arXiv Robotics — research abstracts· Sangwoo Shin, Ashvin Anilkumar, Ryan Gao, Josiah P. Hanna·· 1 days agoEditorial score58
Informationally Decoupled Trajectory Design for Sim-to-Real System Identification
Informationally Decoupled Trajectory Design for Sim-to-Real System Identification
Summary
This research proposes IDTD, a framework for trajectory design in sim-to-real system identification. By leveraging the Schur complement score and Fisher information matrix, IDTD decouples parameter effects, improving parameter estimation accuracy by 39.6% compared to existing methods. The approach is validated across multiple simulation environments and a real K1 humanoid robot.
Source: arXiv Robotics — research abstracts · Read original article ↗
Loading article text…
Source：arXiv Robotics — research abstracts · arxiv.org