日本フィジカルAI新聞

世界のフィジカルAIを、日本語で。

歩行arXiv:2312.06365

12自由度四足ロボットのためのバランス型位置制御アーキテクチャ:シミュレーション検証とハードウェア試験

A Balanced Positional Control Architecture for a 12-DoF Quadruped Robot through Simulation-validation and Hardware Testing

シェア:XThreadsFacebookLINEはてブBluesky

12自由度四足ロボットの脚位置を座標系内でバランスよく制御する新アルゴリズムを提案し、順・逆運動学の導出、Pythonシミュレーション、マイコンとのシリアル通信による実機検証を行い、試作脚でシミュレーション結果と78.9%の精度で一致することを示した。

著者: Abid Shahriar

分類: cs.RO, cs.SY, eess.SY

原文アブストラクト

A multi-joint enabled robot requires extensive mathematical calculations to determine the end effector's position with respect to the other connective joints involved and their corresponding frames in a specific coordinate system. If a control architecture employs fewer positional constraints which cannot precisely determine the end effector's position in all quadrants of a 2D Cartesian plane then the robot is generally under-constrained, leading to challenges in accurate positioning to the end-effector across the entire plane. Consequently, only a subset of the end effector's degree of freedom (DoF) can be assigned for the robot's leg position for pose and trajectory estimation purposes. This paper introduces a novel approach and proposes an algorithm to consider a balanced control of the robot's leg position in a coordinate system so the robot's leg can be precisely determined and the DoF is not limited. Mathematical derivation of the joint angles is derived with forward and inverse kinematics, and Python-based simulation has been done to verify and simulate the robot's locomotion. Using Python-based code for serial communication with a micro-controller unit makes this approach more effective for demonstrating its application on a prototype leg its movement has been realized. The experimental prototype leg exhibits a commendable 78.9% accuracy with the simulated result, validating the robustness of our algorithm in practical scenarios. A comprehensive assessment of the control algorithm with random and continuous data point test has been conducted to ensure performance, so the algorithm can as well be deployed in a physical robot.

関連論文

  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. 3Google DeepMind、ロボット向けAIを発表新製品・発表
  4. 4NVIDIA、CoRLでGR00T 2とIsaac Lab 3.0を発表新製品・発表
  5. 5GXO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

PR・アフィリエイト広告

FRONTIER(GPU搭載BTOパソコン)

FRONTIER(GPU搭載BTOパソコン)の公式広告バナーFRONTIERでGPU搭載PCを見る ↗
PR本紙発行元 EmplifAI