PPO-Exp: Keeping Fixed-Wing UAV Formation with Deep Reinforcement Learning

Flocking for fixed-Wing Unmanned Aerial Vehicles (UAVs) is an extremely complex challenge due to fixed-wing UAV’s control problem and the system’s coordinate difficulty. Recently, flocking approaches based on reinforcement learning have attracted attention. However, current methods also require that each UAV makes decision decentralized, which increases cost computation of whole system. This paper researches a low-cost formation system consisting one leader (equipped with intelligence chip) five followers (without chip), proposes centralized collision-free formation-keeping method. The communication in process considered protocol designed by minimizing cost. In addition, analysis Proximal Policy Optimization (PPO) algorithm provided; derives estimation error bound, reveals relationship between bound exploration. To encourage agent balance their exploration version PPO named PPO-Exploration (PPO-Exp) proposed. It can adjust clip constraint parameter make mechanism more flexible. results experiments show PPO-Exp performs better than algorithms these tasks.