Urban low-altitude UAV path planning by fusing an enhanced A* algorithm with an adaptive artificial potential field method

To address the challenges posed by dense obstacles, complex dynamic avoidance, and poor path continuity in trajectory planning for multi-rotor UAVs operating in intricate urban low-altitude environments, this paper presents a multi-phase trajectory planning method (A*-APF) that integrates an enhanced A* algorithm with an adaptive artificial potential field (APF). The proposed approach supports the UAV’s entire mission profile, including take-off, cruising, and landing. By constructing a 3D voxelized urban airspace model and introducing a phased advancement mechanism to establish sub-goals, together with strategies such as a dual-radius safety and obstacle avoidance model, dynamic step size adjustment, and 3D kinematic constraints, the heuristic function is optimized to enhance both path feasibility and safety. Furthermore, the method employs an improved APF combined force to guide heading direction and enables local obstacle avoidance corrections through the coupling of attractive and repulsive forces. The final path is globally and locally smoothed using cubic/quartic B-spline algorithms, ensuring continuous curvature and dynamic controllability. Validated through comprehensive comparative analyses and ablation studies, the simulation results quantitatively confirm that, compared with existing state-of-the-art improved algorithms, the proposed A*-APF method reduces average computation time by over 36% and path length by over 18%. Furthermore, the ablation experiments substantiate that the synergistic integration of the multi-phase mechanism and adaptive APF is critical for ensuring a 100% obstacle avoidance success rate and effectively responding to emergent threats.