Tree Based Trajectory Optimization Based on Local Linearity of Continuous Non-Linear Dynamics

Chyon Hae Kim, Shigeki Sugano

研究成果: Article査読

9 被引用数 (Scopus)

抄録

This technical note addresses a tradeoff in the tree based trajectory optimization algorithms for open-loop optimal control problem of rigid body system. In this technical note, linear prediction based uniform state sampling method (LPUSS), which relaxes the tradeoff between solutions' quality and computational efficiency, is proposed on the basis of the local linearity of motion equations. LPUSS has a lower calculation order than randomized kinodynamic planning (RKP) and rapid semi-optimal motion planning (RASMO). In the validation using multi degree-of-freedom (DOF) under actuated manipulator models, the solutions' quality and the computational efficiency of LPUSS were better than those of RKP and RASMO. LPUSS finished the optimization for the 6 DOF model within 40 minutes. This was the world's first success of the optimization of the swing up motion of a 6 DOF under actuated manipulator model.

本文言語English
論文番号7322197
ページ(範囲)2610-2617
ページ数8
ジャーナルIEEE Transactions on Automatic Control
61
9
DOI
出版ステータスPublished - 2016 9

ASJC Scopus subject areas

  • 制御およびシステム工学
  • コンピュータ サイエンスの応用
  • 電子工学および電気工学

フィンガープリント

「Tree Based Trajectory Optimization Based on Local Linearity of Continuous Non-Linear Dynamics」の研究トピックを掘り下げます。これらがまとまってユニークなフィンガープリントを構成します。

引用スタイル