A Method to Identify the Nonlinear Stiffness Characteristics of an Elastic Continuum Mechanism

Bastian Deutschmann, Tong Liu, Alexander Dietrich, Christian Ott, Dongheui Lee

Research output: Contribution to journalArticlepeer-review

12 Scopus citations

Abstract

The humanoid robot David is equipped with a novel robotic neck based on an elastic continuum mechanism (ECM). To realize a model-based motion control, the six dimensional stiffness characteristics needs to be known. This letter presents an approach to experimentally identify the stiffness characteristic using a robot manipulator to deflect the ECM and measure the Cartesian wrenches and Cartesian poses with external sensors. A three-step process is proposed to establish Cartesian wrench and pose pairs experimentally. The process consists of a simulation step, to select a good model, a second step that extracts effective poses from workspace which are sampled experimentally and the third step, the pose sampling procedure in which the robot drives the ECM to these effective poses. A full cubic polynomial regression model is adopted based on simulation data to fit the stiffness characteristics. To extract the poses to be sampled in the experiments, two different approaches are evaluated and compared to ensure a well-posed identification. The identification process on the hardware is performed by using Cartesian impedance and inverse kinematics control in combination to comply with the physical constraints imposed by the ECM.

Original languageEnglish
Pages (from-to)1450-1457
Number of pages8
JournalIEEE Robotics and Automation Letters
Volume3
Issue number3
DOIs
StatePublished - Jul 2018
Externally publishedYes

Keywords

  • Model learning for control
  • compliant joint/mechanism
  • soft material robotics

Fingerprint

Dive into the research topics of 'A Method to Identify the Nonlinear Stiffness Characteristics of an Elastic Continuum Mechanism'. Together they form a unique fingerprint.

Cite this