用语义统一程序让机器人任务自动从控制转为学习，无需人工奖励。
SUN: Persistent Programs For Language-Grounded Control-to-Learning-to-Real Policies
- 构建可编译的语义程序，统一控制与学习的几何和接触关系
- 在9个任务中成功率82.03%，比基线高出近50个百分点
- 无需演示或密集奖励，适合长时序机械臂操作研究
将基于模型的控制与学习策略在长时序操作中结合，存在隐性矛盾：控制执行明确目标，学习策略将其行为泛化为反应式策略，但现有方法忽略任务语义，导致奖励需手工设计且行为偏离控制验证结果。本文提出语义统一（SUN）程序，一种类型化的可执行代码，其中几何与接触关系仅定义一次，即可编译为对齐的模型预测控制（MPC）代价、满足谓词、强化学习奖励、转移守卫和诊断信息。由大视觉语言系统驱动的Kuafu系统，能从语言和场景语义自动生成SUN程序，通过MPC筛选可行性，并在训练阶段条件化策略时保留语义。在九项任务中，Kuafu实现82.03%宏成功率，显著优于稀疏奖励（35.67%）和阶段-行为克隆（24.75%）基线。在8192种规模下，每小时人类遥操作生成的可行轨迹时间是基线的10.57倍。每个任务500条轨迹训练的DP3策略，在仿真中成功率达46.0%（对比其他方案22.4%），在真实Franka与Kinova机器人上达34.7%。结果表明，经仿真筛选的任务语义可有效将控制转化为鲁棒策略，无需示范或人工密集奖励，统一符号规划与数据驱动执行。
原文摘要 · Abstract (English)
Bridging model-based control and learned policies in long-horizon manipulation has harbored a silent disagreement: control executes specified objectives, learning amortizes that behavior into a reactive policy, yet existing protocols discard task semantics, leaving rewards hand-crafted and behavior drifting from what control verified.We introduce Semantically UNified (SUN) Programs, typed executables where geometric and contact relations are defined once and compiled into aligned Model Predictive Control (MPC) costs, satisfaction predicates, RL rewards, transition guards, and diagnostics. Our system, Kuafu, driven by large vision language systems, automatically synthesizes SUN Programs from language and scene semantics, screens feasibility via MPC, and retains semantics while training stage-conditioned policies. Across nine tasks, Kuafu achieves 82.03% macro-success, outperforming sparse-reward (35.67%) and Stage-BC (24.75%) baselines. At 8192-way scale, it generates 10.57x the successful trajectory time per hour of human teleoperation. With 500 trajectories per task, Kuafu data trains DP3 policies to 46.0% simulation success (vs. 22.4% for alternatives) and 34.7% on physical Franka and Kinova robots. These results establish that simulation-screened task semantics can effectively amortize control into robust policies, without demonstrations or manual dense rewards, unifying symbolic planning and data-driven execution.
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。