Denglin Cheng

Master’s Student in Robotics @ Johns Hopkins University.
Robotics • Optimization • Control • Estimation

selfie.jpg

About Me

I am a Master’s student in Robotics (MSE) at Johns Hopkins University, advised by Prof. Jin-Seob Kim. Before that, I completed my bachelor’s degree in Mechanical Engineering at Zhejiang University.

During the Spring and Summer of 2025, I was fortunate to work as a research assistant at the WELL Lab at the University of Wisconsin–Madison, with Prof. Xiaobin Xiong and Jiarong Kang on legged-robot state estimation and calibration.

My primary research goal is to enable dynamic robots to autonomously generate and robustly execute complex dynamic behaviors in changing environments, by leveraging model-based optimization and optimal control.

Research Interests

Optimization and numerical methods for large-scale, structured optimal control/estimation problems;

Contact-aware and contact-implicit hybrid dynamical models for legged robots (All models are wrong!);

Bridging model-based optimal control with reinforcement learning, e.g. Actor-critic with Trajectory Optimization.

news

Jul 21, 2025 Preliminary result of our work Covariance Calibration for State Estimation on Mobile Robots via Bi-level Optimization will be presented at TC Virtual Poster Session and Networking Event 2025. :smile:

selected publications

  1. ICRA
    legged-bilevel.webp
    Simultaneous Calibration of Noise Covariance and Kinematics for State Estimation of Legged Robots via Bi-level Optimization
    Denglin Cheng*, Jiarong Kang*, and Xiaobin Xiong
    IEEE International Conference on Robotics and Automation (ICRA), 2026, Under review, Sep 2025