通过时间行为树引导轨迹修复学习不完美的示范
Learning from Imperfect Demonstrations via Temporal Behavior Tree-Guided Trajectory Repair
AI总结：
本文提出利用时间行为树修复不完美示范轨迹，以提升机器人控制策略的学习效果，通过生成逻辑一致且可解释的数据集，指导强化学习 agent 向任务一致的 state space 区域发展。
摘要：
从示范中学习机器人控制策略是一种强大范式，但现实数据往往次优、噪声或不完美，对模仿和强化学习构成重大挑战。本文提出一个正式框架，利用时间行为树（TBT），即扩展了信号时间逻辑（STL）的行为树语义，以在用于下游策略学习之前修复次优轨迹。给定违反 TBT 规范的示范，基于模型的修复算法修正轨迹段以满足正式约束，生成逻辑一致且可解释的数据集。修复后的轨迹用于提取潜在函数，塑造强化学习的奖励信号，引导 agent 向任务一致的 state space 区域发展，无需了解 agent 的运动学模型。我们在离散网格世界导航和连续单agent及多agent reach-avoid 任务中展示了该框架的有效性，突显了其在无法假设高质量示范的机器人学习中的潜力。
英文摘要：
Learning robot control policies from demonstrations is a powerful paradigm, yet real-world data is often suboptimal, noisy, or otherwise imperfect, posing significant challenges for imitation and reinforcement learning. In this work, we present a formal framework that leverages Temporal Behavior Trees (TBT), an extension of Signal Temporal Logic (STL) with Behavior Tree semantics, to repair suboptimal trajectories prior to their use in downstream policy learning. Given demonstrations that violate a TBT specification, a model-based repair algorithm corrects trajectory segments to satisfy the formal constraints, yielding a dataset that is both logically consistent and interpretable. The repaired trajectories are then used to extract potential functions that shape the reward signal for reinforcement learning, guiding the agent toward task-consistent regions of the state space without requiring knowledge of the agent's kinematic model. We demonstrate the effectiveness of this framework on discrete grid-world navigation and continuous single and multi-agent reach-avoid tasks, highlighting its potential for data-efficient robot learning in settings where high-quality demonstrations cannot be assumed.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
