Environment-adaptive multi-UAV task allocation and path planning

Light and small UAVs have become a research hot spot in recent years because of their good adaptability and low cost. Trajectory planning is the key to autonomous UAV flight, and the path planning problem in the case of multi-UAV cooperation can be divided into two sub-problems, task allocation and trajectory planning. In this paper, channel features are extracted from two dimensions, time and network type, respectively, through the graph structure. Taking this feature as a part of the environmental factors, an MDP model is established to perform task allocation and path planning with the goal of maximizing the information throughput per unit of energy. Finally, a large number of simulation tests are carried out to verify the stable performance of the algorithm in various environments.