Sim-to-real control of bipedal Cassie robot
Developed sim-to-real control strategies for the Cassie bipedal robot, focusing on trajectory optimization, feedback control, system identification and state estimation.
In this project, I collaborated with William Yang on developing sim-to-real control strategies for the Cassie bipedal robot. The work focused on bridging the simulation-to-reality gap through system identification, state estimation, and model-based control. We developed methods to transfer results from simulation to the physical robot, addressing challenges such as model mismatch, sensor noise, and actuator dynamics.
