Path Encoding and Decoding for Articulated Robots via Information Extraction

This paper addresses efficient path planning for articulated robots. Traditional methods suffer from high computational costs and tuning parameter issues, worsened by frequent kinematic parameter changes. We propose a novel path generation approach independent of these parameters, i.e., a path reuse...

Full description

Saved in:
Bibliographic Details
Published in2024 SICE Festival with Annual Conference (SICE FES) pp. 169 - 173
Main Authors Nishida, Takeshi, Watanabe, Tsubasa
Format Conference Proceeding
LanguageEnglish
Published The Society of Instrument and Control Engineers - SICE 27.08.2024
Subjects
articulated robots
Collision avoidance
Computational efficiency
Decoding
Encoding
Kinematics
Path planning
robot configuration space
Robots
Shape
Tuning
Vectors
Online AccessGet full text

Cover

More Information
Summary:This paper addresses efficient path planning for articulated robots. Traditional methods suffer from high computational costs and tuning parameter issues, worsened by frequent kinematic parameter changes. We propose a novel path generation approach independent of these parameters, i.e., a path reuse method. Typically, robot paths are planned in task space (T-space) and robot configuration space (C-space), with collision checks dominating computational costs. Prior methods such as random sampling and gradient search are inefficient for dynamic environments. To address this, a novel encoding method for paths in C-space is introduced, capturing key path characteristics. This encoding facilitates path reconstruction, and the fast-decoding technique enables adaptation to different start/goal positions, significantly reducing collision checks and computational costs.