Xianjin Zhu
Papers
10
Total Citations
57
H-Index
5
About
Xianjin Zhu is a robotics and control systems researcher whose work sits at the intersection of reinforcement learning, sliding mode control, and autonomous mobile robotics. His research focuses primarily on the balance and motion control of underactuated robotic platforms — particularly bicycle robots and single-track two-wheeled (STTW) robots — operating in challenging, unstructured environments such as curved pavements, rough terrain, and narrow mountain passages. Zhu's most significant contributions involve developing intelligent adaptive controllers that combine reinforcement learning with classical control strategies. His pioneering work on online reinforcement-learning-based adaptive terminal sliding mode control addresses the notoriously difficult problem of stabilizing underactuated robots under matched and mismatched disturbances, earning 12 citations since 2022. He has also advanced sim-to-real transfer techniques, proposing action mapping and state prediction frameworks that help bridge the gap between simulation training and real-world deployment — a critical challenge in applied deep reinforcement learning. Collectively accumulating over 57 citations across a focused body of work, Zhu has demonstrated a consistent ability to push the boundaries of autonomous robot control under realistic constraints. His research holds particular promise for applications in search-and-rescue operations, military reconnaissance, and terrain exploration, where compact, agile robotic platforms must navigate demanding environments with minimal human intervention.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3Natural Residual Reinforcement Learning for Bicycle Robot Control9 citations · 2021
- 4
- 5
- 6
- 7
- 8
- 9
- 10Machine Learning and Formal Methods (Dagstuhl Seminar 17351)2 citations · 2018