UAV clusters should take immediate action to avoid impediments in the event of a sudden change in the external environment or a UAV malfunctioning while completing tasks. Therefore, intelligent evasion technology has emerged as an important study area in UAV cluster control. This work provides a new method that can be detached from the local extreme point to solve extreme points of equilibrium between gravitation and repulsive force in a standard artificial potential field. First, the orientation of the resultant force was modified by changing the repulsive force's composition, which caused velocity to deflect at a tiny angle, isolating the local extreme point. Second, frictional resistance was added to the UAV in order to optimize vibration and ensure that it remained balanced after reaching the desired endpoints. Furthermore, the algorithm's stability was improved by optimizing relevant parameters and imposing appropriate velocity and acceleration limits. Finally, using four calculation scenarios, the control method was simulated and tested, and the dynamic obstacle avoidance of ten UAVs in various formations was achieved. The simulation findings show that the UAV cluster can rebuild, form, and sustain formations in a dispersed way while avoiding dynamic barriers. In conclusion, this study is a valuable resource for engineers working with UAV cluster flying through formations.