日本フィジカルAI新聞

世界のフィジカルAIを、日本語で。

連続体ロボット経路計画arXiv:1812.03615

多節連続体アームのための準最適平滑経路計画

Near-optimal Smooth Path Planning for Multisection Continuum Arms

シェア:XThreadsFacebookLINEはてブBluesky

3節・各3筋アクチュエータの連続体アームに対し、構成グラフに基づく経路計画法を提案し、障害物の有無両方で完全性を示すとともに逆運動学手法より高い成功率を達成した。

著者: Jiahao Deng, Brandon H. Meng, Iyad Kanj, Isuru S. Godage

分類: cs.RO

原文アブストラクト

We study the path planning problem for continuum-arm robots, in which we are given a starting and an end point, and we need to compute a path for the tip of the continuum arm between the two points. We consider both cases where obstacles are present and where they are not. We demonstrate how to leverage the continuum arm features to introduce a new model that enables a path planning approach based on the configurations graph, for a continuum arm consisting of three sections, each consisting of three muscle actuators. The algorithm we apply to the configurations graph allows us to exploit parallelism in the computation to obtain efficient implementation. We conducted extensive tests, and the obtained results show the completeness of the proposed algorithm under the considered discretizations, in both cases where obstacles are present and where they are not. We compared our approach to the standard inverse kinematics approach. While the inverse kinematics approach is much faster when successful, our algorithm always succeeds in finding a path or reporting that no path exists, compared to a roughly 70% success rate of the inverse kinematics approach (when a path exists).

  1. 1Boston Dynamics、車輪付きAtlasの研究試作機を紹介研究・技術
  2. 2newmoとKDDIが資本業務提携 2028年に堺市で自動運転タクシー商用化目指す提携・買収
  3. 3GXOとApptronik、Apolloの物流現場試験を拡大 最大1,000台の展開を構想量産・展開
  4. 4NVIDIA、CoRLでGR00T 2とIsaac Lab 3.0を発表新製品・発表
  5. 5UC BerkeleyとNVIDIAの研究者、AIエージェントでロボット制御コードを仮想環境で改善研究・技術

PR・アフィリエイト広告

MINISFORUM(ミニPC)

MINISFORUM(ミニPC)の公式広告バナーMINISFORUM公式ストアを見る ↗
PR本紙発行元 EmplifAI