Reproduce “LEAP: Gradient Flow Augmented Lagrangian Trajectory Optimization for Contact Constrained Robots” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
César E. Ramos Chuquiure, Ryan Tsai, Yutong Zhou, Zihang Huang, Vansh Thakur, Akshunn Trivedi, Zachary Lin, Pearl Lin and 3 more
LEAP reshapes contact-constrained plans through a gradient flow, using an augmented Lagrangian and an integrator that accepts descending steps. It certifies friction-cone feasibility across intervals when Bézier wrench control points are feasible, while checking other constraints through mesh refinement. In ten simulated cases, references met tracking thresholds in 10 rollouts, versus seven for Crocoddyl and eight for Aligator.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.