Zhouheng Li

I am currently a 3rd year Ph.D. student in the College of Control Science and Engineering at Zhejiang University , Hangzhou, China, under the supervision of Prof. Lei Xie and Prof. Hongye Su. Previously, I had the privilege of being mentored by Prof. Yilun Du from Harvard University. ✨ Enjoyments of life: 🎲 Board Games (Splendor, Seven Wonders: Duel, etc), πŸ‘£ hiking, 🎾 tennis, πŸ“ ping-pong, πŸ—ΊοΈ traveling.

🎯 Research

My ultimate goal is to develop embodied intelligent autonomous systems capable of seamlessly interacting with the physical world. (πŸ“ Publications). To achieve this, my research focuses on motion planning techniques driven by physics-informed generative models, enabling safe generalization to out-of-distribution (OOD) scenarios. Currently, I am exploring planning strategies for both autonomous vehicle racing and drone racing, as well as advancing world models, with particular emphasis on the following key areas:
⭐ End-to-End Motion Planning for Unmanned Systems: End-to-end approaches can directly learn motion patterns from data distributions, effectively bypassing the need for complex planning and control (PnC) rules. Conditional diffusion models, informed by physical and task-specific information, enable stable and safe generalization in motion generation, thus providing strong guarantees for overall system performance.
⭐ Multimodal Generative Models: Multimodal inputs form the foundation of intelligent systems. By leveraging classifier-free guidance (CFG) for model composition, diverse modalitiesβ€”such as images, language, and statesβ€”can be efficiently integrated, enabling the embodied intelligent motion generation.
⭐ Physics-Informed World Models: At present, world models still face challenges in maintaining temporal consistency during long-horizon sequence generation. By incorporating physical knowledge, world models can generate in latent spaces that adhere to physical laws, thereby ensuring stability and consistency in their predictions.
I am also actively involved in applying these techniques to Roboracer competition. If any of these topics caught your interest, feel free to drop me emails (πŸ“¨ zh_li@zju.edu.cn). I enjoy collaborating on interesting projects and making amazing things happen together!

πŸš€ Spotlights

To demonstrate the effectiveness and robustness of my approaches, I showcase some demos that address distinct challenges: navigation, which emphasizes complex task and motion constraints, and racing, which requires highly integrated trajectory planning and control.

End2End Navigation

Static Model replay
Dynamic Model replay
Composed Result replay
This diffusion composition approach guarantees safe, collision-free trajectory planning in unseen scenes for safety. It can make test-time decisions to generate safe behaviors, such as accelerating to bypass or decelerating to avoid obstacles, thus allowing safe trajecoty planning for diverse scenes without retraining.

Vehicle Racing

Implementation of the proposed VPMPCC planner on the F1TENTH platform. The safe and aggressive cornering velocity with in the sharp U-turn is 11.5 km/h.

racing


πŸ”₯ News

πŸ“ Selected Publications

    * means equal contribution.
  • Rapid and Safe Trajectory Planning over Diverse Scenes through Diffusion Composition

    Wule Mao*, Zhouheng Li*, Yunhao Luo, Yilun Du, Lei Xie

  • A Data-Driven Aggressive Autonomous Racing Framework Utilizing Local Trajectory Planning with Velocity Prediction

    Zhouheng Li, Bei Zhou, Cheng Hu, Lei Xie, Hongye Su

  • A rapid iterative trajectory planning method for automated parking through differential flatness

    Zhouheng Li, Lei Xie, Cheng Hu, Hongye Su

  • Reduce Lap Time for Autonomous Racing with Curvature-Integrated MPCC Local Trajectory Planning Method

    Zhouheng Li, Lei Xie, Cheng Hu, Hongye Su

  • An aggressive cornering framework for autonomous vehicles combining trajectory planning and drift control

    Wangjia Weng, Cheng Hu, Zhouheng Li, Hongye Su, Lei Xie

  • Adaptive Learning-based Model Predictive Control Strategy for Drift Vehicles

    Bei Zhou, Cheng Hu, Jun Zeng, Zhouheng Li, Johannes Betz, Lei Xie, Hongye Su

  • A Learning-based Planning and Control Framework for Inertia Drift Vehicles

    Bei Zhou, Zhouheng Li, Lei Xie, Hongye Su, Johannes Betz

  • Learning to Race in Extreme Turning Scene with Active Exploration and Gaussian Process Regression-based MPC

    Guoqiang Wu*, Cheng Hu*, Wangjia Weng, Zhouheng Li, Yonghao Fu, Lei Xie, Hongye Su

  • An Overtaking Trajectory Planning Framework Based on Spatio-temporal Topology and Reachable Set Analysis Ensuring Time Efficiency

    Wule Mao, Zhouheng Li, Lei Xie, Hongye Su

🏁 Competitions


πŸ€— So Glad You're Here!

I truly appreciate everyone who takes the time to visit my homepage :)