Dynamic-Risk-Field-Guided Global–Local Coordination for Urban Low-Altitude UAV Trajectory Planning
Trajectory planning for urban low-altitude unmanned aerial vehicles (UAVs) must coordinate moving-obstacle avoidance with temporary airspace restrictions within limited maneuvering space. This study presents Dynamic-Risk-Field-Guided Global–Local Coordination (DRF-GLC), which couples a fixed static-risk-guided reference with online prediction and a composite risk field incorporating released rules. Event-triggered finite-horizon rollout selects local maneuvers, auxiliary hysteretic recovery supports reference re-entry, and a common projection–screening–fallback chain checks command admissibility. Controlled synthetic comparisons separate the effects of reference selection, local motion generation, and execution screening. In 30 paired nominal trials, DRF-GLC completed every mission without collision or physical no-fly-zone entry. Relative to A* with the same online chain, mean normalized risk-field exposure decreased by 0.0461 and mission time by 1.67 s, alongside lower no-fly-zone clearance and an uncertain dynamic-margin difference. In a separate Monte Carlo family of 60 independently randomized scenes with moving obstacles and temporary restrictions, DRF-GLC and the A*-reference configuration each completed 59 trials, compared with 37 for the calibrated reactive generator. These findings support integrated risk-and-rule coordination while showing that reference and generator choices produce distinct clearance–exposure–efficiency trade-offs.
Authors
- Yiming Li (ORCID: https://orcid.org/0009-0003-6143-5286)
- Chenfeng Zhang
- Yixuan Shan
- Honghai Zhang
Institutions
- Nanjing University of Aeronautics and Astronautics (CN)
Publication Details
- Journal
- Aerospace
- Published
- 2026-11-18
- DOI
- https://doi.org/10.3390/aerospace13111079
- Primary Topic
- Robotic Path Planning Algorithms
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- National Social Science Fund of China