Table of Contents
Fetching ...

On robotic manipulators with time-dependent inertial parameters: From physical consistency to boundedness of the mass matrix

Tom Kaufmann, Johann Reger

TL;DR

<3-5 sentence high-level summary> This work extends robotic manipulator dynamics to accommodate time-varying inertial parameters and causative mass-distribution changes, linking physical consistency of inertia to boundedness of the mass matrix. It develops a Lagrangian-based generalized robotics equation with a regressor structure that separates pose-dependent and parameter-dependent terms, and introduces the notions of uniform physical consistency and upper boundedness for time-varying inertial parameters. The authors prove conditions under which the mass matrix M(q,Θ(t)) is uniformly positive definite and bounded, leveraging a normal Jacobian and forward-kinematic constraints akin to Gorbel’s results. These results provide a theoretical foundation for robust adaptive control of manipulators subject to load changes and internal mass redistribution, by guaranteeing a positive definite mass matrix over time.</paper_summary>

Abstract

We generalize the robotics equation describing the dynamics of an open kinematic chain to include the effect of time-dependent change of inertial parameters as well as the effects of its cause, i.e. time dependency of the distributions of mass originating from parasitic movements of mass-carrying particles. The results generate insight that allows linking the novel concepts of uniform physical consistency and upper boundedness of inertial parameters -- ruling out approaching the edge to physical inconsistency or to diverge -- with the existence of finite, positive uniform bounds of the mass matrix.

On robotic manipulators with time-dependent inertial parameters: From physical consistency to boundedness of the mass matrix

TL;DR

<3-5 sentence high-level summary> This work extends robotic manipulator dynamics to accommodate time-varying inertial parameters and causative mass-distribution changes, linking physical consistency of inertia to boundedness of the mass matrix. It develops a Lagrangian-based generalized robotics equation with a regressor structure that separates pose-dependent and parameter-dependent terms, and introduces the notions of uniform physical consistency and upper boundedness for time-varying inertial parameters. The authors prove conditions under which the mass matrix M(q,Θ(t)) is uniformly positive definite and bounded, leveraging a normal Jacobian and forward-kinematic constraints akin to Gorbel’s results. These results provide a theoretical foundation for robust adaptive control of manipulators subject to load changes and internal mass redistribution, by guaranteeing a positive definite mass matrix over time.</paper_summary>

Abstract

We generalize the robotics equation describing the dynamics of an open kinematic chain to include the effect of time-dependent change of inertial parameters as well as the effects of its cause, i.e. time dependency of the distributions of mass originating from parasitic movements of mass-carrying particles. The results generate insight that allows linking the novel concepts of uniform physical consistency and upper boundedness of inertial parameters -- ruling out approaching the edge to physical inconsistency or to diverge -- with the existence of finite, positive uniform bounds of the mass matrix.

Paper Structure

This paper contains 17 sections, 10 theorems, 47 equations.

Key Result

Theorem 1

(traversaro_identification_2016 and Theorem $3$, Eq. (16)-(17) in wensing_linear_2018) A vector $\Phi\in\mathbb{R}^{10}$ with entries according to $(eq:Phi)$ is physically consistent if and only if $f(\Phi)\succ 0$, where $f:\mathbb{R}^{10}\to\mathrm{sym}(4)$ is a function with and $h \in \mathbb{R}^3$, $I \in \mathrm{sym}(3)$ are structured as in $(eq:h)$, $(eq:I)$, respectively. The inverse tra

Theorems & Definitions (21)

  • Definition 1
  • Theorem 1
  • Remark 1
  • Definition 2
  • Definition 3
  • Theorem 2
  • proof
  • Lemma 1
  • proof
  • Theorem 3
  • ...and 11 more