Intercepting enemy UAV in a low-altitude flight environment is a challenging task. Imitation learning provides a feasible solution for such challenge. Generative Adversarial Imitation Learning (GAIL) can learn expert strategies without a carefully designed reward function. In this paper, we propose a GAIL-based method to improve the autonomous interception of UAVs problem. In the method, we augment the expert data and introduce semantic condition into the Generative Adversarial Network (GAN) to improve learning efficiency. In addition, in the designed interactive environment, we use the proposed method to complete the UAV interception decision-making task and compare with two baseline methods by evaluating the interception performance through three performance metrics. The simulation results show that our method is better than all baseline methods on all of the three metrics.
Autonomous UAV Interception via Augmented Adversarial Inverse Reinforcement Learning
Lect. Notes Electrical Eng.
International Conference on Autonomous Unmanned Systems ; 2021 ; Changsha, China September 24, 2021 - September 26, 2021
Proceedings of 2021 International Conference on Autonomous Unmanned Systems (ICAUS 2021) ; Chapter : 205 ; 2073-2084
2022-03-18
12 pages
Article/Chapter (Book)
Electronic Resource
English
Environment Adversarial Reinforcement Learning
AIAA | 2024
|Multiagent modeling of pedestrian-vehicle conflicts using Adversarial Inverse Reinforcement Learning
Taylor & Francis Verlag | 2023
|