Papers
3
Total Citations
19
H-Index
3
About
Chen Lyu is a robotics and autonomous systems researcher focused on deep reinforcement learning (DRL) for navigation in complex, partially-observable environments. His work bridges the gap between simulation and real-world deployment, particularly for mobile robots and unmanned aerial vehicles (UAVs). Lyu’s major contributions include the development of a Dueling-DDPG architecture for laser-based path planning, which improves sample efficiency and stability in obstacle-dense settings. He also pioneered a partially-observable monocular navigation system for UAVs, demonstrated in a video presentation at an AIAA conference, enabling drones to avoid obstacles using only a single camera. Additionally, Lyu introduced a virtual end-to-end learning system that leverages temporal dependencies to convert camera streams into steering commands, overcoming the limitations of frame-by-frame CNN models. With over 19 citations across his top papers, Lyu’s work has influenced the integration of DRL into practical navigation pipelines. His research is notable for its focus on real-time, low-latency control, making autonomous navigation more robust and accessible for field robotics.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3