基于强化学习的能量受限无人机通信感知轨迹规划方法OA北大核心CSTPCD

For the unmanned aerial vehicle(UAV)trajectory planning problem in UAV-assisted communication and sensing with energy constraints,the authors adopt laser wireless charging to provide additional energy for the UAV,and at the same time consider the constraints of the UAV dynamics and sensing communication rate,and establishe the moving target sensing and communication trajectory planning problem with the goal of maximizing the mutual information of moving target sensing.In order to solve the established optimization problem containing a large number of complex constraints,the original optimization problem is established as a Markov Decision Process,and the processes of UAV motion,energy change,target sensing,and base station communication are modeled as the environment space,and the UAV motor rotation speed is designed as the action space,and a deep reinforcement learning method is used for training to achieve UAV trajectory planning.Due to the consideration of UAV dynamics,the planned trajectory is more in line with the UAV motion characteristics,and the optimal control sequence obtained from the training can be directly applied to the UAV motor,which reduces the difficulty of UAV control.In the designed experimental scenarios,the mutual information of moving target sensing of the proposed method is improved by a factor of about three compared with that of the traditional optimal control method.