Data Collection and Trajectory Planning for Energy-Constrained UAV Swarm: A Q-Learning based Approach

This paper studies the energy-constrained path planning of multiple unmanned aerial vehicles (UAVs) based on the data collection of ground sensors in large-scale farms to achieve crop data monitoring. To reduce the energy consumption of the UAV swarm and improve the task completion rate, when all nodes are connected, the relay UAV is dynamically allocated according to the location information of the UAV swarm. The relay UAV collects all data and flies to the base station for data transmission. According to this scenario, a multi-UAV trajectory optimization problem including task time, energy consumption and node access constraints is established. Then, an optimization algorithm for UAV swarm trajectory design is proposed. This algorithm is based on Q-learning and takes the changes in the distance between the UAV and the node, the UAV's flight angle, and the safe distance as the reward function. The Q table is updated according to the awards to guide the flight route of the UAV. The analysis shows that the proposed algorithm can effectively improve the completion rate of data collection tasks and reduce the overall energy consumption of the UAV swarm.