弱信息交互条件下的无人机集群决策方法

The development of unmanned systems and intelligent technology has presented a broad application prospect of UAV swarms, one of the typical applications of unmanned systems in both civilian and military fields. When the swarm size is large, however, the traditional networking communication method will be limited by bandwidth and interference, which greatly affects the cooperative combat effectiveness of UAV swarms. This paper proposes a weak information interaction UAV swarm model (WIIUSM), not relying on two-way data interaction between UAVs but achieving the desired swarm behavior by using only one-way visual perception. Firstly, this paper establishes a weak information-interaction UAV swarm model. Next, an improved genetic algorithm (IGA) is used as an optimization method for the decision model, and several simulation tests are conducted with the area search task. A comparison with the snake search method based on top-level planning reveals the effectiveness of search efficiency of the proposed method. The degradation of search effectiveness under the conditions of different proportions of UAV failure is also tested, showing the robustness of our methods compared with the snake method.