Two-step Planning of Dynamic UAV Trajectories Using Iterative $$\delta $$-Spaces

UAV trajectory planning is often done in a two-step approach, where low-dimensional path refined to dynamic trajectory. The resulting trajectories are only locally optimal, however. On the other hand, direct higher-dimensional state spaces generates globally optimal solutions but time-consuming and thus infeasible for time-constrained applications. To address this issue, we propose $$\delta $$ -Spaces, pruned high-dimensional space representation refinement. It does not contain area around single lower-dimensional consists of union multiple near-optimal paths. Thus, it less prone local minima. Furthermore, an anytime algorithm using -Spaces increasing sizes. We compare our method against state-of-the-art search-based methods evaluate 2D 3D environments generate second-order third-order trajectories.