# Unitree G1验证速度条件全身控制，单动作训练实现98%仿真成功率

> 原文链接：https://robotics.hellobike.com/papers/2026-11-17-2611-20888-482d30dd
> 频道：papers | 更新日期：2026-11-17 | 标签：auto, arxiv
> 核心摘要：提出速度条件全身控制策略，仅用单条随机游走数据训练即可在Unitree G1真机稳定复现人类步态风格。仿真成功率达98%，关节误差控制在参考轨迹8%以内，解决传统模仿学习依赖实时参考轨迹及真机部署难问题。

# Unitree G1验证速度条件全身控制，单动作训练实现98%仿真成功率

::: v-pre
> 本文由 AI 根据公开内容整理，可能有误，请以原文为准：[Driving the Style: Velocity-Conditioned Whole-Body Control from Motion Demonstrations](https://arxiv.org/abs/2611.20888)（arXiv 2611.20888）

## 核心内容

针对人形机器人模仿学习中对抗模仿难部署、全身跟踪依赖实时参考轨迹的痛点，研究提出一种速度条件全身控制方法。该方法利用单条随机游走动作捕捉数据训练策略，通过跟踪奖励编码风格而非向Actor输入参考轨迹。实验显示，该策略在仿真中达到至少98%成功率，并在Unitree G1人形机器人上完成所有真机测试，而传统全身跟踪与命令条件蒸馏在慢走场景下真机部署均告失败。

技术架构上，Actor网络仅接收紧凑运动指令与本体观测，特权Critic与奖励函数由参考动作监督，无需教师蒸馏。在保持训练动作风格时，策略关节位置误差仅为全身跟踪方法的8%以内，且风格覆盖率优于对抗运动先验（AMP）。对比实验中，未受保护的AMP在真机直接摔倒，而新策略在无安全吊绳条件下实现了基于速度指令的稳定风格化行走，验证了低数据依赖下的Sim-to-Real迁移有效性。

## 关键看点

- **数据效率**：仅使用单条随机游走MoCap数据训练，无需大规模动作库即可编码全身运动风格
- **真机性能**：在Unitree G1上无安全吊绳完成全部测试，仿真成功率≥98%，关节误差&lt;8%
- **架构创新**：Actor仅依赖速度指令与观测，参考轨迹仅用于奖励监督，解除运行时轨迹依赖
- **对比优势**：风格覆盖率优于AMP，解决传统全身跟踪在分布外速度下真机失效问题

## 工程与产业观察

该工作证明了人形机器人风格化控制可摆脱对海量MoCap数据及实时参考轨迹的依赖。通过将风格信息内化于奖励函数并解耦Actor输入，显著降低了Sim-to-Real门槛。这对具身智能工程化具有参考价值：在数据采集受限场景下，可通过精简的状态空间设计与奖励塑形，实现低成本、高鲁棒性的真机运动策略部署。

## 英文摘要 (Abstract)

Achieving human-like behavior on humanoid robots has advanced rapidly by leveraging motion-capture demonstrations rather than hand-crafted cost functions. Adversarial imitation supplies a style signal that lets a policy fulfill a task while adhering to a desired style, but is difficult to transfer to hardware. Whole-body tracking through reinforcement learning instead deploys reliably and imitates demonstrations faithfully, but the resulting policies are conditioned on the future reference, so they require a reference trajectory at runtime and can fail when the motion differs from the training data. We train a humanoid policy directly with whole-body tracking rewards while restricting its actor to compact locomotion commands and robot observations. Reference motion supervises the reward and privileged critic, without supplying reference joint trajectories to the actor or requiring teacher distillation. The demonstrated style is thereby encoded in the policy through the tracking reward, so the robot reproduces the whole-body style under commanded velocities. Trained on a single random-walk motion and commanded with velocity profiles from held-out motions, our policy reaches at least 98\% success in simulation and completes every trial on the Unitree G1. On held-out slow walking, whole-body tracking drops to 2\% success and command-conditioned distillation to 56\%, and both fail on the real robot. On the trained motions, the joint position error of our policy stays within 8\% of whole-body tracking, and it has a higher style coverage than adversarial motion priors (AMP), which fell on the real robot without a safety crane.

作者：Murad Dawood, Oezge Drama Berkenkamp, Antonio Gonzales Marin, Nutan Chen
:::

## 原始参考来源 (Sources)

- [arXiv 2611.20888：Driving the Style: Velocity-Conditioned Whole-Body Control from Motion Demonstrations](https://arxiv.org/abs/2611.20888)