Deep Reinforcement Learning for Multi-UAVs Collaborative Task Assignment in Logistic Scenarios
Xulin Wang, Yongzhao Hua, Xiwang Dong, Shuobo Wang, Zheng Zhang · 2024
This paper proposes a method based on deep reinforcement learning algorithm to solve the collaborative task assignment of multi-UAV s in logistic scenarios. Firstly, the practical logistic scenario is analyzed, a task model based on MDP is established, and the constraints of the logistic assignment problem are given. Secondly, the design and implementation of the state transition function and reward function are implemented based on the established model. This paper averages the final reward into each step to alleviates the problem of reward sparseness in task assignment problem. Then the deep reinforcement learning algorithm Soft Actor-Critic is used to solve the optimization problem. Finally, the experimental results show that the SAC algorithm takes less time to calculate than the traditional algorithm and can have higher average earnings in random environments.