Hierarchical Path Planning and Safety Mechanisms for UAV Multi-target Inspection in Complex Low-Altitude Environments

A hierarchical safety framework is proposed for UAV multi-target inspection in complex low-altitude environments. The framework integrates: (1) multi-objective TSP for global task planning; (2) an improved DWA scheme for local path planning; (3) APF/PN velocity fusion for cooperative static–dynamic threat avoidance; and (4) a discrete-time CBF layer with multi-level escape strategies for formal safety protection. A unified environmental model with static and dynamic obstacles is constructed, and extensive Monte Carlo simulations under different map complexities and ablation settings are carried out. Results show task success rate improvements of 22.7%/13.4%/43.9% and path efficiency gains of 14.8%/11.0%/16.2% over APF/DWA/RRT.