General Extremal Field Method for Time-Optimal Trajectory Planning in Flow Fields

We present an algorithm to compute time-optimal trajectories for light vehicles in unsteady flow fields, with applications to long-range, low-power aircraft as well as underwater vehicles in ocean currents. The proposed approach aims at unifying various works from the literature on extremal fields and extends it by several features. First, we propose an exact scheme to deal with still obstacles. While being directly useful for pure obstacles, it is also of particular interest to ensure the validity of computation at the borders of the problem domain. Second, we demonstrate the method ability to deal with trajectory planning for long-range airborne missions with real weather data. Lastly, the source code, written in Python, is made open to the community to accelerate research in the domain.