforms, whose advantageous compliance and flexibility is paid with dynamics that are highly non-linear and hard to model. For this reason, a data-driven solution based on long-short term memory networks is introduced. The methodology is then tested both on simulated and real continuum robots. The results show that the controller allows to accurately follow trajectories in the task-space with an average error lower than 4mm.

Open-loop Model-free Dynamic Control of a Soft Manipulator for Tracking Tasks

Egidio Falotico
2021

Abstract

This paper presents a novel controller for soft arms to be employed in dynamic tracking tasks. Creating dynamic controllers for continuum soft robots has been among the most important objectives for this field. This is because relying on the steady-state assumption of static controllers greatly limits the capabilities of these robotic platforms, whose advantageous compliance and flexibility is paid with dynamics that are highly non-linear and hard to model. For this reason, a data-driven solution based on long-short term memory networks is introduced. The methodology is then tested both on simulated and real continuum robots. The results show that the controller allows to accurately follow trajectories in the task-space with an average error lower than 4mm.
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: http://hdl.handle.net/11382/546213
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 1
social impact