Modeling and estimation of friction, extension, and coupling effects in multisegment continuum robots

Rajarshi Roy, Long Wang, Nabil Simaan

Research output: Contribution to journalArticlepeer-review

81 Scopus citations

Abstract

The prevalence of wire-actuated mechanisms and continuum robots (CRs) for surgical applications stems from several advantages due to the remote location of actuators from the end-effectors (e.g., improved down scalability and sterilization). These advantages, however, come at the expense of inherent uncertainties due to backlash effects, compliance, and friction in the actuation lines, which in turn, limit their precision. In addition, multisegment CRs suffer from actuation coupling, diminishing the accuracy of their kinematic model. This paper aims to address these two gaps by presenting a model-based estimation and actuation compensation framework enabling the online estimation of modeling uncertainties and friction, despite possible temporal changes in these parameters and cross coupling between segments. A capstan friction model accounting for friction-induced actuation line extensions in both fixed and variable geometry conduits is presented. A modified statics model for multisegment robots is presented to account for cross-coupling effects between subsequent CR segments. A sequential estimation approach using the robot inverse kinematics, statics, and measurements of actuated joint positions and forces is then presented. These approaches are then validated experimentally on a two-segment single port access surgery CR.

Original languageEnglish
Article number7795260
Pages (from-to)909-920
Number of pages12
JournalIEEE/ASME Transactions on Mechatronics
Volume22
Issue number2
DOIs
StatePublished - Apr 2017

Keywords

  • Actuation compensation
  • continuum robot
  • extension modeling
  • friction estimation
  • surgical robot

Fingerprint

Dive into the research topics of 'Modeling and estimation of friction, extension, and coupling effects in multisegment continuum robots'. Together they form a unique fingerprint.

Cite this