A branch-and-prune method to solve closure equations in dual quaternions

Altres autors/es

Institut de Robòtica i Informàtica Industrial

Universitat Politècnica de Catalunya. KRD - Cinemàtica i Disseny de Robots

Data de publicació

2021

Resum

© 2021 Elsevier


Using dual quaternions, the closure equations of a kinematic loop can be expressed as a system of multiaffine quations. In this paper, this property is leveraged to introduce a branch-and-prune method specially tailored for solving such systems of equations. The new method is objectively simpler (in the sense that it is easier to understand and to implement) than previous approaches relying on general techniques such as interval Newton methods or methods based on Bernstein polynomials or linear relaxations. Moreover, it relies on two basic operations —linear interpolation and projection onto coordinate planes— that can be e¿ciently computed. The generality of the proposed method is evaluated on position analysis problems with 0-, 1-, and 2-dimensional solution sets, including the inverse kinematics of serial robots and the forward kinematics of parallel ones. The results obtained on these problems show that the efficiency of the method compares favorably to state-of-the-art alternatives.


This work was partially supported by the Spanish Ministry of Economy and Competitiveness through the project DPI2017- 88282-P.


Peer Reviewed


Postprint (author's final draft)

Tipus de document

Article

Llengua

Anglès

Documents relacionats

https://www.sciencedirect.com/science/article/abs/pii/S0094114X21001828?via%3Dihub

info:eu-repo/grantAgreement/AEI/Plan Estatal de Investigación Científica y Técnica y de Innovación 2013-2016/DPI2017-88282-P/ES/PLANIFICACION CINETODINAMICA DE MOVIMIENTOS ROBOTICOS EFICIENTES Y AGILES/

Citació recomanada

Aquesta citació s'ha generat automàticament.

Drets

http://creativecommons.org/licenses/by-nc-nd/3.0/es/

Open Access

Attribution-NonCommercial-NoDerivs 3.0 Spain

Aquest element apareix en la col·lecció o col·leccions següent(s)

E-prints [73034]