Zhiyu Ding
Papers
5
Total Citations
230
H-Index
4
About
Zhiyu Ding is a leading researcher in mobile robot motion planning, with a focus on overcoming the challenges of navigation in complex, trapped, and human-populated environments. His most impactful contribution is the development of a **Generalized Voronoi Diagram (GVD)-based heuristic** to dramatically improve the efficiency of Rapidly-exploring Random Trees (RRTs). His seminal 2021 paper on this method has garnered **172 citations**, establishing a foundational solution to the "trap space" problem in mazes and S-shaped corridors. Building on this, Ding introduced a reusable GVD-based feature tree for fast planning in trapped environments and has since shifted his focus to **socially adaptive navigation**. His recent work on **PRTIRL and NRTIRL** (Preference/Nonlinear Reward Trajectory Inverse Reinforcement Learning) integrates obstacle avoidance learning with human-robot interaction, enabling robots to navigate crowds proactively rather than merely avoiding static obstacles. With a growing body of work that bridges geometric path planning and human-aware decision-making, Ding is shaping the next generation of autonomous mobile robots capable of safe, efficient, and socially compliant movement in real-world settings.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4PRTIRL Based Socially Adaptive Path Planning for Mobile Robots12 citations · 2022
- 5NRTIRL Based NN-RRT* Path Planner in Human-Robot Interaction Environment4 citations · 2022