Design and Application of an Intelligent Decision-Making System for Unmanned Aerial Vehicles Based on Deep Reinforcement Learning
Year: 2025
Type: article
Abstract: This paper uses deep reinforcement learning algorithms to achieve intelligent decision-making for unmanned aerial vehicles (UAVs) to realize autonomous obstacle avoidance and target tracking of UAVs. This research proposes an improved Twin Delayed Deep Deterministic Policy Gradient (ITD3) algorithm combined with a multi-experience pool mechanism. The UAV obstacle avoidance task is modeled as a Partially Observable Markov Decision Process (POMDP) model. The experimental results show that the ITD3 algorithm demonstrates significant advantages in the simulated flight environment, with the task success rate increasing to 93.7%, and it also has excellent anti-noise interference ability and generalization.
Source: IEEE Access
Author: Xiaobo Song
Institution: Air Force Engineering University
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: gold
Estimated APC paid: $2,075
OpenAlex ID: W4414647314