Papers

18

Total Citations

277

H-Index

8

About

Shangke Lyu is a leading researcher in robotics and control systems, specializing in adaptive control, human-robot interaction, and reinforcement learning for legged locomotion. His major contributions include pioneering data-driven approaches for robot control with unknown Jacobians, enabling precise manipulation without exact kinematic models—work that has garnered 57 citations. Lyu also developed the Dynamic Modularity Approach for adaptive control of robotic systems with closed architectures, a breakthrough for industrial robots (54 citations). His innovative contact force estimation technique, using Gaussian Process Adaptive Disturbance Kalman Filters, addresses imperfect dynamic models in manipulators (45 citations). Lyu’s notable achievements include the RL2AC framework for rapid online adaptive control in legged robots, enhancing robust locomotion, and GeRM, a generalist robotic model using mixture-of-experts for quadruped robots. His work on continual reinforcement learning and aerial manipulator control further underscores his impact, with over 200 total citations. Lyu’s research bridges theory and practice, offering scalable solutions for real-world robotic autonomy and safety in human-robot collaboration.

Research Focus

Key Achievements

8
H-Index
18
Papers
277
Total Citations
15
Avg Citations/Paper
🏆 Most Cited Paper
Data-driven learning for robot control with unknown Jacobian
57 citations · 2020
📈 Most Prolific Year: 2018 (4 Papers)
🤝 Key Collaborators: 34
🏛 Institutions: Beihang University, Nanyang Technological University, Westlake University

Top Papers

  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8
  9. 9
  10. 10

Key Collaborators

Contact & Links

Available for collaboration
Content generated · 13 days ago