Shijie Yuan

Shanghai, China

I am currently an independent research focused on reinforcement learning and motion planning for robotic systems. I received my M.E. in Mechanical Engineering from Hunan University in 2024 and my B.E. in Vehicle Engineering from Dalian University of Technology in 2021.

At Li Auto (2024 – present), I developed learning-based trajectory planning modules that have been deployed in millions of production vehicles. My research interests include safe reinforcement learning, risk-aware motion planning, and constrained optimization for complex driving scenarios. My long-term goal is to develop safe, intelligent, and trustworthy robotic systems that assist human beings, augment human capabilities, and improve quality of life.

Education

2021.09–2024.06

Hunan University

M.E. in Mechanical Engineering · GPA: 3.57

Thesis: The Trajectory Planning Method for Intelligent Vehicles Using Reinforcement Learning and Risk-Reachable Sets

2017.09–2021.06

Dalian University of Technology

B.E. in Vehicle Engineering · GPA: 3.45

Excellent Graduate of Dalian University of Technology

Projects

The Trajectory Planning Method for Intelligent Vehicles Using Reinforcement Learning and Risk-Reachable Sets

Safe RL and constrained optimization for structured-road trajectory planning.

Builds uncertainty-aware risk-reachable sets.
Uses PPO-Lagrangian to generate a safe initial trajectory.
Refines the trajectory with constrained iLQR.

Automated Valet Parking

Hybrid A* and optimization-based trajectory planning in narrow parking spaces.

Searches for a collision-free path with Hybrid A*.
Smooths the path with cubic interpolation.
Plans speed and control with IPOPT.

Human-Aligned Safe RL for Highway On-Ramp Merging

Preference-aware constrained RL, action shielding, and MPC for safe highway merging.

Uses discrete SAC for merging decisions.
Rejects unsafe actions with a trajectory shield.
Executes smooth vehicle motion with MPC.

View all projects →

news

May 28, 2026	Our paper on preference-aware safe reinforcement learning for on-ramp merging was published in Machines.
Jul 01, 2024	Joined Li Auto as an Autonomous Driving Algorithm Engineer, working on learning-based planning for urban driving.
Jun 30, 2024	Received my M.E. in Mechanical Engineering from Hunan University.
Dec 19, 2022	Our paper on optimal trajectory planning in narrow corridors was published in Electronics.
Jun 30, 2021	Received my B.E. in Vehicle Engineering from Dalian University of Technology, with the honor of Excellent Graduate.

selected publications

Adaptive Constraint Regulation for Human Preference-Aware Safe Reinforcement Learning of On-Ramp Merging

Jingjia Teng, Wenjie Huang, Shijie Yuan, and 5 more authors

Machines, May 2026

Abs DOI Bib HTML Code

Safe reinforcement learning for autonomous driving requires balancing safety constraints with driving efficiency. This paper proposes an adaptive constraint regulation framework for human preference-aware safe reinforcement learning in highway on-ramp merging scenarios. The method dynamically adjusts constraint limits according to traffic conditions and driver risk preferences, combined with an action shielding mechanism that filters unsafe decisions before execution. A hierarchical architecture separates high-level maneuver decisions from low-level model predictive control, ensuring both safety compliance and driving efficiency. Simulation experiments across varying traffic densities demonstrate that the proposed approach achieves a superior safety–efficiency trade-off compared to fixed-constraint baselines, maintaining collision-free merging while improving travel efficiency.
@article{teng2026adaptive, title = {Adaptive Constraint Regulation for Human Preference-Aware Safe Reinforcement Learning of On-Ramp Merging}, author = {Teng, Jingjia and Huang, Wenjie and Yuan, Shijie and Hu, Manjiang and Qin, Hongmao and Li, Yang and Bian, Yougang and Li, Bai}, journal = {Machines}, volume = {14}, number = {6}, pages = {605}, year = {2026}, month = may, publisher = {MDPI}, doi = {10.3390/machines14060605}, }
Space Discretization-Based Optimal Trajectory Planning for Automated Vehicles in Narrow Corridor Scenes

Biao Xu, Shijie Yuan, Xuerong Lin, and 3 more authors

Electronics, Dec 2022

Abs DOI Bib HTML

Autonomous vehicles operating in narrow corridor scenarios face challenging motion planning problems with tight spatial constraints. This paper presents a space-discretization-based optimal trajectory planning method for automated vehicles in narrow corridors. The approach formulates the planning problem as a quadratic programming optimization that incorporates vehicle kinematic constraints, boundary constraints, comfort requirements, and driving efficiency objectives within a discretized space representation. The method generates smooth, dynamically feasible trajectories that safely navigate constrained environments while optimizing for passenger comfort and travel time. The proposed framework is validated through extensive simulations and real-world field experiments, demonstrating reliable trajectory generation in various narrow-corridor configurations.
@article{xu2022space, title = {Space Discretization-Based Optimal Trajectory Planning for Automated Vehicles in Narrow Corridor Scenes}, author = {Xu, Biao and Yuan, Shijie and Lin, Xuerong and Hu, Manjiang and Bian, Yougang and Qin, Zhaobo}, journal = {Electronics}, volume = {11}, number = {24}, pages = {4239}, year = {2022}, month = dec, publisher = {MDPI}, doi = {10.3390/electronics11244239}, }