用Transformer预测机器人难测状态，提升四足机器人跳跃能力
State Estimation Transformers for Agile Legged Locomotion
- 将状态估计建模为条件序列生成，用因果掩码Transformer预测关键状态
- 在仿真和真实机器人上，跳跃成功率与恢复控制器触发率均优于现有方法
- 适合需要高动态运动控制的四足机器人研究者参考
我们提出一种状态估计方法，可精准预测机器人难以直接获取的特权状态，从而推动四足机器人在野外执行高级动作（如跳跃）的能力边界。具体而言，我们设计了状态估计变压器（SET），将状态估计问题转化为条件序列建模。通过使用因果掩码Transformer，SET基于机器人历史状态，自动回归预测体高、速度等难以直接测量的状态，在高度动态运动中仍保持高精度。我们在低成本四足机器人Cyberdog2上评估了三种任务：跑步跳跃、跑后空翻和跑侧滑。结果表明，SET在仿真中不仅估计精度和迁移能力优于其他方法，在真实世界中也显著提升了跳跃成功率和恢复控制器触发率，验证了基于Transformer的显式状态估计器在高动态运动任务中的优势。
原文摘要 · Abstract (English)
We propose a state estimation method that can accurately predict the robot's privileged states to push the limits of quadruped robots in executing advanced skills such as jumping in the wild. In particular, we present the State Estimation Transformers (SET), an architecture that casts the state estimation problem as conditional sequence modeling. SET outputs the robot states that are hard to obtain directly in the real world, such as the body height and velocities, by leveraging a causally masked Transformer. By conditioning an autoregressive model on the robot's past states, our SET model can predict these privileged observations accurately even in highly dynamic locomotions. We evaluate our methods on three tasks -- running jumping, running backflipping, and running sideslipping -- on a low-cost quadruped robot, Cyberdog2. Results show that SET can outperform other methods in estimation accuracy and transferability in the simulation as well as success rates of jumping and triggering a recovery controller in the real world, suggesting the superiority of such a Transformer-based explicit state estimator in highly dynamic locomotion tasks.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。