Derivative-free online learning of inverse dynamics models

  • 2018-09-13 17:29:35
  • Diego Romeres, Mattia Zorzi, Raffaello Camoriano, Silvio Traversaro, Alessandro Chiuso
  • 1

Abstract

This paper discusses online algorithms for inverse dynamics modelling inrobotics. Several model classes including rigid body dynamics (RBD) models,data-driven models and semiparametric models (which are a combination of theprevious two classes) are placed in a common framework. While model classesused in the literature typically exploit joint velocities and accelerations,which need to be approximated resorting to numerical differentiation schemes,in this paper a new `derivative-free' framework is proposed that does notrequire this preprocessing step. An extensive experimental study with real datafrom the right arm of the iCub robot is presented, comparing different modelclasses and estimation procedures, showing that the proposed `derivative-free'methods outperform existing methodologies.

 

Introduction (beta)

None

 

Conclusion (beta)

None