Reduced-Order Model-Guided Reinforcement Learning for Demonstration-Free Humanoid Locomotion

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Liu, Shuai, Lau, Meng Cheng
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908555678842880
author Liu, Shuai
Lau, Meng Cheng
author_facet Liu, Shuai
Lau, Meng Cheng
contents We introduce Reduced-Order Model-Guided Reinforcement Learning (ROM-GRL), a two-stage reinforcement learning framework for humanoid walking that requires no motion capture data or elaborate reward shaping. In the first stage, a compact 4-DOF (four-degree-of-freedom) reduced-order model (ROM) is trained via Proximal Policy Optimization. This generates energy-efficient gait templates. In the second stage, those dynamically consistent trajectories guide a full-body policy trained with Soft Actor--Critic augmented by an adversarial discriminator, ensuring the student's five-dimensional gait feature distribution matches the ROM's demonstrations. Experiments at 1 meter-per-second and 4 meter-per-second show that ROM-GRL produces stable, symmetric gaits with substantially lower tracking error than a pure-reward baseline. By distilling lightweight ROM guidance into high-dimensional policies, ROM-GRL bridges the gap between reward-only and imitation-based locomotion methods, enabling versatile, naturalistic humanoid behaviors without any human demonstrations.
format Preprint
id arxiv_https___arxiv_org_abs_2509_19023
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Reduced-Order Model-Guided Reinforcement Learning for Demonstration-Free Humanoid Locomotion
Liu, Shuai
Lau, Meng Cheng
Robotics
Artificial Intelligence
We introduce Reduced-Order Model-Guided Reinforcement Learning (ROM-GRL), a two-stage reinforcement learning framework for humanoid walking that requires no motion capture data or elaborate reward shaping. In the first stage, a compact 4-DOF (four-degree-of-freedom) reduced-order model (ROM) is trained via Proximal Policy Optimization. This generates energy-efficient gait templates. In the second stage, those dynamically consistent trajectories guide a full-body policy trained with Soft Actor--Critic augmented by an adversarial discriminator, ensuring the student's five-dimensional gait feature distribution matches the ROM's demonstrations. Experiments at 1 meter-per-second and 4 meter-per-second show that ROM-GRL produces stable, symmetric gaits with substantially lower tracking error than a pure-reward baseline. By distilling lightweight ROM guidance into high-dimensional policies, ROM-GRL bridges the gap between reward-only and imitation-based locomotion methods, enabling versatile, naturalistic humanoid behaviors without any human demonstrations.
title Reduced-Order Model-Guided Reinforcement Learning for Demonstration-Free Humanoid Locomotion
topic Robotics
Artificial Intelligence
url https://arxiv.org/abs/2509.19023