I study how control and learning can help robots move, adapt, and interact with the physical world. My current work focuses on legged and humanoid locomotion.
I earned an M.S. in Aerospace Science and Technology from Beihang University in 2026, advised by Prof. Kexin Guo. Previously, I completed a dual bachelor's degree in Materials Science and Engineering at Northwestern Polytechnical University and Queen Mary University of London.
Email snrt_zzhan@buaa.edu.cn
Publications
My published work spans learning-based planning and control, as well as control-theoretic methods that make learned dynamics more robust and generalizable. Publications are ordered newest first; featured entries foreground first- and equal-author contributions. * denotes equal contribution.
Accepted by npj Robotics, 2026
A lightweight, model-agnostic physics-filtered feedback module that improves robot-policy generalization under dynamics uncertainty with limited training data.
International Conference on Robotics and Automation, 2026
A meta-adaptation framework for predicting general non-structural disturbances, followed by feedback-calibrated online adaptation.
International Conference on Robotics and Automation, 2026
A Chebyshev-polynomial approach that decouples disturbances and learns the external-input-related component online.
arXiv preprint, 2026
A trajectory-optimization approach for taming model uncertainty by minimizing residual dynamics.
International Conference on Learning Representations, 2025
Feedback neural networks improve the generalization of neural ODEs through real-time state feedback in continuous-time tasks.
International Conference on Control and Automation, 2025
A bidirectionally coupled architecture for co-optimizing motion and energy in hydrogen-powered hybrid UAVs.
International Conference on Control and Automation, 2024
Closed-loop trajectory refinement using differential flatness for safe, accurate quadrotor maneuvers without controller fine-tuning.
Research journey
My path moved from materials and aircraft design to simulation, control, and learning. Those foundations now converge in embodied intelligence and adaptive robot behavior.
2019–2023
Materials science, aircraft design, simulation, and early experiments with learning-based engineering tools.
2023–2025
Robot planning, UAV control, disturbance estimation, and feedback-informed learning for reliable autonomy.
2025–present
Legged and humanoid locomotion research as a research intern at Galbot.
Next
Pursuing a PhD focused on robots that can move, adapt, and manipulate across varied environments.
Earlier work
These projects span different fields and do not all resemble my work today. I keep them here because each one taught me how to build, model, test, or rethink a physical system—and without them, my current research path would not exist.
Multi-agent planning and tracking control for quadrotor formations under wind disturbances. Successive convexification produces collision-free trajectories, paired with a disturbance-observer-based controller.
Design and prototyping of a long-endurance Prandtl wing, including dynamic modeling, LQR control synthesis, and differential-thrust stabilization of the Dutch-roll mode.
A framework combining time-optimal trajectory optimization with dynamic-programming-based task allocation for multi-robot delivery scenarios.
A coupled porosity, heat-conduction, and sintering-stress model investigating temperature gradients and potential stress overload in large porous green bodies.
An iterative-LQR solution to optimal transportation in time-varying turbulent flow, with local Jacobians approximated using second-order central finite differences.
A ResNet-based rapid-prototyping tool trained on 10,000 XFOIL-generated samples to predict candidate airfoil shapes from target pressure distributions.
An iterative vortex-panel solver that maps CFD-derived lift loss to an equivalent decambered airfoil under attached-flow assumptions.
Team design and prototyping of a fixed-wing competition aircraft combining a 16 kg payload capacity, high lift-to-drag ratio, and strong maneuverability.