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
Top Papers
- 1Data-driven learning for robot control with unknown Jacobian57 citations · 2020
- 2
- 3
- 4Human–Robot Interaction Control Based on a General Energy Shaping Method29 citations · 2019
- 5
- 6
- 7
- 8GeRM: A Generalist Robotic Model with Mixture-of-experts for Quadruped Robot10 citations · 2024
- 9Continual Reinforcement Learning for Quadruped Robot Locomotion8 citations · 2024
- 10Human-guided Optical Manipulation of Multiple Microscopic Objects8 citations · 2018