一、开题答辩题目：Deep Reinforcement Learning control for Line-of-sight rate stabilization and target tracking using gimbaled camera on a UAV Platform.
二、开题答辩人：Khan Dawood Ahmed
三、开题答辩时间：June 10, 2026, 11:00
四、开题答辩地点：长安校区 自动化学院楼 323
五、开题答辩内容简介：This research presents a unified deep reinforcement learning (DRL) framework for three-axis UAV-mounted gimbaled cameras, focusing on line-of-sight stabilization and dynamic target tracking in real-world conditions. Unlike conventional methods that separate stabilization and tracking or use simplified models, our approach learns continuous torque commands for yaw, pitch, and roll through a continuous-state, continuous-action RL formulation. The reward function jointly optimizes target-centering accuracy and minimizes angular rates.
六、开题答辩人简介
Khan Dawood Ahmed, male is a doctoral candidate in Control Science and Engineering at the School of Automation, Northwestern Polytechnical University, class of 2024. His main research interests are multi-axis gimbal control.
