From a Proximal Quadratic Programming Solver to Model Predictive Control Applications

This research focuses on two key areas:

QP Solver Development of a new Quadratic Programming (QP) solver that combine sproximal regularization and augmented Lagrangian methods. It uses Reinforcement Learning (RL) to boost speed and robustness.

MPC for Path Planning Application of Nonlinear Model Predictive Control (MPC) for path planning indrones and ground vehicles, including pilot assistance and dynamic docking maneuvers.