Zhelin Zhang
Papers
2
Total Citations
9
H-Index
2
About
Zhelin Zhang is a rising researcher in legged robotics, specializing in deep reinforcement learning (DRL) and whole-body control for multi-legged locomotion. His work addresses the critical challenge of enabling high-dimensional systems—such as hexapod robots—to learn stable, efficient gaits despite the sim-to-real gap. In his most-cited paper (2024, 7 citations), Zhang introduced an imitation-enhanced DRL framework with a privileged smooth transition mechanism, significantly improving policy learning for hexapod locomotion by leveraging expert demonstrations to guide exploration. His 2025 follow-up (2 citations) tackles safety and explainability in real-world deployment, proposing a whole-body constrained learning approach that integrates hierarchical optimization to enforce joint limits and stability constraints, reducing risky behaviors in sim-to-real transfer. Though early in his career, Zhang’s work is gaining traction for its practical focus on bridging simulation and reality—a key bottleneck in legged robotics. His contributions are particularly notable for advancing control strategies in underactuated, high-degree-of-freedom systems, offering a pathway toward more reliable and interpretable autonomous robots.
Research Focus
Key Achievements
Top Papers
- 1
- 2