The rapid advancement of networked UAV technology has broadened its applications in logistics, disaster relief, agriculture, and military. However, precise trajectory planning in complex dynamic environments remains a key challenge. Traditional methods suffer from low computational efficiency and poor real-time response, limiting practical use. This paper proposes a deep learning-based trajectory optimization strategy for networked UAVs, integrating reinforcement learning and deep neural networks to autonomously learn environmental information and collaborative patterns. Simulation experiments in realistic scenarios validate its effectiveness: the method enhances trajectory planning accuracy and robustness while reducing computational complexity and improving real-time response. This research provides new theoretical and technical foundations for promoting networked UAV applications.
Shouting Xin (Wed,) studied this question.