UAV Trajectory Optimization via Improved Noisy Deep Q-Network
This is an incremental improvement for UAV navigation using deep reinforcement learning, enhancing efficiency and reliability in simulated settings.
This paper tackled UAV trajectory optimization by proposing an Improved Noisy Deep Q-Network, which achieved up to +40 higher rewards and faster convergence compared to standard DQN in a simulated grid navigation environment.
This paper proposes an Improved Noisy Deep Q-Network (Noisy DQN) to enhance the exploration and stability of Unmanned Aerial Vehicle (UAV) when applying deep reinforcement learning in simulated environments. This method enhances the exploration ability by combining the residual NoisyLinear layer with an adaptive noise scheduling mechanism, while improving training stability through smooth loss and soft target network updates. Experiments show that the proposed model achieves faster convergence and up to $+40$ higher rewards compared to standard DQN and quickly reach to the minimum number of steps required for the task 28 in the 15 * 15 grid navigation environment set up. The results show that our comprehensive improvements to the network structure of NoisyNet, exploration control, and training stability contribute to enhancing the efficiency and reliability of deep Q-learning.