arXiv：cs.AI· Aditya Azad, Arthur Huang, Guanrui Li·· 6 小时前
ClearNav：用差分物理实现感知对齐的四旋翼飞行
ClearNav: Perception-Aligned Quadrotor Flight Using Differential-Physics
AI 导读
ClearNav 是一种将运动约束在机载相机可视方向上的四旋翼导航策略，通过前视航向目标、横向速度惩罚和近障碍减速三项奖励训练。在仿真中 18 个未见场景、每个策略 2304 组配对飞行测试中，成功率提升 6.8 个百分点，碰撞率较 SOTA 降低 65%，最高飞行速度 5m/s；仅用近障碍惩罚项时碰撞率仍降低 70%。该策略还在真实杂乱环境中验证，能让四旋翼安全通过。
正文
Abstract:Autonomous quadrotor flight in cluttered environment is an key capability for tasks like inspection and search. Navigation policies trained in simulation and transferred to hardware can fly at high speed in open space, but they can still collide as obstacle clearance shrinks. We attribute most of these collisions to a single mechanism, namely motion in directions the onboard camera does not observe. Heading that is not aligned with the direction of travel and lateral excursion during obstacle avoidance both carry the vehicle toward obstacles outside its field of view. We introduce ClearNav, a navigation policy trained with two reward terms that constrain motion to the observed direction, a forward-looking heading target and a penalty on lateral velocity. A third term penalizes speed near obstacles so that an obstacle entering the field of view remains avoidable. In simulation, across 18 held-out scenarios with 2304 paired flights per policy unseen in training, ClearNav improves success by 6.8 points, cuts the collision rate by 65% relative to the state of the arts, and flies at maximum 5m/s. For the proximity term alone under matched commanded speeds, the collision reduction still holds at 70%. The effect is significant on static scenes and consistent in direction on moving obstacles, though not significant there at the suite level, showing that safer flight in clutter is possible even on the limited sensors and compute of small drones. We also validate our navigation policy in a real-world cluttered environment, showing that it enables the quadrotor to navigate through the cluttered environment successfully and safely.
| Subjects: | Robotics (cs.RO); Artificial Intelligence (cs.AI); Machine Learning (cs.LG) |
| Cite as: | arXiv:2610.17419 [cs.RO] |
| (or arXiv:2610.17419v1 [cs.RO] for this version) | |
| https://doi.org/10.48550/arXiv.2610.17419 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Aditya Azad [view email]
[v1]
Mon, 12 Oct 2026 21:33:50 UTC (6,514 KB)
来源：arXiv：cs.AI · arxiv.org