面向水培生菜机器人采摘的路径安全目标到达仿真到现实强化学习
Sim-to-Real Reinforcement Learning for Path-Safe Target Reaching in Robotic Harvesting of Hydroponic Lettuce
AI总结：
本研究提出PSA-PPO算法，在Isaac Lab训练后直接迁移至UR10e机械臂，在水培生菜采摘任务中，其到达率、成功率均优于经典规划器，且 jerk 更低。
摘要：
机器人采摘需要精确的操作以确保末端执行器的准确定位，尤其是在约束工作空间中。本研究提出了路径安全近端策略优化（Path-Safe Approach Proximal Policy Optimization，PSA-PPO），这是一种为温室中营养膜技术（Nutrient Film Technique，NFT）水培生菜的机器人采摘设计的目标条件局部到达框架。所提出的PSA-PPO保留了标准的裁剪式PPO目标，同时引入了面向采摘的策略公式，该公式整合了目标条件观测、走廊感知空间约束、基于轨迹历史的安全编码以及面向约束机器人操作的任务特定成功标准。PSA-PPO完全在Isaac Lab中进行训练，无需任何现实世界的策略微调即可成功迁移至物理UR10e机械臂。在针对150个预定目标的受控物理基准测试中，PSA-PPO达到了94.7%的最高到达率，优于Cartesian+TOTG（86.0%）、RRTConnect（88.0%）和RRT*（80.0%），同时在最终接近段相对于经典规划器将中位数RMS jerk降低了39.8%至73.0%。对200个模拟目标进行的公式消融评估表明，路径安全和侧走廊 shaping是实现采摘兼容性能的主要驱动因素：即使在保留或提高终端端点精度的情况下，移除任一术语都会大幅降低严格成功率，这证实了仅针对端点进行优化对于约束作物邻近操作而言是一个不足的目标。最后，PSA-PPO与RGB-D视觉模块和末端执行器集成，并作为完整的温室采摘系统在144次生菜采摘尝试中进行评估，实现了94.4%的视觉引导目标到达率和86.1%的完整采摘成功率。
英文摘要：
Robotic harvesting requires precise manipulation to ensure accurate positioning of the end-effector, especially in a constraint workspace. This study introduces the Path-Safe Approach Proximal Policy Optimization (PSA-PPO), a target-conditioned local reaching framework designed for robotic harvesting of nutrient film technique (NFT)-hydroponic lettuce in greenhouses. The proposed PSA-PPO preserves the standard clipped PPO objective while introducing a harvesting-oriented policy formulation that integrates target-conditioned observations, corridor-aware spatial constraints, trajectory-history-based safety encoding, and task-specific success criteria for constrained robotic manipulation. PSAPPO was trained entirely in Isaac Lab and successfully transferred to a physical UR10e robotic arm without any real-world policy fine-tuning. In a controlled physical benchmark against 150 scheduled targets, PSA-PPO achieved the highest reaching rate of 94.7%, outperforming Cartesian+TOTG (86.0%), RRTConnect (88.0%), and RRT* (80.0%), while reducing median RMS jerk by 39.8 to 73.0% relative to the classical planners during the final approach segment. A formulation ablation evaluated over 200 simulation targets demonstrates that path-safety and side-corridor shaping are the principal drivers of harvesting-compatible performance: removing either term substantially degrades strict success even when terminal endpoint accuracy is preserved or improved, confirming that optimization toward the endpoint alone is an insufficient objective for constrained crop-proximal manipulation. Finally, PSA-PPO was integrated with an RGB-D vision module and an end-effector and evaluated as a complete greenhouse harvesting system across 144 lettuce harvesting attempts, achieving 94.4% vision-guided target reaching and 86.1% complete harvesting success.
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。
