Reinforcement Learning-Based Flocking Control for Aerial and Ground Vehicles
Tomoya Masaoka, Naomi Kuze · 2025
Due to the high mobility of unmanned aerial vehicles (UAVs), UAV-assisted network systems have been widely developed. Specifically, while the cooperation between aerial and ground vehicles has attracted attention in scenarios such as disaster response, the heterogeneity of vehicles has not been sufficiently considered. We consider a heterogeneous vehicular network consisting of one aerial vehicle and several ground vehicles, and propose a cooperative path planning mechanism based on reinforcement learning and flocking. Through simulation experiments, we show that the proposed mechanism can construct safe and effective vehicle paths that account for differing accessible areas.