Flocking and Collision Avoidance for a Dynamic Squad of Fixed-Wing UAVs Using Deep Reinforcement Learning
Chao Yan,Xiaojia Xiang,Chang Wang,Zhen Lan,Chao Yan,Xiaojia Xiang,Chang Wang,Zhen Lan
Developing the flocking behavior for a dynamic squad of fixed-wing UAVs is still a challenge due to kinematic complexity and environmental uncertainty. In this paper, we deal with the decentralized flocking and collision avoidance problem through deep reinforcement learning (DRL). Specifically, we formulate a decentralized DRL-based decision making framework from the perspective of every follower,...
