High-Maneuverability Air-to-Air Interception Control for UAVs based on DRL with Shooting Mode
X.-G. Li, Zhihong Liu, Guanzheng Wang, Tianyou Liu
National University of Defense Technology
阅读操作
确认中在文库中上传 PDF 后可生成中文音频讲解。
摘要与影响
Air-to-air interception by unmanned aerial vehicles (UAVs) requires aggressive acceleration toward the target to ensure effective impact. However, conventional deep reinforcement learning (DRL) controllers designed for point-reaching tasks often suffer from an over penalty issue when the agent overshoots the target, resulting in undesirable deceleration during the terminal phase. To address this problem, this paper proposes a novel shooting mode task formulation that terminates the episode when the UAV passes through the target plane, thus eliminating incentive slowdown. A neural network controller based on the Proximal Policy Optimization (PPO) algorithm is designed to directly map state errors to bodyrates and thrust commands. Simulation experiments demonstrate that, compared to the point-reaching mode, the proposed method effectively avoids terminal deceleration and achieves a high-speed interception at 15 m/s with a precision of 0.2 m. The results validate the method’s capability to support high-maneuverability aerial interception tasks.
逐年被引趋势
暂无年度引用数据
关键指标
同类平均 = 1
同领域 · 同年份 · 同类型
Google Scholar 与 OpenAlex 的被引统计范围不同,数值存在差异属正常。
AI 辅助阅读
依据:摘要
可就本文提问;依据不足时会说明。
学术脉络
学科主题
工程Guidance and Control Systems
Reinforcement Learning in Robotics · Aerospace Engineering and Control Systems