Zheng Gong

Email: zhgong (at) ucsd.edu

Zheng Gong is a Ph.D. student in the Jacobs School of Engineering at the University of California, San Diego. His research interest lies in safety-critical systems. Currently, he focuses on finding connections between Control-Lyapunov Functions and Hamilton-Jacobi Reachability Analysis. He also has a broad interest in reinforcement learning and robotics.

Before joining UCSD, he received his master’s degree from Duke University, May 2021, where He studied stability for LTI systems with time delays. In June 2019, he completed his bachelor’s degree at the Beijing Institute of Technology.

In his free time, he enjoys playing basketball, video games (CIV 5/6, 2K22, RDR2), and cooking (home-style Chinese food).

Selected Projects

Control Lyapunov Value Functions: A constructive and general method for producing control Lyapunov-like functions that stabilize to an equilibrium (or control invariant set) at a user-defined rate  

Decomposition (+ Composition) of Value Functions: Decomposing coupled dynamic models to generate smaller value functions in each subspace that can be composed to find the backward reachable set (BRS) for a goal or an obstacle

Safety in Reinforcement Learning: Improving the optimization procedure for reinforcement learning with hard and soft safety constraints using reachability estimation

Related Papers

*co-first authors