Implement an additional robust control layer on top of the nominal Model Predictive Control (MPC), formulate a convex optimization problem subject to Control Lyapunov Function LMI to guarantee the safety and stability in 2D.

First successful hardware experiment with the MPC, though it clearly requires further tuning. Working on hardware feels different and more challenging compared to the high-fidelity simulations in MuJoCo.

Implement a Model Predicted Control strategy in MuJoCo-C, using convex optimization solved with qpSWIFT, and employ a Kalman Filter for velocity estimation to achieve basic forward and sideways trotting gaits.

This Model Predictive Control using SRB was applied on 2D Unitree A1 model in MuJoCoPy to realize quadruped uncertain terrain robust walking.

This is Model Predictive Control in high level using differential car as the model to realize the leminscate curve tracking. The high level trot gait controller was developed by Dr. Bhounsule where the robot's trotting velocity and direction can be directly manipulated.

This is my first project in robotics field. A simple Raibert Controller to realize the single serial leg robot hopping horizontal speed and vertical height control.