Chengye Wang | 王成烨

Chengye Wang | 王成烨

he/his

I am currently pursuing my graduate studies at the School of Mechanical Engineering of Beijing Institute of Technology under the supervision of Prof. Yechen Qin

Model–Data-Driven Trajectory Planning

Research Area 03

Brief

This research explores trajectory planning method for autonomous vehicles operating in interactive, uncertain, and dynamically demanding scenarios. The goal is to generate trajectory that is not only collision-free, but also dynamically feasible for the vehicle to execute safely.

Motivation

Under extreme conditions, a dynamically infeasible trajectories may result in vehicle instability, or even accidents under extreme conditions, making model-based prediction and constraint enforcement essential for ensuring dynamic feasibility and safety. Meanwhile, interactive traffic involves uncertain and context-dependent responses that are difficult to represent using handcrafted models alone, motivating data-driven learning to acquire adaptive interaction strategies from experience. Therefore, an effective planner should integrate model-based optimization for physics-consistent motion generation with data-driven decision-making for complex and uncertain interactions.

Approach

  • Multi-coupled dynamics trajectory planning scheme for autonomous driving in extreme conditions and Real-vehicle test cooperated with BYD.

If the embedded video cannot be played, watch it directly on YouTube →

  • Onsite Autonomous Driving Algorithm Challenge 2025, 4th prize in City and High-Speed Road, vedio here.

  • Risk-Aware Decision Making and Motion Planning Framework for Interactive Autonomous Driving.

If the embedded video cannot be played, watch it directly on YouTube →

  1. Shaoyang Shi, Yu Zhang, Kai Yang, Chengye Wang, Zhenfeng Wang, and Yechen Qin, “A Risk-Aware Decision Making and Motion Planning Framework for Interactive Autonomous Driving,” ITSC 2026, accepted.
  2. Xuepeng Hu, Yu Zhang, Chengye Wang, Zhenfeng Wang, and Yechen Qin, “MCTP: A Multi-Coupled Dynamics Trajectory Planning Scheme for Autonomous Driving in Extreme Conditions,” IEEE Transactions on Automation Science and Engineering, accepted.
  3. Shaoyang Shi, Kai Yang, Chengye Wang, Dongwei Li, Yu Zhang and Yechen Qin, “Self-Evolving Reinforcement Learning Framework for Autonomous Driving via Uncertainty-Aware Collision-Risk Supervision”, IEEE Transactions on Automation Science and Engineering. (Under review, IF=7.9, Q2, TOP)

← Back to all research