Abstract: In environments characterized by high complexity and dense obstacles, the actual flight trajectories of unmanned aerial vehicles (UAVs) often deviate from predetermined optimal paths, particularly when shortest path algorithms are used for route planning. This deviation is especially pronounced in dense environments, increasing the risk of collisions with obstacles. To address this challenge, we introduce the concept of geofence buffer zones. By establishing buffer areas of specific widths around obstacles and deploying artificial potential field functions within these areas, we enhance existing path-tracking algorithms, substantially reducing the likelihood of UAV obstacle collisions. Furthermore, we conduct comparative experiments on multi-rotor UAVs in a virtually dense obstacle environment using software-in-the-loop simulation technology. The experimental conditions include no minimum speed limit, a set minimum speed limit, and various turning radii. The results quantitatively confirm the significant effectiveness of geofence buffer zones in obstacle avoidance.