Unmanned aerial vehicle cluster collaborative obstacle avoidance method based on deep reinforcement learning

The invention discloses an unmanned aerial vehicle cluster collaborative obstacle avoidance method based on deep reinforcement learning. According to the invention, based on the obstacle avoidance result of each unmanned aerial vehicle in the unmanned aerial vehicle cluster, the obstacle avoidance result is used as the initial obstacle avoidance result, and the whole cluster is regarded as an overall state for deep reinforcement learning training, so that a more complete result is obtained. The reward function comprises the distance between the unmanned aerial vehicle cluster and a target position after obstacle avoidance, the formation recovery degree after obstacle avoidance of the unmanned aerial vehicle cluster, and the difference between the obstacle avoidance direction of the unmanned aerial vehicle determining the initial obstacle avoidance direction in the unmanned aerial vehicle cluster and the initial obstacle avoidance direction of the unmanned aerial vehicle, which is called the difference degree