Papers
3
Total Citations
193
H-Index
2
About
Haojie Shi is a robotics and control systems researcher whose work bridges the gap between advanced machine learning techniques and real-world robotic applications. His research spans two interconnected domains: data-driven control for nonlinear systems and agile locomotion for quadrupedal robots. Shi's most influential contribution, "Deep Koopman Operator With Control for Nonlinear Systems" (2022, 136 citations), addresses a fundamental challenge in control theory by leveraging Koopman operator theory to transform complex nonlinear dynamics into tractable linear representations, enabling real-time control without explicit system models. This work has garnered significant attention from both the control and machine learning communities. In the field of legged robotics, Shi has made notable strides in developing scalable locomotion frameworks. His reinforcement learning approach incorporating evolutionary trajectory generation (2022, 55 citations) elegantly tackles reward sparsity and complex dynamics in quadrupedal systems. More recently, he has explored model-based learning alternatives that improve sample efficiency and reduce the sim-to-real gap, pushing toward more practical deployment of agile motor skills. Collectively, Shi's research reflects a commitment to making intelligent autonomous systems more adaptable, efficient, and deployable in real-world environments — an increasingly vital pursuit in modern robotics research.
Research Focus
Key Achievements
Top Papers
- 1Deep Koopman Operator With Control for Nonlinear Systems136 citations · 2022
- 2
- 3