Safe UAV Navigation through Wind-Aware Trajectory Planning
Mehnaz Sharna, Shafi Al Salman Romeo · 2024
Safe trajectory planning for small unmanned aerial vehicles in cluttered environments and unpredictable wind gusts is critical for a wide range of applications, including agriculture surveillance and mapping, package delivery, and rescue missions. Existing approaches typically plan trajectories assuming either no wind or a constant wind velocity. Robust internal loop controllers are then utilized to ensure that the path is followed precisely even in the presence of actual wind. In this paper, we present a strategy for obtaining a collision-free path in the presence of wind uncertainty. We create a dynamically feasible trajectory utilizing several types of trajectory planners (Kino-FMT, CC-RRT, etc.) that take into account the real wind model in order to predict deviations from planned path segments. By simulating a quadrotor performing collision avoidance, the practicable trajectory is verified. The simulation results show that the developed technique can efficiently compute wind-aware pathways for navigation with the best inputs to reach the desired position.