外观
论文arXiv 2608.229762 个月前
特权Critic训练使端到端RL中的无传感器推进器故障适应成为可能
原文链接：Privileged Critic Training Enables Sensor-Free Thruster Fault Adaptation in End-to-End RL（arXiv 2608.22976 · 发布日期：2026-08-24）
Privileged Critic Training Enables Sensor-Free Thruster Fault Adaptation in End-to-End RL
针对推进器故障难观测且缺少专用传感器的问题，提出RAFT用特权非对称Critic和循环记忆训练，并在浮动平台机器人上验证。
核心要点
- 推进器可能连续退化、失效或卡在开启，传统检测需部署时不可用的专用传感器。
- RAFT在训练中让PPO价值函数访问真实退化状态，Actor仅接收标准任务观测。
- 在最多四个同时推进器故障下，RAFT四故障成功率70.2%，接近部署时可见完整退化状态的oracle策略。
- 论文称所有代码、检查点和数据均已开源。
验证范围与局限：验证范围为一台浮动平台机器人（8个推进器、1个反作用轮）上最多四个同时推进器故障。
英文摘要 (Abstract)
Fault-tolerant navigation for thruster-actuated robots requires online adaptation to failures that are neither binary nor fully observable: thrusters may degrade continuously, fail dead, or jam stuck-open. Classical fault detection pipelines require dedicated sensors unavailable at deployment; oracle controllers that observe the true failure state are equally impractical. We show that privileged critic training is sufficient for sensor-free fault adaptation: giving the PPO value function access to the true degradation state dgt during training, while the actor receives only standard task observations, shapes a policy that compensates for failures at deployment without any dedicated fault sensing. We propose RAFT (Recurrent Asymmetric Fault Tolerant), a policy with recurrent memory trained with a privileged asymmetric critic. Evaluated on a floating-platform robot (8 thrusters, 1 reaction wheel) under up to four simultaneous thruster failures, RAFT achieves 70.2% success at four concurrent failures, closing 84% of the gap from a failure-naive baseline (4.8%) to an oracle policy that sees the full degradation state at deployment (82.4%). All code, checkpoints, and data are open-source.
作者：Ricard Marsal I Castan, Miguel A. Olivares-Méndez
来源
本页事实对应的公开页面，以原文为准。