Papers
16
Total Citations
170
H-Index
9
About
Jiatao Ding is a leading researcher in legged robotics, specializing in robust locomotion and control for both humanoid and quadrupedal systems. His work bridges model-based optimization and learning-based methods, with major contributions to disturbance rejection and dynamic balance. Ding pioneered an observer-based cascaded model predictive control (MPC) approach that exploits multiple balance strategies—ankle, hip, stepping, and height variation—enabling humanoid robots to withstand dynamic disturbances with unprecedented robustness. His research on curriculum-based reinforcement learning has achieved reference-free quadrupedal jumping, while his impact-aware landing framework leverages parallel elasticity for explosive, efficient motions. With over 130 citations across his top ten papers, Ding’s work is widely recognized for advancing safe, adaptive locomotion in complex environments. Notable achievements include developing walking stabilization on unknown slopes, constrained task-space imitation learning for passive safety, and novel parallel elastic actuator designs that enhance joint performance and energy efficiency. His hierarchical optimization methods for versatile reactive locomotion have set new standards in the field, making his research essential reading for anyone working on robust, real-world legged robot control.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4
- 5
- 6
- 7
- 8
- 9
- 10