arXiv cs.AI· Stergios E. Bachoumas, Panagiotis Artemiadis·· 2 天前AI 评分23
SOMA：四足机器人如何用分层强化学习完成物体堆叠操作
SOMA: Stack-of-Objects Manipulation using a Quadruped Robot
AI 导读
SOMA 是一个面向四足机器人的分层强化学习框架，可实现多物体堆叠的全身非抓取式操作，并首次引入显式机身高度控制，将传统二维平面推动扩展到三维推动策略。该框架用 S5 编码器的高层操作策略预测目标跟踪指令，配合动态调整腿部站姿的底层运动策略，在仿真中单物体重定位成功率达 90.3%，双箱动态堆叠达 90.1%（位置误差 0.05 m、朝向误差 1°）。
正文 · 原文
Abstract:Executing heavy-duty physical manipulation in unpredictable environments remains a major challenge for quadrupedal robots lacking dedicated manipulators. This paper presents SOMA, a hierarchical reinforcement learning framework for whole-body non-prehensile manipulation of multi-object stacks using a quadruped robot. Unlike traditional methods restricted to two-dimensional planar pushing, SOMA integrates explicit robot base-height control, unlocking full three-dimensional pushing strategies. This capability enables the quadruped to manipulate complex payload stacks as a unified system while preventing relative sliding or tipping of upper layers. The architecture factorizes control into a high-level manipulation policy using a Simplified Structured State Space Sequence (S5) encoder to reason over long-horizon temporal history and predict bounded target tracking commands via Beta distributions, and a low-level locomotion policy that dynamically adapts leg stance configurations. In extensive simulation benchmarks, SOMA achieves a $90.3\%$ success rate for single-object reposing and up to $90.1\%$ for dynamic two-box stacks under stringent error margins ($0.05\text{ m}$ position error, $1^\circ$ orientation error). Finally, we validate zero-shot sim-to-real transfer on a Unitree Go1 quadruped robot across diverse payload masses and geometries, demonstrating active base-height adjustments to maintain traction and balance during heavy pushing tasks.
| Comments: | 9 pages |
| Subjects: | Robotics (cs.RO); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.20968 [cs.RO] |
| (or arXiv:2610.20968v1 [cs.RO] for this version) | |
| https://doi.org/10.48550/arXiv.2610.20968 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Stergios Evangelos Bachoumas [view email]
[v1]
Wed, 14 Oct 2026 21:43:14 UTC (5,239 KB)
来源：arXiv cs.AI · arxiv.org