An Online Attack Decision Method for Unmanned Aerial Vehicle Cluster in Uncertain Environments

Viewing stationary targets with priority order in an uncertain environment, the cooperative strike decision of distributed unmanned aerial vehicle (UAV) cluster is explored in this article. With regard to the decision for UAV cluster in dynamic environments, it has always been a difficult problem. In this article, a Q-learning-based attack decision method is proposed, and a two-layer Q-network is designed, aiming at solving the decision problem of UAVs in cruise in terms of attack time and attack plan. The upper Q network is used to find the time to attack the target during the cruise, while the lower Q network is used to determine the attack plan based on the current position of the UAV and the target after the attack time is determined. Ultimately, the effectiveness of our proposed method is validated through simulation, with comparative simulations against the dual-layer crayfish optimization algorithm (Dual-COA) demonstrating the superiority of our approach. This provides a reference for future operational endeavors of UAV clusters.