基于人工智能算法的无人机自动控制研究

Research of Automatic Control of UAV Based on Artificial Intelligence Algorithm

提出了基于深度强化学习的无人机控制(DRLUAVC)策略,设计了新颖的价值函数作为性能指标,并设计了基于Critic-Actor神经网络的强化学习算法来计算价值函数和确定最优控制,提出了事件触发机制以减少无人机的计算开销。使用仿真平台进行实验,实验结果表明,提出的策略能实现高效的自动控制,降低计算开销。

A drone control(DRLUAVC)strategy based on deep reinforcement learning was proposed.A novel value function was designed as the performance indicator and a reinforcement learning algorithm based on Critic-Actor neural network was designed to calculate the value function and determine the optimal control.An event trigger mechanism was proposed to reduce the computational overhead of UAVs.The simulation platform was used for experiments and the experimental results showed that the proposed strategy can realize efficient automatic control and reduce computational overhead.